diff --git a/__pycache__/track_reid.cpython-39.pyc b/__pycache__/track_reid.cpython-39.pyc index 47bda46..736b8a3 100644 Binary files a/__pycache__/track_reid.cpython-39.pyc and b/__pycache__/track_reid.cpython-39.pyc differ diff --git a/contrast/__pycache__/genfeats.cpython-39.pyc b/contrast/__pycache__/genfeats.cpython-39.pyc index ca68b22..fefdfa9 100644 Binary files a/contrast/__pycache__/genfeats.cpython-39.pyc and b/contrast/__pycache__/genfeats.cpython-39.pyc differ diff --git a/contrast/feat_extract/__pycache__/config.cpython-39.pyc b/contrast/feat_extract/__pycache__/config.cpython-39.pyc index dd9fad2..3792f4c 100644 Binary files a/contrast/feat_extract/__pycache__/config.cpython-39.pyc and b/contrast/feat_extract/__pycache__/config.cpython-39.pyc differ diff --git a/contrast/feat_extract/config.py b/contrast/feat_extract/config.py index 3bb15c9..9e9c6b3 100644 --- a/contrast/feat_extract/config.py +++ b/contrast/feat_extract/config.py @@ -62,7 +62,7 @@ class Config: # test_val = "./data/test_data_100" # test_model = "checkpoints/best_resnet18_v11.pth" - test_model = "checkpoints/zhanting_cls22_v11.pth" + test_model = "checkpoints/zhanting_res_801.pth" diff --git a/contrast/genfeats.py b/contrast/genfeats.py index ddaa509..2bb391b 100644 --- a/contrast/genfeats.py +++ b/contrast/genfeats.py @@ -136,6 +136,8 @@ def stdfeat_infer(imgPath, featPath, bcdSet=None): continue featpath = os.path.join(featPath, f"{bcd}.pickle") + # if os.path.isfile(featpath): + # continue stdbDict = {} t1 = time.time() diff --git a/contrast/feat_select.py b/contrast/input_getout_compare.py similarity index 100% rename from contrast/feat_select.py rename to contrast/input_getout_compare.py diff --git a/contrast/one2one_contrast.py b/contrast/one2one_contrast.py index 1d61aed..69f96ae 100644 --- a/contrast/one2one_contrast.py +++ b/contrast/one2one_contrast.py @@ -11,7 +11,7 @@ Created on Fri Aug 30 17:53:03 2024 标准特征提取，并保存至文件夹 stdFeaturePath 中， 也可在运行过程中根据与购物事件集合 barcodes 交集执行 2. 1:1 比对性能测试， - func: one2one_eval(resultPath) + func: one2one_eval(similPath) (1) 求购物事件和标准特征级 Barcode 交集，构造 evtDict、stdDict (2) 构造扫 A 放 A、扫 A 放 B 组合，mergePairs = AA_list + AB_list (3) 循环计算 mergePairs 中元素 "(A, A) 或 (A, B)" 相似度; @@ -32,6 +32,7 @@ import os import sys import random import pickle +import json # import torch import time # import json @@ -47,10 +48,12 @@ from datetime import datetime # from feat_inference import inference_image sys.path.append(r"D:\DetectTracking") -from tracking.utils.read_data import extract_data, read_tracking_output, read_one2one_simi, read_deletedBarcode_file - -from config import config as conf -from genfeats import model_init, genfeatures, stdfeat_infer +from tracking.utils.read_data import extract_data, read_tracking_output, read_similar, read_deletedBarcode_file +from tracking.utils.plotting import Annotator, colors +from feat_extract.config import config as conf +from feat_extract.inference import FeatsInterface +from utils.event import Event +from genfeats import gen_bcd_features IMG_FORMAT = ['.bmp', '.jpg', '.jpeg', '.png'] @@ -107,6 +110,10 @@ def creat_shopping_event(eventPath): evtType = 'other' '''================ 1. 构造事件描述字典，暂定 9 items ===============''' + + + + event = {} event['barcode'] = barcode event['type'] = evtType @@ -118,7 +125,8 @@ def creat_shopping_event(eventPath): event['back_feats'] = np.empty((0, 256), dtype=np.float64) event['front_feats'] = np.empty((0, 256), dtype=np.float64) event['feats_compose'] = np.empty((0, 256), dtype=np.float64) - event['one2one_simi'] = None + event['one2one'] = None + event['one2n'] = None event['feats_select'] = np.empty((0, 256), dtype=np.float64) @@ -145,8 +153,9 @@ def creat_shopping_event(eventPath): event['front_feats'] = tracking_output_feats if dataname.find("process.data")==0: - simiDict = read_one2one_simi(datapath) - event['one2one_simi'] = simiDict + simiDict = read_similar(datapath) + event['one2one'] = simiDict['one2one'] + event['one2n'] = simiDict['one2n'] if len(event['back_boxes'])==0 or len(event['front_boxes'])==0: @@ -215,6 +224,52 @@ def creat_shopping_event(eventPath): return event + + +def plot_save_image(event, savepath): + cameras = ('front', 'back') + for camera in cameras: + if camera == 'front': + boxes = event.front_trackerboxes + imgpaths = event.front_imgpaths + else: + boxes = event.back_trackerboxes + imgpaths = event.back_imgpaths + + def array2list(bboxes): + '''[x1, y1, x2, y2, track_id, score, cls, frame_index, box_index]''' + frame_ids = bboxes[:, 7].astype(int) + fID = np.unique(bboxes[:, 7].astype(int)) + fboxes = [] + for f_id in fID: + idx = np.where(frame_ids==f_id)[0] + box = bboxes[idx, :] + fboxes.append((f_id, box)) + return fboxes + + fboxes = array2list(boxes) + + for fid, fbox in fboxes: + imgpath = imgpaths[int(fid-1)] + + image = cv2.imread(imgpath) + + annotator = Annotator(image.copy(), line_width=2) + for i, *xyxy, tid, score, cls, fid, bid in enumerate(fbox): + label = f'{int(id), int(cls)}' + if tid >=0 and cls==0: + color = colors(int(cls), True) + elif tid >=0 and cls!=0: + color = colors(int(id), True) + else: + color = colors(19, True) # 19为调色板的最后一个元素 + annotator.box_label(xyxy, label, color=color) + + im0 = annotator.result() + spath = os.path.join(savepath, Path(imgpath).name) + cv2.imwrite(spath, im0) + + def save_event_subimg(event, savepath): ''' 功能: 保存一次购物事件的轨迹子图 @@ -224,160 +279,92 @@ def save_event_subimg(event, savepath): 子图保存次序：先前摄、后后摄，以 k 为编号，和 "feats_compose" 中次序相同 ''' cameras = ('front', 'back') - k = 0 for camera in cameras: if camera == 'front': - boxes = event['front_boxes'] - imgpaths = event['front_imgpaths'] + boxes = event.front_boxes + imgpaths = event.front_imgpaths else: - boxes = event['back_boxes'] - imgpaths = event['back_imgpaths'] + boxes = event.back_boxes + imgpaths = event.back_imgpaths + + + for i, box in enumerate(boxes): x1, y1, x2, y2, tid, score, cls, fid, bid = box - imgpath = imgpaths[i] + imgpath = imgpaths[int(fid-1)] image = cv2.imread(imgpath) + subimg = image[int(y1/2):int(y2/2), int(x1/2):int(x2/2), :] camerType, timeTamp, _, frameID = os.path.basename(imgpath).split('.')[0].split('_') - subimgName = f"{k}_cam-{camerType}_tid-{int(tid)}_fid-({int(fid)}, {frameID}).png" + subimgName = f"cam{camerType}_{i}_tid{int(tid)}_fid({int(fid)}, {frameID}).png" spath = os.path.join(savepath, subimgName) cv2.imwrite(spath, subimg) - k += 1 # basename = os.path.basename(event['filepath']) - print(f"Image saved: {os.path.basename(event['filepath'])}") - - + print(f"Image saved: {os.path.basename(event.eventpath)}") -def one2one_eval(resultPath): - - # stdBarcode = [p.stem for p in Path(stdFeaturePath).iterdir() if p.is_file() and p.suffix=='.pickle'] - stdBarcode = [p.stem for p in Path(stdBarcodePath).iterdir() if p.is_file() and p.suffix=='.pickle'] - - - '''购物事件列表，该列表中的 Barcode 存在于标准的 stdBarcode 内''' - evtList = [(p.stem, p.stem.split('_')[-1]) for p in Path(eventFeatPath).iterdir() - if p.is_file() - and p.suffix=='.pickle' - and (len(p.stem.split('_'))==2 or len(p.stem.split('_'))==3) - and p.stem.split('_')[-1].isdigit() - and p.stem.split('_')[-1] in stdBarcode - ] - - barcodes = set([bcd for _, bcd in evtList]) - - '''标准特征集图像样本经特征提取并保存，运行一次后无需再运行''' - stdfeat_infer(stdBarcodePath, stdFeaturePath, barcodes) - - '''========= 构建用于比对的标准特征字典 =============''' - stdDict = {} - for barcode in barcodes: - stdpath = os.path.join(stdFeaturePath, barcode+'.pickle') - with open(stdpath, 'rb') as f: - stddata = pickle.load(f) - stdDict[barcode] = stddata - - '''========= 构建用于比对的操作事件字典 =============''' - evtDict = {} - for event, barcode in evtList: - evtpath = os.path.join(eventFeatPath, event+'.pickle') - with open(evtpath, 'rb') as f: - evtdata = pickle.load(f) - evtDict[event] = evtdata - - - '''===== 构造 3 个事件对: 扫 A 放 A, 扫 A 放 B, 合并 ====================''' - AA_list = [(event, barcode, "same") for event, barcode in evtList] - AB_list = [] - for event, barcode in evtList: - dset = list(barcodes.symmetric_difference(set([barcode]))) - idx = random.randint(0, len(dset)-1) - AB_list.append((event, dset[idx], "diff")) - - mergePairs = AA_list + AB_list - - '''读取事件、标准特征文件中数据，以 AA_list 和 AB_list 中关键字为 key 生成字典''' - rltdata, rltdata_ft16, rltdata_ft16_ = [], [], [] - for evt, stdbcd, label in mergePairs: - event = evtDict[evt] - ## 判断是否存在轨迹图像文件夹，不存在则创建文件夹并保存轨迹图像 - pairpath = os.path.join(subimgPath, f"{evt}") - if not os.path.exists(pairpath): - os.makedirs(pairpath) - save_event_subimg(event, pairpath) - - ## 判断是否存在 barcode 标准样本集图像文件夹，不存在则创建文件夹并存储 barcode 样本集图像 - stdImgpath = stdDict[stdbcd]["imgpaths"] - pstdpath = os.path.join(subimgPath, f"{stdbcd}") - if not os.path.exists(pstdpath): - os.makedirs(pstdpath) - ii = 1 - for filepath in stdImgpath: - stdpath = os.path.join(pstdpath, f"{stdbcd}_{ii}.png") - shutil.copy2(filepath, stdpath) - ii += 1 - - ##============================================ float32 - stdfeat = stdDict[stdbcd]["feats"] - evtfeat = event["feats_compose"] - - matrix = 1 - cdist(stdfeat, evtfeat, 'cosine') - simi_mean = np.mean(matrix) - simi_max = np.max(matrix) - stdfeatm = np.mean(stdfeat, axis=0, keepdims=True) - evtfeatm = np.mean(evtfeat, axis=0, keepdims=True) - simi_mfeat = 1- np.maximum(0.0, cdist(stdfeatm, evtfeatm, 'cosine')) - rltdata.append((label, stdbcd, evt, simi_mean, simi_max, simi_mfeat[0,0])) - - - ##============================================ float16 - stdfeat_ft16 = stdfeat.astype(np.float16) - evtfeat_ft16 = evtfeat.astype(np.float16) - stdfeat_ft16 /= np.linalg.norm(stdfeat_ft16, axis=1)[:, None] - evtfeat_ft16 /= np.linalg.norm(evtfeat_ft16, axis=1)[:, None] - - - matrix_ft16 = 1 - cdist(stdfeat_ft16, evtfeat_ft16, 'cosine') - simi_mean_ft16 = np.mean(matrix_ft16) - simi_max_ft16 = np.max(matrix_ft16) - stdfeatm_ft16 = np.mean(stdfeat_ft16, axis=0, keepdims=True) - evtfeatm_ft16 = np.mean(evtfeat_ft16, axis=0, keepdims=True) - simi_mfeat_ft16 = 1- np.maximum(0.0, cdist(stdfeatm_ft16, evtfeatm_ft16, 'cosine')) - rltdata_ft16.append((label, stdbcd, evt, simi_mean_ft16, simi_max_ft16, simi_mfeat_ft16[0,0])) - - '''****************** uint8 is ok!!!!!! ******************''' - ##============================================ uint8 - # stdfeat_uint8, stdfeat_ft16_ = ft16_to_uint8(stdfeat_ft16) - # evtfeat_uint8, evtfeat_ft16_ = ft16_to_uint8(evtfeat_ft16) - - stdfeat_uint8 = (stdfeat_ft16*128).astype(np.int8) - evtfeat_uint8 = (evtfeat_ft16*128).astype(np.int8) - stdfeat_ft16_ = stdfeat_uint8.astype(np.float16)/128 - evtfeat_ft16_ = evtfeat_uint8.astype(np.float16)/128 - - - absdiff = np.linalg.norm(stdfeat_ft16_ - stdfeat) / stdfeat.size - - matrix_ft16_ = 1 - cdist(stdfeat_ft16_, evtfeat_ft16_, 'cosine') - simi_mean_ft16_ = np.mean(matrix_ft16_) - simi_max_ft16_ = np.max(matrix_ft16_) - stdfeatm_ft16_ = np.mean(stdfeat_ft16_, axis=0, keepdims=True) - evtfeatm_ft16_ = np.mean(evtfeat_ft16_, axis=0, keepdims=True) - simi_mfeat_ft16_ = 1- np.maximum(0.0, cdist(stdfeatm_ft16_, evtfeatm_ft16_, 'cosine')) - rltdata_ft16_.append((label, stdbcd, evt, simi_mean_ft16_, simi_max_ft16_, simi_mfeat_ft16_[0,0])) - +def data_precision_compare(stdfeat, evtfeat, evtMessage, save=True): + evt, stdbcd, label = evtMessage + rltdata, rltdata_ft16, rltdata_ft16_ = [], [], [] + matrix = 1 - cdist(stdfeat, evtfeat, 'cosine') + simi_mean = np.mean(matrix) + simi_max = np.max(matrix) + stdfeatm = np.mean(stdfeat, axis=0, keepdims=True) + evtfeatm = np.mean(evtfeat, axis=0, keepdims=True) + simi_mfeat = 1- np.maximum(0.0, cdist(stdfeatm, evtfeatm, 'cosine')) + rltdata = [label, stdbcd, evt, simi_mean, simi_max, simi_mfeat[0,0]] - tm = datetime.fromtimestamp(time.time()).strftime('%Y%m%d_%H%M%S') - ##================================================ save as float32, - rppath = os.path.join(resultPath, f'{tm}.pickle') + + ##================================================================= float16 + stdfeat_ft16 = stdfeat.astype(np.float16) + evtfeat_ft16 = evtfeat.astype(np.float16) + stdfeat_ft16 /= np.linalg.norm(stdfeat_ft16, axis=1)[:, None] + evtfeat_ft16 /= np.linalg.norm(evtfeat_ft16, axis=1)[:, None] + + + matrix_ft16 = 1 - cdist(stdfeat_ft16, evtfeat_ft16, 'cosine') + simi_mean_ft16 = np.mean(matrix_ft16) + simi_max_ft16 = np.max(matrix_ft16) + stdfeatm_ft16 = np.mean(stdfeat_ft16, axis=0, keepdims=True) + evtfeatm_ft16 = np.mean(evtfeat_ft16, axis=0, keepdims=True) + simi_mfeat_ft16 = 1- np.maximum(0.0, cdist(stdfeatm_ft16, evtfeatm_ft16, 'cosine')) + rltdata_ft16 = [label, stdbcd, evt, simi_mean_ft16, simi_max_ft16, simi_mfeat_ft16[0,0]] + + '''****************** uint8 is ok!!!!!! ******************''' + ##=================================================================== uint8 + # stdfeat_uint8, stdfeat_ft16_ = ft16_to_uint8(stdfeat_ft16) + # evtfeat_uint8, evtfeat_ft16_ = ft16_to_uint8(evtfeat_ft16) + + stdfeat_uint8 = (stdfeat_ft16*128).astype(np.int8) + evtfeat_uint8 = (evtfeat_ft16*128).astype(np.int8) + stdfeat_ft16_ = stdfeat_uint8.astype(np.float16)/128 + evtfeat_ft16_ = evtfeat_uint8.astype(np.float16)/128 + + absdiff = np.linalg.norm(stdfeat_ft16_ - stdfeat) / stdfeat.size + + matrix_ft16_ = 1 - cdist(stdfeat_ft16_, evtfeat_ft16_, 'cosine') + simi_mean_ft16_ = np.mean(matrix_ft16_) + simi_max_ft16_ = np.max(matrix_ft16_) + stdfeatm_ft16_ = np.mean(stdfeat_ft16_, axis=0, keepdims=True) + evtfeatm_ft16_ = np.mean(evtfeat_ft16_, axis=0, keepdims=True) + simi_mfeat_ft16_ = 1- np.maximum(0.0, cdist(stdfeatm_ft16_, evtfeatm_ft16_, 'cosine')) + rltdata_ft16_ = [label, stdbcd, evt, simi_mean_ft16_, simi_max_ft16_, simi_mfeat_ft16_[0,0]] + + if not save: + return + + + ##========================================================= save as float32 + rppath = os.path.join(similPath, f'{evt}_ft32.pickle') with open(rppath, 'wb') as f: pickle.dump(rltdata, f) - rtpath = os.path.join(resultPath, f'{tm}.txt') + rtpath = os.path.join(similPath, f'{evt}_ft32.txt') with open(rtpath, 'w', encoding='utf-8') as f: for result in rltdata: part = [f"{x:.3f}" if isinstance(x, float) else str(x) for x in result] @@ -385,12 +372,12 @@ def one2one_eval(resultPath): f.write(line + '\n') - ##================================================ save as float16, - rppath_ft16 = os.path.join(resultPath, f'{tm}_ft16.pickle') + ##========================================================= save as float16 + rppath_ft16 = os.path.join(similPath, f'{evt}_ft16.pickle') with open(rppath_ft16, 'wb') as f: pickle.dump(rltdata_ft16, f) - rtpath_ft16 = os.path.join(resultPath, f'{tm}_ft16.txt') + rtpath_ft16 = os.path.join(similPath, f'{evt}_ft16.txt') with open(rtpath_ft16, 'w', encoding='utf-8') as f: for result in rltdata_ft16: part = [f"{x:.3f}" if isinstance(x, float) else str(x) for x in result] @@ -398,42 +385,145 @@ def one2one_eval(resultPath): f.write(line + '\n') - ##================================================ save as uint8, - rppath_uint8 = os.path.join(resultPath, f'{tm}_uint8.pickle') + ##=========================================================== save as uint8 + rppath_uint8 = os.path.join(similPath, f'{evt}_uint8.pickle') with open(rppath_uint8, 'wb') as f: pickle.dump(rltdata_ft16_, f) - rtpath_uint8 = os.path.join(resultPath, f'{tm}_uint8.txt') + rtpath_uint8 = os.path.join(similPath, f'{evt}_uint8.txt') with open(rtpath_uint8, 'w', encoding='utf-8') as f: for result in rltdata_ft16_: part = [f"{x:.3f}" if isinstance(x, float) else str(x) for x in result] line = ', '.join(part) f.write(line + '\n') + +def one2one_simi(): + ''' + stdFeaturePath: 标准特征集地址 + eventDataPath: Event对象地址 + ''' + stdBarcode = [p.stem for p in Path(stdFeaturePath).iterdir() if p.is_file() and p.suffix=='.pickle'] + + '''======1. 购物事件列表，该列表中的 Barcode 存在于标准的 stdBarcode 内 ===''' + evtList = [(p.stem, p.stem.split('_')[-1]) for p in Path(eventDataPath).iterdir() + if p.is_file() + and p.suffix=='.pickle' + and (len(p.stem.split('_'))==2 or len(p.stem.split('_'))==3) + and p.stem.split('_')[-1].isdigit() + and p.stem.split('_')[-1] in stdBarcode + ] + barcodes = set([bcd for _, bcd in evtList]) + + '''======2. 构建用于比对的标准特征字典 =============''' + stdDict = {} + for barcode in barcodes: + stdpath = os.path.join(stdFeaturePath, barcode+'.pickle') + with open(stdpath, 'rb') as f: + stddata = pickle.load(f) + stdDict[barcode] = stddata + + + '''======3. 构建用于比对的操作事件字典 =============''' + evtDict = {} + for evtname, barcode in evtList: + evtpath = os.path.join(eventDataPath, evtname+'.pickle') + with open(evtpath, 'rb') as f: + evtdata = pickle.load(f) + evtDict[evtname] = evtdata + + + '''======4.1 事件轨迹子图保存 ======================''' + error_event = [] + for evtname, event in evtDict.items(): + pairpath = os.path.join(subimgPath, f"{evtname}") + if not os.path.exists(pairpath): + os.makedirs(pairpath) + try: + save_event_subimg(event, pairpath) + except Exception as e: + error_event.append(evtname) + + img_path = os.path.join(imagePath, f"{evtname}") + if not os.path.exists(img_path): + os.makedirs(img_path) + try: + plot_save_image(event, img_path) + except Exception as e: + error_event.append(evtname) + + + errfile = os.path.join(subimgPath, f'error_event.txt') + with open(errfile, 'w', encoding='utf-8') as f: + for line in error_event: + f.write(line + '\n') + + + '''======4.2 barcode 标准图像保存 ==================''' + # for stdbcd in barcodes: + # stdImgpath = stdDict[stdbcd]["imgpaths"] + # pstdpath = os.path.join(subimgPath, f"{stdbcd}") + # if not os.path.exists(pstdpath): + # os.makedirs(pstdpath) + # ii = 1 + # for filepath in stdImgpath: + # stdpath = os.path.join(pstdpath, f"{stdbcd}_{ii}.png") + # shutil.copy2(filepath, stdpath) + # ii += 1 + + '''======5 构造 3 个事件对: 扫 A 放 A, 扫 A 放 B, 合并 ====================''' + AA_list = [(evtname, barcode, "same") for evtname, barcode in evtList] + AB_list = [] + for evtname, barcode in evtList: + dset = list(barcodes.symmetric_difference(set([barcode]))) + if len(dset): + idx = random.randint(0, len(dset)-1) + AB_list.append((evtname, dset[idx], "diff")) + + mergePairs = AA_list + AB_list + + '''======6 计算事件、标准特征集相似度 ==================''' + rltdata = [] + for i in range(len(mergePairs)): + evtname, stdbcd, label = mergePairs[i] + event = evtDict[evtname] + + ##============================================ float32 + stdfeat = stdDict[stdbcd]["feats_ft32"] + evtfeat = event.feats_compose + + if len(evtfeat)==0: continue + + matrix = 1 - cdist(stdfeat, evtfeat, 'cosine') + matrix[matrix < 0] = 0 + + + simi_mean = np.mean(matrix) + simi_max = np.max(matrix) + stdfeatm = np.mean(stdfeat, axis=0, keepdims=True) + evtfeatm = np.mean(evtfeat, axis=0, keepdims=True) + simi_mfeat = 1- np.maximum(0.0, cdist(stdfeatm, evtfeatm, 'cosine')) + rltdata.append((label, stdbcd, evtname, simi_mean, simi_max, simi_mfeat[0,0])) + + '''================ float32、16、int8 精度比较与存储 =============''' + # data_precision_compare(stdfeat, evtfeat, mergePairs[i], save=True) + print("func: one2one_eval(), have finished!") + + return rltdata + -def compute_precise_recall(pickpath): - - pickfile = os.path.basename(pickpath) - file, ext = os.path.splitext(pickfile) - - if ext != '.pickle': return - if file.find('ft16') < 0: return - - with open(pickpath, 'rb') as f: - results = pickle.load(f) - +def compute_precise_recall(rltdata): Same, Cross = [], [] - for label, stdbcd, evt, simi_mean, simi_max, simi_mft in results: + for label, stdbcd, evtname, simi_mean, simi_max, simi_mft in rltdata: if label == "same": Same.append(simi_mean) if label == "diff": Cross.append(simi_mean) - - + Same = np.array(Same) Cross = np.array(Cross) TPFN = len(Same) @@ -480,115 +570,135 @@ def compute_precise_recall(pickpath): ax.set_xlabel(f"Same Num: {TPFN}, Cross Num: {TNFP}") ax.legend() plt.show() - plt.savefig(f'./result/{file}_pr.png') # svg, png, pdf + + rltpath = os.path.join(similPath, 'pr.png') + plt.savefig(rltpath) # svg, png, pdf -def gen_eventdict(eventDatePath, saveimg=True): +def gen_eventdict(sourcePath, saveimg=True): eventList = [] - # k = 0 - for datePath in eventDatePath: - for eventName in os.listdir(datePath): + errEvents = [] + k = 0 + for source_path in sourcePath: + bname = os.path.basename(source_path) + + pickpath = os.path.join(eventDataPath, f"{bname}.pickle") + if os.path.isfile(pickpath): continue + + # if bname != "20241129-100321-a9dae9e3-7db5-4e31-959c-d7dfc228923e_6972636670213": + # continue - pickpath = os.path.join(eventFeatPath, f"{eventName}.pickle") - if os.path.isfile(pickpath): - continue - eventPath = os.path.join(datePath, eventName) + + + # eventDict = creat_shopping_event(eventPath) + # if eventDict: + # eventList.append(eventDict) + # with open(pickpath, 'wb') as f: + # pickle.dump(eventDict, f) + # print(f"Event: {eventName}, have saved!") - eventDict = creat_shopping_event(eventPath) - if eventDict: - eventList.append(eventDict) - with open(pickpath, 'wb') as f: - pickle.dump(eventDict, f) - print(f"Event: {eventName}, have saved!") + # if saveimg and eventDict: + # basename = os.path.basename(eventDict['filepath']) + # savepath = os.path.join(subimgPath, basename) + # if not os.path.exists(savepath): + # os.makedirs(savepath) + # save_event_subimg(eventDict, savepath) + + try: + event = Event(source_path) + eventList.append(event) + with open(pickpath, 'wb') as f: + pickle.dump(event, f) + print(bname) + except Exception as e: + errEvents.append(source_path) + print(e) + + # k += 1 + # if k==10: + # break - # k += 1 - # if k==1: - # break - - ## 保存轨迹中 boxes 子图 - if not saveimg: - return - for event in eventList: - basename = os.path.basename(event['filepath']) - savepath = os.path.join(subimgPath, basename) - if not os.path.exists(savepath): - os.makedirs(savepath) - save_event_subimg(event, savepath) + errfile = os.path.join(eventDataPath, f'error_events.txt') + with open(errfile, 'w', encoding='utf-8') as f: + for line in errEvents: + f.write(line + '\n') def test_one2one(): - eventDatePath = [r'\\192.168.1.28\share\测试_202406\1101\images', - # r'\\192.168.1.28\share\测试_202406\0910\images', - # r'\\192.168.1.28\share\测试_202406\0723\0723_1', - # r'\\192.168.1.28\share\测试_202406\0723\0723_2', - # r'\\192.168.1.28\share\测试_202406\0723\0723_3', - # r'\\192.168.1.28\share\测试_202406\0722\0722_01', - # r'\\192.168.1.28\share\测试_202406\0722\0722_02' - # r'\\192.168.1.28\share\测试_202406\0719\719_3', - # r'\\192.168.1.28\share\测试_202406\0716\0716_1', - # r'\\192.168.1.28\share\测试_202406\0716\0716_2', - # r'\\192.168.1.28\share\测试_202406\0716\0716_3', - # r'\\192.168.1.28\share\测试_202406\0712\0712_1', # 无帧图像 - # r'\\192.168.1.28\share\测试_202406\0712\0712_2', # 无帧图像 - ] - bcdList = [] - for evtpath in eventDatePath: + bcdList, event_spath = [], [] + for evtpath in eventSourcePath: for evtname in os.listdir(evtpath): evt = evtname.split('_') + dirpath = os.path.join(evtpath, evtname) + if os.path.isfile(dirpath): continue if len(evt)>=2 and evt[-1].isdigit() and len(evt[-1])>=10: - bcdList.append(evt[-1]) - + bcdList.append(evt[-1]) + event_spath.append(os.path.join(evtpath, evtname)) + bcdSet = set(bcdList) - - - - - model = model_init(conf) - - '''==== 1. 生成标准特征集, 只需运行一次 ===============''' - genfeatures(model, stdSamplePath, stdBarcodePath, stdFeaturePath, bcdSet) + '''==== 1. 生成标准特征集, 只需运行一次, 在 genfeats.py 中实现 ===========''' + # gen_bcd_features(stdSamplePath, stdBarcodePath, stdFeaturePath, bcdSet) print("stdFeats have generated and saved!") '''==== 2. 生成事件字典, 只需运行一次 ===============''' - - gen_eventdict(eventDatePath) + gen_eventdict(event_spath) print("eventList have generated and saved!") '''==== 3. 1:1性能评估 ===============''' - one2one_eval(resultPath) - for filename in os.listdir(resultPath): - if filename.find('.pickle') < 0: continue - if filename.find('0911') < 0: continue - pickpath = os.path.join(resultPath, filename) - compute_precise_recall(pickpath) - - + rltdata = one2one_simi() + compute_precise_recall(rltdata) + if __name__ == '__main__': ''' - 共6个地址： + 共7个地址： (1) stdSamplePath: 用于生成比对标准特征集的原始图像地址 (2) stdBarcodePath: 比对标准特征集原始图像地址的pickle文件存储，{barcode: [imgpath1, imgpath1, ...]} (3) stdFeaturePath: 比对标准特征集特征存储地址 - (4) eventFeatPath: 用于1:1比对的购物事件特征存储地址、对应子图存储地址 - (5) subimgPath: 1:1比对购物事件轨迹、标准barcode所对应的 subimgs 存储地址 - (6) resultPath: 1:1比对结果存储地址 + (4) eventSourcePath: 事件地址 + (5) resultPath: 结果存储地址 + (6) eventDataPath: 用于1:1比对的购物事件特征存储地址、对应子图存储地址 + (7) subimgPath: 1:1比对购物事件轨迹、标准barcode所对应的 subimgs 存储地址 + (8) similPath: 1:1比对结果存储地址(事件级) ''' - - stdSamplePath = r"\\192.168.1.28\share\已标注数据备份\对比数据\barcode\barcode_500_1979_已清洗" - stdBarcodePath = r"\\192.168.1.28\share\测试_202406\contrast\std_barcodes_2192" - stdFeaturePath = r"\\192.168.1.28\share\测试_202406\contrast\std_features_ft32" - eventFeatPath = r"\\192.168.1.28\share\测试_202406\contrast\events" - subimgPath = r'\\192.168.1.28\share\测试_202406\contrast\subimgs' - resultPath = r"D:\DetectTracking\contrast\result\pickle" - if not os.path.exists(resultPath): - os.makedirs(resultPath) + # stdSamplePath = r"\\192.168.1.28\share\已标注数据备份\对比数据\barcode\barcode_500_1979_已清洗" + # stdBarcodePath = r"\\192.168.1.28\share\测试_202406\contrast\std_barcodes_2192" + # stdFeaturePath = r"\\192.168.1.28\share\测试_202406\contrast\std_features_ft32" + # eventDataPath = r"\\192.168.1.28\share\测试_202406\contrast\events" + # subimgPath = r'\\192.168.1.28\share\测试_202406\contrast\subimgs' + # similPath = r"D:\DetectTracking\contrast\result\pickle" + # eventSourcePath = [r'\\192.168.1.28\share\测试_202406\1101\images'] + + stdSamplePath = r"\\192.168.1.28\share\数据\已完成数据\展厅数据\v1.0\比对数据\整理\zhantingBase" + stdBarcodePath = r"D:\exhibition\dataset\bcdpath" + stdFeaturePath = r"D:\exhibition\dataset\feats" + resultPath = r"D:\exhibition\result\events" + # eventSourcePath = [r'D:\exhibition\images\20241202'] + # eventSourcePath = [r"\\192.168.1.28\share\测试视频数据以及日志\各模块测试记录\展厅测试\1129_展厅模型v801测试组测试"] + eventSourcePath = [r"\\192.168.1.28\share\测试视频数据以及日志\各模块测试记录\展厅测试\1126_展厅模型v801测试"] + + + + '''定义当前事件存储地址及生成相应文件件''' + eventDataPath = os.path.join(resultPath, "1126", "evtobjs") + subimgPath = os.path.join(resultPath, "1126", "subimgs") + imagePath = os.path.join(resultPath, "1126", "image") + similPath = os.path.join(resultPath, "1126", "simidata") + + if not os.path.exists(eventDataPath): + os.makedirs(eventDataPath) + if not os.path.exists(subimgPath): + os.makedirs(subimgPath) + if not os.path.exists(imagePath): + os.makedirs(imagePath) + if not os.path.exists(similPath): + os.makedirs(similPath) test_one2one() diff --git a/contrast/one2one_onsite.py b/contrast/one2one_onsite.py index 94c8c3e..16ecdbe 100644 --- a/contrast/one2one_onsite.py +++ b/contrast/one2one_onsite.py @@ -106,7 +106,9 @@ def test_compare(): def one2one_pr(paths): paths = Path(paths) - evtpaths = [p for p in paths.iterdir() if p.is_dir() and len(p.name.split('_'))>=2] + + # evtpaths = [p for p in paths.iterdir() if p.is_dir() and len(p.name.split('_'))>=2] + evtpaths = [p for p in paths.iterdir() if p.is_dir()] events, similars = [], [] @@ -120,14 +122,19 @@ def one2one_pr(paths): ##===================================== 应用于1:n tpevents, fnevents, fpevents, tnevents = [], [], [], [] tpsimi, fnsimi, tnsimi, fpsimi = [], [], [], [] - + other_event, other_simi = [], [] + + ##===================================== barcodes总数、比对错误事件 + bcdList, one2onePath = [], [] for path in evtpaths: barcode = path.stem.split('_')[-1] datapath = path.joinpath('process.data') if not barcode.isdigit() or len(barcode)<10: continue if not datapath.is_file(): continue - + + bcdList.append(barcode) + try: SimiDict = read_similar(datapath) except Exception as e: @@ -150,13 +157,17 @@ def one2one_pr(paths): one2oneAA.extend(simAA) one2oneAB.extend(simAB) - + one2onePath.append(path.stem) ##===================================== 以下应用适用于展厅 1：N max_idx = similars.index(max(similars)) max_sim = similars[max_idx] # max_bcd = barcodes[max_idx] + if path.stem.find('100321')>0: + print("hhh") + + for i in range(len(one2one)): bcd, simi = barcodes[i], similars[i] if bcd==barcode and simi==max_sim: @@ -172,7 +183,7 @@ def one2one_pr(paths): fp_simi.append(simi) fp_events.append(path.stem) - + ##===================================== 以下应用适用1:n events, evt_barcodes, evt_similars, evt_types = [], [], [], [] for dt in one2n: @@ -197,9 +208,13 @@ def one2one_pr(paths): elif bcd!=barcode and simi!=maxsim: tnsimi.append(simi) tnevents.append(path.stem) - else: + elif bcd!=barcode and simi==maxsim: fpsimi.append(simi) fpevents.append(path.stem) + else: + other_simi.append(simi) + other_event.append(path.stem) + '''命名规则: 1:1 1:n 1:N @@ -228,9 +243,12 @@ def one2one_pr(paths): FN_ = sum(np.array(one2oneAA) < th) TN_ = sum(np.array(one2oneAB) < th) PPrecise_.append(TP_/(TP_+FP_+1e-6)) - PRecall_.append(TP_/(TP_+FN_+1e-6)) + # PRecall_.append(TP_/(TP_+FN_+1e-6)) + PRecall_.append(TP_/(len(one2oneAA)+1e-6)) + NPrecise_.append(TN_/(TN_+FN_+1e-6)) - NRecall_.append(TN_/(TN_+FP_+1e-6)) + # NRecall_.append(TN_/(TN_+FP_+1e-6)) + NRecall_.append(TN_/(len(one2oneAB)+1e-6)) '''============================= 1:n''' TP = sum(np.array(tpsimi) >= th) @@ -238,9 +256,12 @@ def one2one_pr(paths): FN = sum(np.array(fnsimi) < th) TN = sum(np.array(tnsimi) < th) PPrecise.append(TP/(TP+FP+1e-6)) - PRecall.append(TP/(TP+FN+1e-6)) + # PRecall.append(TP/(TP+FN+1e-6)) + PRecall.append(TP/(len(tpsimi)+len(fnsimi)+1e-6)) + NPrecise.append(TN/(TN+FN+1e-6)) - NRecall.append(TN/(TN+FP+1e-6)) + # NRecall.append(TN/(TN+FP+1e-6)) + NRecall.append(TN/(len(tnsimi)+len(fpsimi)+1e-6)) '''============================= 1:N 展厅''' @@ -249,9 +270,12 @@ def one2one_pr(paths): FNX = sum(np.array(fn_simi) < th) TNX = sum(np.array(tn_simi) < th) PPreciseX.append(TPX/(TPX+FPX+1e-6)) - PRecallX.append(TPX/(TPX+FNX+1e-6)) + # PRecallX.append(TPX/(TPX+FNX+1e-6)) + PRecallX.append(TPX/(len(tp_simi)+len(fn_simi)+1e-6)) + NPreciseX.append(TNX/(TNX+FNX+1e-6)) - NRecallX.append(TNX/(TNX+FPX+1e-6)) + # NRecallX.append(TNX/(TNX+FPX+1e-6)) + NRecallX.append(TNX/(len(tn_simi)+len(fp_simi)+1e-6)) '''============================= 1:1 曲线''' fig, ax = plt.subplots() @@ -262,8 +286,8 @@ def one2one_pr(paths): ax.set_xlim([0, 1]) ax.set_ylim([0, 1]) ax.grid(True) - ax.set_title('Precise & Recall') - ax.set_xlabel(f"Num: {len(evtpaths)}") + ax.set_title('1:1 Precise & Recall') + ax.set_xlabel(f"Event Num: {len(one2oneAA)}") ax.legend() plt.show() @@ -286,8 +310,8 @@ def one2one_pr(paths): ax.set_xlim([0, 1]) ax.set_ylim([0, 1]) ax.grid(True) - ax.set_title('Precise & Recall') - ax.set_xlabel(f"Num: {len(evtpaths)}") + ax.set_title('1:n Precise & Recall') + ax.set_xlabel(f"Event Num: {len(one2oneAA)}") ax.legend() plt.show() @@ -317,8 +341,8 @@ def one2one_pr(paths): ax.set_xlim([0, 1]) ax.set_ylim([0, 1]) ax.grid(True) - ax.set_title('Precise & Recall') - ax.set_xlabel(f"Num: {len(evtpaths)}") + ax.set_title('1:N Precise & Recall') + ax.set_xlabel(f"Event Num: {len(one2oneAA)}") ax.legend() plt.show() @@ -338,16 +362,23 @@ def one2one_pr(paths): axes[1, 1].set_title('FN') plt.show() - - + # bcdSet = set(bcdList) + # one2nErrFile = str(paths.joinpath("one_2_Small_n_Error.txt")) + # with open(one2nErrFile, "w") as file: + # for item in fnevents: + # file.write(item + "\n") + + # one2NErrFile = str(paths.joinpath("one_2_Big_N_Error.txt")) + # with open(one2NErrFile, "w") as file: + # for item in fn_events: + # file.write(item + "\n") print('Done!') - - + if __name__ == "__main__": - evtpaths = r"\\192.168.1.28\share\测试视频数据以及日志\各模块测试记录\展厅测试\1120_展厅模型v801测试\扫A放A" + evtpaths = r"\\192.168.1.28\share\测试视频数据以及日志\各模块测试记录\展厅测试\1129_展厅模型v801测试组测试" one2one_pr(evtpaths) diff --git a/contrast/feat_similar.py b/contrast/seqfeat_compare.py similarity index 100% rename from contrast/feat_similar.py rename to contrast/seqfeat_compare.py diff --git a/contrast/feat_analysisi.py b/contrast/stdfeat_analys.py similarity index 100% rename from contrast/feat_analysisi.py rename to contrast/stdfeat_analys.py diff --git a/contrast/utils/__pycache__/event.cpython-39.pyc b/contrast/utils/__pycache__/event.cpython-39.pyc new file mode 100644 index 0000000..25c22c0 Binary files /dev/null and b/contrast/utils/__pycache__/event.cpython-39.pyc differ diff --git a/contrast/utils/event.py b/contrast/utils/event.py new file mode 100644 index 0000000..e28f56e --- /dev/null +++ b/contrast/utils/event.py @@ -0,0 +1,179 @@ +# -*- coding: utf-8 -*- +""" +Created on Tue Nov 26 17:35:05 2024 + +@author: ym +""" +import os +import numpy as np +from pathlib import Path + +import sys +sys.path.append(r"D:\DetectTracking") +from tracking.utils.read_data import extract_data, read_tracking_output, read_similar + +IMG_FORMAT = ['.bmp', '.jpg', '.jpeg', '.png'] +VID_FORMAT = ['.mp4', '.avi'] + +class Event: + def __init__(self, eventpath, stype="data"): + '''stype: str, 'video', 'image', 'data', ''' + + self.eventpath = eventpath + self.evtname = str(Path(eventpath).stem) + self.barcode = '' + self.evtType = '' + + '''=========== path of image and video =========== ''' + self.back_videopath = '' + self.front_videopath = '' + self.back_imgpaths = [] + self.front_imgpaths = [] + + '''=========== process.data ===============================''' + self.one2one = None + self.one2n = None + + '''=========== 0/1_track.data =============================''' + self.back_yolobboxes = np.empty((0, 6), dtype=np.float64) + self.back_yolofeats = np.empty((0, 256), dtype=np.float64) + self.back_trackerboxes = np.empty((0, 9), dtype=np.float64) + self.back_trackerfeats = np.empty((0, 256), dtype=np.float64) + self.back_trackingboxes = np.empty((0, 9), dtype=np.float64) + self.back_trackingfeats = np.empty((0, 256), dtype=np.float64) + + self.front_yolobboxes = np.empty((0, 6), dtype=np.float64) + self.front_yolofeats = np.empty((0, 256), dtype=np.float64) + self.front_trackerboxes = np.empty((0, 9), dtype=np.float64) + self.front_trackerfeats = np.empty((0, 256), dtype=np.float64) + self.front_trackingboxes = np.empty((0, 9), dtype=np.float64) + self.front_trackingfeats = np.empty((0, 256), dtype=np.float64) + + '''=========== 0/1_tracking_output.data ===================''' + self.back_boxes = np.empty((0, 9), dtype=np.float64) + self.front_boxes = np.empty((0, 9), dtype=np.float64) + self.back_feats = np.empty((0, 256), dtype=np.float64) + self.front_feats = np.empty((0, 256), dtype=np.float64) + self.feats_compose = np.empty((0, 256), dtype=np.float64) + self.feats_select = np.empty((0, 256), dtype=np.float64) + + if stype=="data": + self.from_datafile(eventpath) + + if stype=="video": + self.from_video(eventpath) + + if stype=="image": + self.from_image(eventpath) + + def from_datafile(self, eventpath): + evtList = self.evtname.split('_') + if len(evtList)>=2 and len(evtList[-1])>=10 and evtList[-1].isdigit(): + self.barcode = evtList[-1] + if len(evtList)==3 and evtList[-1]== evtList[-2]: + self.evtType = 'input' + else: + self.evtType = 'other' + + '''================ path of image =============''' + frontImgs, frontFid = [], [] + backImgs, backFid = [], [] + for imgname in os.listdir(eventpath): + name, ext = os.path.splitext(imgname) + if ext not in IMG_FORMAT or name.find('frameId') < 0: continue + if len(name.split('_')) != 3 and not name.split('_')[3].isdigit(): continue + + CamerType = name.split('_')[0] + frameId = int(name.split('_')[3]) + imgpath = os.path.join(eventpath, imgname) + if CamerType == '0': + backImgs.append(imgpath) + backFid.append(frameId) + if CamerType == '1': + frontImgs.append(imgpath) + frontFid.append(frameId) + ## 生成依据帧 ID 排序的前后摄图像地址列表 + frontIdx = np.argsort(np.array(frontFid)) + backIdx = np.argsort(np.array(backFid)) + self.front_imgpaths = [frontImgs[i] for i in frontIdx] + self.back_imgpaths = [backImgs[i] for i in backIdx] + + + '''================ path of video =============''' + for vidname in os.listdir(eventpath): + name, ext = os.path.splitext(vidname) + if ext not in VID_FORMAT: continue + vidpath = os.path.join(eventpath, vidname) + + CamerType = name.split('_')[0] + if CamerType == '0': + self.back_videopath = vidpath + if CamerType == '1': + self.front_videopath = vidpath + + '''================ process.data =============''' + procpath = Path(eventpath).joinpath('process.data') + if procpath.is_file(): + SimiDict = read_similar(procpath) + self.one2one = SimiDict['one2one'] + self.one2n = SimiDict['one2n'] + + + '''=========== 0/1_track.data & 0/1_tracking_output.data =======''' + for dataname in os.listdir(eventpath): + datapath = os.path.join(eventpath, dataname) + if not os.path.isfile(datapath): continue + CamerType = dataname.split('_')[0] + + '''========== 0/1_track.data ==========''' + if dataname.find("_track.data")>0: + bboxes, ffeats, trackerboxes, tracker_feat_dict, trackingboxes, tracking_feat_dict = extract_data(datapath) + if CamerType == '0': + self.back_yolobboxes = bboxes + self.back_yolofeats = ffeats + self.back_trackerboxes = trackerboxes + self.back_trackerfeats = tracker_feat_dict + self.back_trackingboxes = trackingboxes + self.back_trackingfeats = tracking_feat_dict + if CamerType == '1': + self.front_yolobboxes = bboxes + self.front_yolofeats = ffeats + self.front_trackerboxes = trackerboxes + self.front_trackerfeats = tracker_feat_dict + self.front_trackingboxes = trackingboxes + self.front_trackingfeats = tracking_feat_dict + + '''========== 0/1_tracking_output.data ==========''' + if dataname.find("_tracking_output.data")>0: + tracking_output_boxes, tracking_output_feats = read_tracking_output(datapath) + if CamerType == '0': + self.back_boxes = tracking_output_boxes + self.back_feats = tracking_output_feats + elif CamerType == '1': + self.front_boxes = tracking_output_boxes + self.front_feats = tracking_output_feats + self.select_feat() + self.compose_feats() + + + def compose_feats(self): + '''事件的特征集成''' + feats_compose = np.empty((0, 256), dtype=np.float64) + if len(self.front_feats): + feats_compose = np.concatenate((feats_compose, self.front_feats), axis=0) + if len(self.back_feats): + feats_compose = np.concatenate((feats_compose, self.back_feats), axis=0) + self.feats_compose = feats_compose + + def select_feats(self): + '''事件的特征选择''' + if len(self.front_feats): + self.feats_select = self.front_feats + else: + self.feats_select = self.back_feats + + + + + + \ No newline at end of file diff --git a/pipeline.py b/pipeline.py index 354563c..4faba3f 100644 --- a/pipeline.py +++ b/pipeline.py @@ -5,11 +5,19 @@ Created on Sun Sep 29 08:59:21 2024 @author: ym """ import os +import sys import cv2 import pickle +import argparse import numpy as np from pathlib import Path -from track_reid import parse_opt, yolo_resnet_tracker +from track_reid import parse_opt +from track_reid import yolo_resnet_tracker +# FILE = Path(__file__).resolve() +# ROOT = FILE.parents[0] # YOLOv5 root directory +# if str(ROOT) not in sys.path: +# sys.path.append(str(ROOT)) # add ROOT to PATH +# ROOT = Path(os.path.relpath(ROOT, Path.cwd())) # relative from tracking.dotrack.dotracks_back import doBackTracks from tracking.dotrack.dotracks_front import doFrontTracks @@ -35,38 +43,27 @@ def get_interbcd_inputenents(): return input_enents -def pipeline(eventpath, stdfeat_path=None, SourceType = "image"): - ''' - inputs: - eventpath: 事件文件夹 - stdfeat_path: 标准特征文件地址 - outputs: - - ''' - # SourceType = "image" # image - # eventpath = r"\\192.168.1.28\share\测试_202406\0918\images1\20240918-110822-1bc3902e-5a8e-4e23-8eca-fb3f02738551_6938314601726" - - savepath = r"D:\contrast\detect" - - opt = parse_opt() - optdict = vars(opt) - optdict["project"] = savepath - - eventname = os.path.basename(eventpath) - # barcode = eventname.split('_')[-1] - - +def pipeline( + eventpath, + savepath, + SourceType = "image", # video + stdfeat_path = None + ): if SourceType == "video": vpaths = get_video_pairs(eventpath) elif SourceType == "image": vpaths = get_image_pairs(eventpath) - - + '''======== 函数 yolo_resnet_tracker() 的参数字典 ========''' + opt = parse_opt() + optdict = vars(opt) + optdict["is_save_img"] = True + optdict["is_save_video"] = True + event_tracks = [] for vpath in vpaths: '''事件结果文件夹''' - save_dir_event = Path(savepath) / Path(eventname) + save_dir_event = Path(savepath) / Path(os.path.basename(eventpath)) if isinstance(vpath, list): save_dir_video = save_dir_event / Path("images") else: @@ -78,8 +75,7 @@ def pipeline(eventpath, stdfeat_path=None, SourceType = "image"): '''Yolo + Resnet + Tracker''' optdict["source"] = vpath optdict["save_dir"] = save_dir_video - optdict["is_save_img"] = True - optdict["is_save_video"] = True + tracksdict = yolo_resnet_tracker(**optdict) @@ -138,6 +134,7 @@ def pipeline(eventpath, stdfeat_path=None, SourceType = "image"): '''前后摄轨迹选择''' + if stdfeat_path is not None: with open(stdfeat_path, 'rb') as f: featDict = pickle.load(f) @@ -171,22 +168,29 @@ def main_loop(): stdfeat_path = os.path.join(bcdpath, f"{bcd}.pickle") input_enents.append((event_path, stdfeat_path)) - + parmDict = {} + parmDict["SourceType"] = "image" + parmDict["savepath"] = r"D:\contrast\detect" for eventpath, stdfeat_path in input_enents: - pipeline(eventpath, stdfeat_path, SourceType) + parmDict["eventpath"] = eventpath + parmDict["stdfeat_path"] = stdfeat_path - -def main(): - eventpath = r"D:\datasets\ym\exhibition\175836" + pipeline(**parmDict) - eventpath = r"\\192.168.1.28\share\测试视频数据以及日志\各模块测试记录\展厅测试\1120_展厅模型v801测试\扫A放A\20241121-144855-dce94b09-1100-43f1-92e8-33a1b538b159_6924743915848_6924743915848" - - SourceType = 'image' - stdfeat_path = None - - pipeline(eventpath, stdfeat_path, SourceType) - +def main(): + ''' + 函数：pipeline()，遍历事件文件夹，选择类型 image 或 video, + ''' + parmDict = {} + parmDict["eventpath"] = r"\\192.168.1.28\share\测试视频数据以及日志\各模块测试记录\展厅测试\1120_展厅模型v801测试\扫A放A\20241121-144855-dce94b09-1100-43f1-92e8-33a1b538b159_6924743915848_6924743915848" + + parmDict["savepath"] = r"D:\contrast\detect" + parmDict["SourceType"] = "image" # video, image + parmDict["stdfeat_path"] = None + + pipeline(**parmDict) + if __name__ == "__main__": main() diff --git a/time_devide.py b/time_devide.py index c035bd6..1836112 100644 --- a/time_devide.py +++ b/time_devide.py @@ -204,11 +204,6 @@ def state_measure(periods, weights, hands, spath=None): state1 = frstate_1[:,2][:, None] state11 = frstate_1[:,3][:, None] - - - - - @@ -428,8 +423,8 @@ def run_tracking(trackboxes, MotionSlice): def show_seri(): - datapath = r"\\192.168.1.28\share\realtime\eventdata\1731316835560" - savedir = r"D:\DetectTracking\realtime" + datapath = r"\\192.168.1.28\share\个人文件\wqg\realtime\eventdata\1731316835560" + savedir = r"D:\DetectTracking\realtime\1" imgdir = datapath.split('\\')[-2] + "_" + datapath.split('\\')[-1] @@ -475,7 +470,7 @@ def show_seri(): def main(): - runyolo() + # runyolo() show_seri() diff --git a/track_reid.py b/track_reid.py index 9ccf665..76ddb95 100644 --- a/track_reid.py +++ b/track_reid.py @@ -64,6 +64,9 @@ from contrast.feat_extract.config import config as conf from contrast.feat_extract.inference import FeatsInterface ReIDEncoder = FeatsInterface(conf) +IMG_FORMATS = '.bmp', '.dng', '.jpeg', '.jpg', '.mpo', '.png', '.tif', '.tiff', '.webp', '.pfm' # include image suffixes +VID_FORMATS = '.asf', '.avi', '.gif', '.m4v', '.mkv', '.mov', '.mp4', '.mpeg', '.mpg', '.ts', '.wmv' # include video suffixes + # from tracking.trackers.reid.reid_interface import ReIDInterface # from tracking.trackers.reid.config import config as ReIDConfig # ReIDEncoder = ReIDInterface(ReIDConfig) @@ -141,6 +144,9 @@ def yolo_resnet_tracker( name='exp', # save results to project/name save_dir = '', + is_save_img = False, + is_save_video = True, + tracker_yaml = "./tracking/trackers/cfg/botsort.yaml", imgsz=(640, 640), # inference size (height, width) conf_thres=0.25, # confidence threshold @@ -153,18 +159,15 @@ def yolo_resnet_tracker( save_csv=False, # save results in CSV format save_conf=False, # save confidences in --save-txt labels save_crop=False, # save cropped prediction boxes - nosave=False, # do not save images/videos - is_save_img = False, - is_save_video = True, - - + update=False, # update all models + exist_ok=False, # existing project/name ok, do not increment + classes=None, # filter by class: --class 0, or --class 0 2 3 agnostic_nms=False, # class-agnostic NMS augment=False, # augmented inference visualize=False, # visualize features - update=False, # update all models - exist_ok=False, # existing project/name ok, do not increment + line_thickness=3, # bounding box thickness (pixels) hide_labels=False, # hide labels hide_conf=False, # hide confidencesL @@ -179,6 +182,8 @@ def yolo_resnet_tracker( model = DetectMultiBackend(weights, device=device, dnn=dnn, data=data, fp16=half) stride, names, pt = model.stride, model.names, model.pt imgsz = check_img_size(imgsz, s=stride) # check image size + + # Dataloader bs = 1 # batch_size @@ -203,8 +208,8 @@ def yolo_resnet_tracker( # Inference with dt[1]: - visualize = increment_path(save_dir / Path(path).stem, mkdir=True) if visualize else False - pred = model(im, augment=augment, visualize=visualize) + visualize = increment_path(project / Path(path).stem, mkdir=True) if visualize else False + pred = model(im, augment=augment, visualize=False) # NMS with dt[2]: @@ -273,19 +278,27 @@ def yolo_resnet_tracker( annotator.box_label(xyxy, label, color=color) '''====== Save results (image and video) ======''' - save_path = str(save_dir / Path(path).name) # 带有后缀名 + # save_path = str(save_dir / Path(path).name) # 带有后缀名 im0 = annotator.result() if is_save_img: - save_path_img, ext = os.path.splitext(save_path) + save_path_img = str(save_dir / Path(path).stem) if dataset.mode == 'image': imgpath = save_path_img + ".png" else: imgpath = save_path_img + f"_{frameId}.png" cv2.imwrite(Path(imgpath), im0) - if dataset.mode == 'video' and is_save_video: - if vid_path[i] != save_path: # new video - vid_path[i] = save_path + # if dataset.mode == 'video' and is_save_video: + + if is_save_video: + if dataset.mode == 'video': + vdieo_path = str(save_dir / Path(path).stem) + '.mp4' # 带有后缀名 + else: + videoname = str(Path(path).stem).split('_')[0] + '.mp4' + vdieo_path = str(save_dir / videoname) + + if vid_path[i] != vdieo_path: # new video + vid_path[i] = vdieo_path if isinstance(vid_writer[i], cv2.VideoWriter): vid_writer[i].release() # release previous video writer if vid_cap: # video @@ -293,9 +306,9 @@ def yolo_resnet_tracker( w = int(vid_cap.get(cv2.CAP_PROP_FRAME_WIDTH)) h = int(vid_cap.get(cv2.CAP_PROP_FRAME_HEIGHT)) else: # stream - fps, w, h = 30, im0.shape[1], im0.shape[0] - save_path = str(Path(save_path).with_suffix('.mp4')) # force *.mp4 suffix on results videos - vid_writer[i] = cv2.VideoWriter(save_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h)) + fps, w, h = 25, im0.shape[1], im0.shape[0] + vdieo_path = str(Path(vdieo_path).with_suffix('.mp4')) # force *.mp4 suffix on results videos + vid_writer[i] = cv2.VideoWriter(vdieo_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h)) vid_writer[i].write(im0) # Print time (inference-only) @@ -344,6 +357,9 @@ def run( vid_stride=1, # video frame-rate stride data=ROOT / 'data/coco128.yaml', # dataset.yaml path ): + ''' + source: 视频文件或图像列表 + ''' source = str(source) # filename = os.path.split(source)[-1] @@ -355,6 +371,9 @@ def run( if is_url and is_file: source = check_file(source) # download + + + # spth = source.split('\\')[-2] + "_" + Path(source).stem save_dir = Path(project) / Path(source.split('\\')[-2] + "_" + str(Path(source).stem)) @@ -440,8 +459,7 @@ def run( - p = Path(p) # to Path - save_path = str(save_dir / p.name) # im.jpg + s += '%gx%g ' % im.shape[2:] # print string # im0_ant = im0.copy() @@ -552,28 +570,33 @@ def run( # Save results (image and video with tracking) im0 = annotator.result() - save_path_img, ext = os.path.splitext(save_path) + + p = Path(p) # to Path + save_path = str(save_dir / p.name) # im.jpg if save_img: + save_path_img, ext = os.path.splitext(save_path) if dataset.mode == 'image': - imgpath = save_path_img + f"_{dataset}.png" + imgpath = save_path_img + ".png" else: imgpath = save_path_img + f"_{frameId}.png" - cv2.imwrite(Path(imgpath), im0) - if vid_path[i] != save_path: # new video - vid_path[i] = save_path - if isinstance(vid_writer[i], cv2.VideoWriter): - vid_writer[i].release() # release previous video writer - if vid_cap: # video - fps = vid_cap.get(cv2.CAP_PROP_FPS) - w = int(vid_cap.get(cv2.CAP_PROP_FRAME_WIDTH)) - h = int(vid_cap.get(cv2.CAP_PROP_FRAME_HEIGHT)) - else: # stream - fps, w, h = 30, im0.shape[1], im0.shape[0] - save_path = str(Path(save_path).with_suffix('.mp4')) # force *.mp4 suffix on results videos - vid_writer[i] = cv2.VideoWriter(save_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h)) - vid_writer[i].write(im0) + if dataset.mode == 'video': + + if vid_path[i] != save_path: # new video + vid_path[i] = save_path + if isinstance(vid_writer[i], cv2.VideoWriter): + vid_writer[i].release() # release previous video writer + if vid_cap: # video + fps = vid_cap.get(cv2.CAP_PROP_FPS) + w = int(vid_cap.get(cv2.CAP_PROP_FRAME_WIDTH)) + h = int(vid_cap.get(cv2.CAP_PROP_FRAME_HEIGHT)) + else: # stream + fps, w, h = 30, im0.shape[1], im0.shape[0] + save_path = str(Path(save_path).with_suffix('.mp4')) # force *.mp4 suffix on results videos + vid_writer[i] = cv2.VideoWriter(save_path, cv2.VideoWriter_fourcc(*'mp4v'), fps, (w, h)) + vid_writer[i].write(im0) + # Print time (inference-only) LOGGER.info(f"{s}{'' if len(det) else '(no detections), '}{dt[1].dt * 1E3:.1f}ms") @@ -672,40 +695,25 @@ def parse_opt(): print_args(vars(opt)) return opt -def find_files_in_nested_dirs(root_dir): +def find_video_imgs(root_dir): all_files = [] extensions = ['.mp4'] for dirpath, dirnames, filenames in os.walk(root_dir): for filename in filenames: file, ext = os.path.splitext(filename) - if ext in extensions: + if ext in IMG_FORMATS + VID_FORMATS: all_files.append(os.path.join(dirpath, filename)) return all_files -print('=======') - -def main(opt): - check_requirements(ROOT / 'requirements.txt', exclude=('tensorboard', 'thop')) - optdict = vars(opt) - - p = r"D:\datasets\ym" - p = r"D:\exhibition\images\153112511_0_seek_105.mp4" - - optdict["project"] = r"D:\exhibition\result" - - files = [] - if os.path.isdir(p): - files.extend(sorted(glob.glob(os.path.join(p, '*.*')))) - optdict["source"] = files - elif os.path.isfile(p): - optdict["source"] = p - - run(**optdict) -def main_loop(opt): +def main(): + ''' + run(): 单张图像或单个视频文件的推理，不支持图像序列， + ''' + check_requirements(ROOT / 'requirements.txt', exclude=('tensorboard', 'thop')) - + opt = parse_opt() optdict = vars(opt) # p = r"D:\datasets\ym\永辉测试数据_比对" @@ -714,28 +722,22 @@ def main_loop(opt): # p = r"D:\datasets\ym\实验室测试" # p = r"D:\datasets\ym\永辉双摄视频\新建文件夹" # p = r"\\192.168.1.28\share\测试_202406\0723\0723_2\20240723-112522_" - p = r"D:\datasets\ym\联华中环" + # p = r"D:\datasets\ym\联华中环" + p = r"D:\exhibition\images\153112511_0_seek_105.mp4" + # p = r"D:\exhibition\images\image" - k = 0 + + optdict["project"] = r"D:\exhibition\result" if os.path.isdir(p): - files = find_files_in_nested_dirs(p) - - # files = [r"D:\datasets\ym\广告板遮挡测试\8\6926636301004_20240508-175300_back_addGood_70f754088050_215_17327712807.mp4", - # r"D:\datasets\ym\videos\标记视频\test_20240402-173935_6920152400975_back_174037372.mp4", - # r"D:\datasets\ym\videos\标记视频\test_20240402-173935_6920152400975_front_174037379.mp4", - # r"D:\datasets\ym\广告板遮挡测试\8\2500441577966_20240508-175946_front_addGood_70f75407b7ae_155_17788571404.mp4" - # ] - - # files = [r"\\192.168.1.28\share\测试_202406\0723\0723_2\20240723-095838_\1_seek_193.mp4"] - - + files = find_video_imgs(p) + k = 0 for file in files: optdict["source"] = file run(**optdict) - # k += 1 - # if k == 10: - # break + k += 1 + if k == 1: + break elif os.path.isfile(p): optdict["source"] = p run(**optdict) @@ -744,10 +746,8 @@ def main_loop(opt): if __name__ == '__main__': - opt = parse_opt() - - main(opt) - # main_loop(opt) + main() + diff --git a/tracking/utils/__pycache__/read_data.cpython-39.pyc b/tracking/utils/__pycache__/read_data.cpython-39.pyc index 5f04eaa..f761704 100644 Binary files a/tracking/utils/__pycache__/read_data.cpython-39.pyc and b/tracking/utils/__pycache__/read_data.cpython-39.pyc differ diff --git a/tracking/utils/read_data.py b/tracking/utils/read_data.py index 1a12b73..32032d2 100644 --- a/tracking/utils/read_data.py +++ b/tracking/utils/read_data.py @@ -35,9 +35,6 @@ def find_samebox_in_array(arr, target): - - - def extract_data(datapath): ''' 0/1_track.data 数据读取 @@ -71,7 +68,7 @@ def extract_data(datapath): boxes, feats, tboxes, tfeats = [], [], [], [] - if line.find("box:") >= 0 and line.find("output_box:") < 0: + if line.find("box:") >= 0 and line.find("output_box:")<0 and line.find("out_boxes")<0: box = line[line.find("box:") + 4:].strip() # if len(box)==6: boxes.append(str_to_float_arr(box)) @@ -122,7 +119,9 @@ def extract_data(datapath): for line in lines: line = line.strip() # 去除行尾的换行符和可能的空白字符 if not line: # 跳过空行 - continue + tracking_flag = False + continue + if tracking_flag: if line.find("tracking_") >= 0: tracking_flag = False @@ -176,8 +175,10 @@ def read_tracking_output(filepath): boxes.append(data) if data.size == 256: feats.append(data) - - assert(len(feats)==len(boxes)), f"{filepath}, len(feats)!=len(boxes)" + + + if len(feats) != len(boxes): + return np.array([]), np.array([]) return np.array(boxes), np.array(feats) @@ -331,7 +332,6 @@ def read_similar(filePath): line = line[:-1] Dict = {} - if not line: if len(one2one_list): SimiDict['one2one'] = one2one_list if len(one2n_list): SimiDict['one2n'] = one2n_list diff --git a/说明文档.txt b/说明文档.txt index 00b3ee9..37d2026 100644 --- a/说明文档.txt +++ b/说明文档.txt @@ -38,7 +38,7 @@ 需分 2 步运行模块: (1) runyolo() - 该模块调用 imgs_inference.py 中模块 run_yolo + 该模块调用 imgs_inference.py 中模块 run_yolo, 该模块重新定义了类 LoadImages, 对图像进行了旋转。 后续工作： 1). 将run_yolo模块与track_redi.yolo_resnet_tracker模块合并 2). 图像文件名标准化 @@ -126,14 +126,14 @@ ./contrast - feat_similar.py + seqfeat_compare.py similarity_compare_sequence(root_dir) inputs: root_dir：文件夹，包含"subimgs"字段，对该文件夹中的相邻图像进行相似度比较 silimarity_compare() 功能：对imgpaths文件夹中的图像进行相似度比较 - feat_select.py + input_getout_compare.py creatd_deletedBarcode_front(filepath) (1) 基于 deletedBarcode.txt, 构造取出事件和相应的放入事件，构成列表并更新这些列表。 MatchList = [(getout_event, InputList), ...] @@ -145,6 +145,9 @@ precision_compare(filepath, savepath) 读取 deletedBarcode.txt 和 deletedBarcodeTest.txt 中的数据，进行相似度比较 + stdfeat_analys() + + genfeats.py get_std_barcodeDict(bcdpath, savepath) 功能: 生成并保存只有一个key值的字典 {barcode: [imgpath1, imgpath1, ...]} @@ -207,6 +210,7 @@ one2one_contrast.py + 已修改，未更新。 共6个地址： (1) stdSamplePath: 用于生成比对标准特征集的原始图像地址 (2) stdBarcodePath: 比对标准特征集原始图像地址的pickle文件存储，{barcode: [imgpath1, imgpath1, ...]} @@ -283,6 +287,11 @@ (3) featpath：调用 inference_image(), 对每一个barcode，生成字典并进行存储 - +time_devide.py + runyolo() + 执行 imgs_inference.py 中的 run_yolo()模块，该模块重新定义了类 LoadImages, 对图像进行了旋转。 + + show_seri() + \ No newline at end of file