pipeline.py 等更新

contrast/event_test.py

@@ -169,9 +169,6 @@ def calc_simil(event, stdfeat):
         Similar = None   # 在event.front_feats和event.back_feats同时为空时
 
     return Similar
-    
-    
-
 
 
 def simi_matrix():

contrast/feat_extract/config.py

@@ -61,8 +61,9 @@ class Config:
     test_val = "D:/比对/cl"
     # test_val = "./data/test_data_100"
     
-    test_model = "checkpoints/best_resnet18_v12.pth"   
+    # test_model = "checkpoints/best_resnet18_v12.pth"   
     # test_model = "checkpoints/zhanting_res_801.pth"
+    test_model = "checkpoints/zhanting_res_abroad_8021.pth"
     
     
 

contrast/feat_extract/inference.py

@@ -61,8 +61,17 @@ class FeatsInterface:
         batch_patches = []
         patches = []
         for i, img in enumerate(images):
-            img = img.copy()   
+            img = img.copy()
-            patch = self.transform(img)            
+            
+            ## 对 img 进行补黑边，生成新的图像new_img
+            width, height = img.size
+            new_size = max(width, height)
+            new_img = Image.new("RGB", (new_size, new_size), (0, 0, 0))
+            paste_x = (new_size - width) // 2
+            paste_y = (new_size - height) // 2
+            new_img.paste(img, (paste_x, paste_y))
+            
+            patch = self.transform(new_img)            
             if str(self.device) != "cpu":
                 patch = patch.to(device=self.device).half()
             else:
@@ -107,10 +116,12 @@ class FeatsInterface:
             patch = self.transform(img1)
 
             # patch = patch.to(device=self.device).half()
-            if str(self.device) != "cpu":
+            # if str(self.device) != "cpu":
-                patch = patch.to(device=self.device).half()
+            #     patch = patch.to(device=self.device).half()
-            else:
+            #     patch = patch.to(device=self.device)
-                patch = patch.to(device=self.device)
+            # else:
+                # patch = patch.to(device=self.device)
+            patch = patch.to(device=self.device)
             
             patches.append(patch)
             if (d + 1) % self.batch_size == 0:

contrast/genfeats.py

@@ -158,6 +158,8 @@ def stdfeat_infer(imgPath, featPath, bcdSet=None):
                 
                 feature /= np.linalg.norm(feature, axis=1)[:, None]
                 
+                feature_ft32 = feature.astype(np.float32)
+                
                 # float16
                 feature_ft16 = feature.astype(np.float16)
                 feature_ft16 /= np.linalg.norm(feature_ft16, axis=1)[:, None]
@@ -173,7 +175,7 @@ def stdfeat_infer(imgPath, featPath, bcdSet=None):
         ##================== float32
         stdbDict["barcode"] = barcode
         stdbDict["imgpaths"] = imgpaths
-        stdbDict["feats_ft32"] = feature
+        stdbDict["feats_ft32"] = feature_ft32
         stdbDict["feats_ft16"] = feature_ft16
         stdbDict["feats_uint8"] = feature_uint8
         
@@ -202,9 +204,13 @@ def gen_bcd_features(imgpath, bcdpath, featpath, bcdSet=None):
     stdfeat_infer(bcdpath, featpath, bcdSet)
     
 def main():
-    imgpath = 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"
+    imgpath = r"\\192.168.1.28\share\数据\已完成数据\展厅数据\v2.0_abroad\比对数据\all_base_二筛"
-    bcdpath = r"D:\exhibition\dataset\bcdpath"  
+    bcdpath = r"D:\exhibition\dataset\bcdpath_abroad"  
-    featpath = r"D:\exhibition\dataset\feats"
+    featpath = r"D:\exhibition\dataset\feats_abroad"
+    if not os.path.exists(bcdpath): 
+        os.makedirs(bcdpath)
+    if not os.path.exists(featpath): 
+        os.makedirs(featpath) 
 
 
     gen_bcd_features(imgpath, bcdpath, featpath)

contrast/one2n_contrast.py

@@ -92,9 +92,9 @@ def simi_calc(event, o2nevt, typee=None):
         feat2 = o2nevt.front_feats
     
     '''自定义事件特征选择'''    
-    if typee==3:
+    if typee==3 and len(event.feats_compose) and len(o2nevt.feats_compose):
-        feat1 = event.feats_compose
+        feat1 = [event.feats_compose]
-        feat2 = o2nevt.feats_compose
+        feat2 = [o2nevt.feats_compose]
         
     
     if len(feat1) and len(feat2):
@@ -109,15 +109,18 @@ def one2n_pr(evtDicts, pattern=1):
     '''
     pattern:
         1: process.data 中记录的相似度
-        2: 根据 process.data 中标记的 type 选择特征计算
+        2: 根据 process.data 中标记的 type 选择特征计算相似度
-        3: 以其它方式选择特征计算   
+        3: 以其它方式选择特征计算相似度
     '''
     
     tpevents, fnevents, fpevents, tnevents = [], [], [], []
     tpsimi, fnsimi, tnsimi, fpsimi = [], [], [], []
-    errorFile_one2n = []
+    one2nFile, errorFile_one2n = [], []
     for evtname, event in evtDicts.items():
         evt_names, evt_barcodes, evt_similars, evt_types = [], [], [], [] 
+                
+        if len(event.barcode)==0:
+            continue
         
         for ndict in event.one2n:
             nname = ndict["event"]
@@ -149,6 +152,13 @@ def one2n_pr(evtDicts, pattern=1):
                     continue
                 evt_similars.append(simival)
         
+        ## process.data的oneTon的各项中，均不包括当前事件的barcode
+        if event.barcode not in evt_barcodes:
+            errorFile_one2n.append(evtname)
+            continue
+        else:
+            one2nFile.append(evtname)
+
         if len(evt_names)==len(evt_barcodes) and len(evt_barcodes)==len(evt_similars) \
             and len(evt_similars)==len(evt_types) and len(evt_names)>0:  
             
@@ -166,7 +176,7 @@ def one2n_pr(evtDicts, pattern=1):
                 elif bcd!=event.barcode and simi!=maxsim: 
                     tnsimi.append(simi)
                     tnevents.append(evtname)
-                elif bcd!=event.barcode and simi==maxsim and event.barcode in evt_barcodes:
+                elif bcd!=event.barcode and simi==maxsim:
                     fpsimi.append(simi)
                     fpevents.append(evtname)
                 else:
@@ -187,7 +197,7 @@ def one2n_pr(evtDicts, pattern=1):
         TN = sum(np.array(tnsimi) < th)
         
         PPrecise.append(TP/(TP+FP+1e-6))
-        PRecall.append(TP/(len(tpsimi)+len(fnsimi)+1e-6)) 
+        PRecall.append(TP/(len(one2nFile)+1e-6)) 
         NPrecise.append(TN/(TN+FN+1e-6))
         NRecall.append(TN/(len(tnsimi)+len(fpsimi)+1e-6))
         
@@ -202,23 +212,23 @@ def one2n_pr(evtDicts, pattern=1):
     ax.set_ylim([0, 1])
     ax.grid(True)
     ax.set_title('1:n Precise & Recall')
-    ax.set_xlabel(f"Event Num: {len(tpsimi)+len(fnsimi)}")     
+    ax.set_xlabel(f"Event Num: {len(one2nFile)}")     
     ax.legend()
     plt.show()
     ## ============================= 1:n 直方图'''
     fig, axes = plt.subplots(2, 2)
     axes[0, 0].hist(tpsimi, bins=60, range=(-0.2, 1), edgecolor='black')
     axes[0, 0].set_xlim([-0.2, 1])
-    axes[0, 0].set_title('TP')
+    axes[0, 0].set_title(f'TP: {len(tpsimi)}')
     axes[0, 1].hist(fpsimi, bins=60, range=(-0.2, 1), edgecolor='black')
     axes[0, 1].set_xlim([-0.2, 1])
-    axes[0, 1].set_title('FP')
+    axes[0, 1].set_title(f'FP: {len(fpsimi)}')
     axes[1, 0].hist(tnsimi, bins=60, range=(-0.2, 1), edgecolor='black')
     axes[1, 0].set_xlim([-0.2, 1])
-    axes[1, 0].set_title('TN')
+    axes[1, 0].set_title(f'TN: {len(tnsimi)}')
     axes[1, 1].hist(fnsimi, bins=60, range=(-0.2, 1), edgecolor='black')
     axes[1, 1].set_xlim([-0.2, 1])
-    axes[1, 1].set_title('FN')
+    axes[1, 1].set_title(f'FN: {len(fnsimi)}')
     plt.show()
       
     return fpevents
@@ -226,14 +236,14 @@ def one2n_pr(evtDicts, pattern=1):
 def main():
     
     '''1. 生成事件字典并保存至 eventDataPath, 只需运行一次 '''
-    init_eventdict(eventSourcePath, stype="source")
+    init_eventdict(eventSourcePath, stype="data")
  
     '''2. 读取事件字典 '''
     evtDicts = read_eventdict(eventDataPath)
     
     
     '''3. 1:n 比对事件评估 '''
-    fpevents = one2n_pr(evtDicts, pattern=3)
+    fpevents = one2n_pr(evtDicts, pattern=1)
 
     fpErrFile = str(Path(resultPath).joinpath("one2n_fp_Error.txt"))
     with open(fpErrFile, "w") as file:
@@ -243,15 +253,16 @@ def main():
             
             
 if __name__ == '__main__':
-    eventSourcePath  = r"\\192.168.1.28\share\测试视频数据以及日志\算法全流程测试\202412\result\ShoppingDict_pkfile"
+    eventSourcePath  = r"\\192.168.1.28\share\测试视频数据以及日志\海外展厅测试数据\比对数据"
-    resultPath = r"\\192.168.1.28\share\测试视频数据以及日志\算法全流程测试\202412\result\contrast"
+    resultPath = r"\\192.168.1.28\share\测试视频数据以及日志\海外展厅测试数据\testing"
     
     eventDataPath = os.path.join(resultPath, "evtobjs")
-    similPath = os.path.join(resultPath, "simidata")
     if not os.path.exists(eventDataPath): 
         os.makedirs(eventDataPath) 
-    if not os.path.exists(similPath): 
+        
-        os.makedirs(similPath)
+    # similPath = os.path.join(resultPath, "simidata")
+    # if not os.path.exists(similPath): 
+    #     os.makedirs(similPath)
     
     main()
     

contrast/one2one_contrast.py

@@ -626,12 +626,13 @@ if __name__ == '__main__':
     # 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"]
     
     
-    stdSamplePath = r"\\192.168.1.28\share\数据\已完成数据\比对数据\barcode\all_totalBarocde\totalBarcode"
+    stdSamplePath = r"\\192.168.1.28\share\数据\已完成数据\展厅数据\v2.0_abroad\比对数据\all_base_二筛"
-    stdBarcodePath = r"D:\全实时\source_data\bcdpath"
+    stdBarcodePath = r"\\192.168.1.28\share\测试视频数据以及日志\海外展厅测试数据\testing\bcdpath"
-    stdFeaturePath = r"D:\全实时\source_data\stdfeats"
+    stdFeaturePath = r"\\192.168.1.28\share\测试视频数据以及日志\海外展厅测试数据\testing\stdfeats"
     
-    eventSourcePath  = [r"\\192.168.1.28\share\测试视频数据以及日志\算法全流程测试\202412\result\ShoppingDict_pkfile"]
+    eventSourcePath  = [r"\\192.168.1.28\share\测试视频数据以及日志\海外展厅测试数据\比对数据"]
-    resultPath = r"\\192.168.1.28\share\测试视频数据以及日志\算法全流程测试\202412\result\contrast"
+    
+    resultPath = r"\\192.168.1.28\share\测试视频数据以及日志\海外展厅测试数据\testing\evtobjs"
     eventDataPath = os.path.join(resultPath, "evtobjs")
     similPath = os.path.join(resultPath, "simidata")
     if not os.path.exists(eventDataPath): 
@@ -639,7 +640,7 @@ if __name__ == '__main__':
     if not os.path.exists(similPath): 
         os.makedirs(similPath)
 
-    # test_one2one()
+    test_one2one()
     
     test_one2SN()
 

contrast/onsite_contrast_pr.py

@@ -495,7 +495,7 @@ def contrast_pr(paths):
 
 
 if __name__ == "__main__":
-    evtpaths = r"D:\全实时\source_data\2024122416"
+    evtpaths = r"\\192.168.1.28\share\测试视频数据以及日志\算法全流程测试\202412\images"
     contrast_pr(evtpaths)
     
     

contrast/utils/event.py

@@ -202,10 +202,10 @@ class ShoppingEvent:
         self.front_trackingfeats = frontdata[5]
 
         '''===========对应于 0/1_tracking_output.data ============================='''
-        self.back_boxes = back_outdata[0]
+        self.back_boxes = back_outdata
-        self.back_feats = back_outdata[1]
+        self.back_feats = back_outdata
-        self.front_boxes = front_outdata[0]
+        self.front_boxes = front_outdata
-        self.front_feats = front_outdata[1]
+        self.front_feats = front_outdata
 
 
     def from_datafile(self, eventpath):

pipeline.py

@@ -103,14 +103,16 @@ def pipeline(
 
     # if pf_path.exists():
     #     return
-
+    
+    
+    '''====================== 构造 ShoppingDict 模块 ======================='''
     ShoppingDict = {"eventPath": eventpath, 
                     "eventName": evtname,
                     "barcode": barcode,
                     "eventType": '',   # "input", "output", "other"
                     "frontCamera": {}, 
                     "backCamera": {},
-                    "one2n": []
+                    "one2n": []        # 
                     }
     
     
@@ -121,7 +123,7 @@ def pipeline(
 
 
     for vpath in vpaths:
-        '''相机事件字典构造'''
+        '''================= 1. 构造相机事件字典 ================='''
         CameraEvent = {"cameraType": '',   # "front", "back"
                        "videoPath": '',
                        "imagePaths": [],
@@ -140,21 +142,21 @@ def pipeline(
         if  bname.split('_')[0] == "1" or bname.find('front')>=0:
             CameraEvent["cameraType"] = "front"
             
-        '''事件结果存储文件夹'''
+        '''================= 2. 事件结果存储文件夹 ================='''
-        
         if isinstance(vpath, list):
             savepath_pipeline_imgs = savepath_pipeline / Path("images") 
         else:
             savepath_pipeline_imgs = savepath_pipeline / Path(str(Path(vpath).stem))
+        
         if not savepath_pipeline_imgs.exists():
-            savepath_pipeline_imgs.mkdir(parents=True, exist_ok=True)
+            savepath_pipeline_imgs.mkdir(parents=True, exist_ok=True) 
-            
+        
         savepath_pipeline_subimgs = savepath_pipeline / Path("subimgs") 
         if not savepath_pipeline_subimgs.exists():
             savepath_pipeline_subimgs.mkdir(parents=True, exist_ok=True)
 
         
-        '''Yolo + Resnet + Tracker'''
+        '''================= 3. Yolo + Resnet + Tracker ================='''
         optdict["source"] = vpath
         optdict["save_dir"] = savepath_pipeline_imgs        
         
@@ -162,11 +164,8 @@ def pipeline(
         
         CameraEvent["yoloResnetTracker"] = yrtOut
         
-        
+        '''================= 4. tracking ================='''
-        # bboxes = np.empty((0, 9), dtype = np.float32)
+        '''(1) 生成用于 tracking 模块的 boxes、feats'''  
-        # for frameDict in yrtOut:            
-        #     bboxes = np.concatenate([bboxes, frameDict["tboxes"]], axis=0)
-            
         trackerboxes = np.empty((0, 9), dtype=np.float64)
         trackefeats = {}
         for frameDict in yrtOut:            
@@ -179,7 +178,7 @@ def pipeline(
                 trackefeats.update({f"{fid}_{bid}": ffeats[f"{fid}_{bid}"]})
             
 
-        '''tracking'''
+        '''(2) tracking, 后摄'''
         if CameraEvent["cameraType"] == "back":
             vts = doBackTracks(trackerboxes, trackefeats)
             vts.classify()
@@ -187,7 +186,8 @@ def pipeline(
             
             CameraEvent["tracking"] = vts
             ShoppingDict["backCamera"] = CameraEvent
-            
+        
+        '''(2) tracking, 前摄'''    
         if  CameraEvent["cameraType"] == "front":
             vts = doFrontTracks(trackerboxes, trackefeats)
             vts.classify()
@@ -196,11 +196,12 @@ def pipeline(
             CameraEvent["tracking"] = vts
             ShoppingDict["frontCamera"] = CameraEvent
     
-   
+    '''========================== 保存模块 ================================='''
+    '''(1) 保存 ShoppingDict 事件'''
     with open(str(pf_path), 'wb') as f:
         pickle.dump(ShoppingDict, f)
-        
+ 
-        
+    '''(2) 保存 Tracking 输出的运动轨迹子图，并记录相似度'''
     for CamerType, vts in event_tracks:
         if len(vts.tracks)==0: continue
         if CamerType == 'front':
@@ -215,16 +216,14 @@ def pipeline(
             imgdict.update(y["imgs"])
             featdict.update(y["feats"])
             simidict.update(y["featsimi"])
-            
+
- 
         for track in vts.Residual:
             if isinstance(track, np.ndarray):
                 save_subimgs(imgdict, track, savepath_pipeline_subimgs, ctype, featdict)
             else:
                 save_subimgs(imgdict, track.slt_boxes, savepath_pipeline_subimgs,  ctype, featdict)
-  
 
-    '''轨迹显示模块'''
+    '''(3) 轨迹显示与保存'''
     illus = [None, None]
     for CamerType, vts in event_tracks:
         if len(vts.tracks)==0: continue
@@ -279,15 +278,15 @@ def main():
         if item.is_dir():            
             # item = evtdir/Path("20241209-160201-b97f7a0e-7322-4375-9f17-c475500097e9_6926265317292")
             parmDict["eventpath"] = item
-            pipeline(**parmDict)
+            # pipeline(**parmDict)
             
-            # try:
+            try:
-            #     pipeline(**parmDict)
+                pipeline(**parmDict)
-            # except Exception as e:
+            except Exception as e:
-            #     errEvents.append(str(item))
+                errEvents.append(str(item))
-        k+=1
+        # k+=1
-        if k==2:
+        # if k==100:
-            break
+        #     break
             
     errfile = os.path.join(parmDict["savepath"], f'error_events.txt')
     with open(errfile, 'w', encoding='utf-8') as f:

track_reid.py

@@ -216,7 +216,7 @@ def yolo_resnet_tracker(
                 # Rescale boxes from img_size to im0 size
                 det[:, :4] = scale_boxes(im.shape[2:], det[:, :4], im0.shape).round()
                 
-                # det = det.cpu().numpy()                
+                det = det.cpu().numpy()                
                 ## ================================================================ writed by WQG            
                 '''tracks: [x1, y1, x2, y2, track_id, score, cls, frame_index, box_index]
                             0   1   2   3      4        5     6       7           8

tracking/module_analysis.py

@@ -2,7 +2,10 @@
 """
 Created on Thu May 30 14:03:03 2024
 
-现场测试性能分析
+轨迹分析现场测试性能分析:
+(1) 读取 data 文件中的轨迹数据，绘制轨迹图
+(2) 读取本地运行 Yolo+Rsenet+Tracker+Tracking 的数据，绘制轨迹图
+
 
 @author: ym
 """