3588适配

.gitignore

@@ -90,6 +90,7 @@ parts/
 sdist/
 var/
 wheels/
+paper_data/
 <<<<<<< HEAD
 *.egg-info/
 wandb/

ieemoo-ai-zhanting.py

@@ -1,5 +1,6 @@
 # -*- coding: utf-8 -*-
 from flask import request, Flask
+import pdb
 import numpy as np
 import json
 import time
@@ -10,6 +11,8 @@ import torch
 from PIL import Image
 import datetime
 from torchvision import transforms
+from models.experimental import attempt_load
+from utils.torch_utils import select_device
 # from models.modeling import VisionTransformer, CONFIGS
 from gevent.pywsgi import WSGIServer
 sys.path.insert(0, ".")
@@ -42,7 +45,7 @@ parser.add_argument('--source', type=str, default='../module/ieemoo-ai-zhanting/
 parser.add_argument('--img-size', type=int, default=640, help='inference size (pixels)')
 parser.add_argument('--conf-thres', type=float, default=0.60, help='object confidence threshold')
 parser.add_argument('--iou-thres', type=float, default=0.45, help='IOU threshold for NMS')
-parser.add_argument('--device', default='', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
+parser.add_argument('--device', default='0', help='cuda device, i.e. 0 or 0,1,2,3 or cpu')
 parser.add_argument('--view-img', type=bool, default=True, help='display results')
 parser.add_argument('--save-txt', type=bool, default=True, help='save results to *.txt')
 parser.add_argument('--save-conf', type=bool, default=True, help='save confidences in --save-txt labels')
@@ -57,6 +60,17 @@ parser.add_argument('--exist-ok', type=bool, default=True, help='existing projec
 
 #opt = parser.parse_args()
 opt, unknown = parser.parse_known_args()
+def init_model(opt):
+     device = select_device(opt.device)
+     #half = device.type != 'cpu'
+     model = attempt_load(opt.weights, map_location=device)
+     stride = int(model.stride.max())
+     #if half:
+     model.half()
+     model.eval()
+     return model, stride
+model,stride = init_model(opt)
+
 @app.route("/zhanting", methods=['POST'])
 def get_isempty():
     data = request.get_data()
@@ -73,7 +87,11 @@ def get_isempty():
         file = open(image_path, 'wb')
         file.write(imgdata)
         img = cv2.imread(image_path)
-        site = np.array([[[0, 1024], [0, 571], [313, 365], [949, 367], [1277, 596], [1280, 1024]]], dtype=np.int32)
+        #pdb.set_trace()
+        if img.shape[0] == 1280:
+            site = np.array([[[0, 1280],[0, 671], [300, 390], [740, 390], [1024, 635], [1024,1280]]], dtype=np.int32)
+        else:
+            site = np.array([[[0, 1024],[0, 571], [313, 365], [949, 367], [1277, 596], [1280, 1024]]], dtype=np.int32)
         im = np.zeros(img.shape[:2], dtype="uint8")
         cv2.polylines(im, site, 1, 255)
         cv2.fillPoly(im, site, 255)
@@ -81,7 +99,7 @@ def get_isempty():
         masked = cv2.bitwise_or(img, img, mask=mask)
         img0 = masked
         cv2.imwrite("../module/ieemoo-ai-zhanting/imgs/1.jpg",img0)
-        pred = detect.detect(opt)
+        pred = detect.detect(opt, model, stride)
         logger.info(pred)
     except Exception as e:
         logger.warning(e)