First model version

app.py

@@ -25,8 +25,11 @@ def infer(filepath):
     )
     image = cv2.imread(filepath)
     result_polygons, result_masks, result_boxes = det_demo.run_on_opencv_image(image)
-    visual_image = det_demo.visualization(image, result_polygons, result_masks, result_boxes)
-    cv2.imwrite('result.jpg', image)
+    patchs = [image[box[1]:box[3], box[0]:box[2], :] for box in result_boxes]
+    patchs = [cv2.resize(patch, (128,32)) for patch in patchs]
+    patchs = np.stack(patchs, axis=0).transpose(0,3,1,2)
+    visual_image = det_demo.visualization(image.copy(), result_polygons, result_masks, result_boxes)
+    cv2.imwrite('result.jpg', visual_image)
     return 'result.jpg'#, pd.DataFrame(result_words)
 
 

det_demo.py

@@ -580,8 +580,10 @@ class DetDemo(object):
         green[...,2] = 0
         for mask in masks:
             image[mask] = image[mask] * 0.5 + green[mask] * 0.5
+        '''
         for box in boxes:
-            cv2.rectangle(image,(box[0], box[1]), (box[2], box[3]), (0,0,255), 5)
+            cv2.rectangle(image,(box[0], box[1]), (box[2], box[3]), (0,0,255), 2)
+        '''
         '''
         for polygon in polygons:
             pts = np.array(polygon, np.int32)