update crop face

@@ -203,7 +203,8 @@ class Face_Detector(object):
                                    (resize_b[13], resize_b[14])])
     
         # import time
-        # cv2.imwrite('results/0.jpg', frame)
+        # image_name = str(time.time()) + '.jpg'
+        # cv2.imwrite('results/{}'.format(image_name), frame)
         return face_bboxes, face_landmarks, max_index
                
 

@@ -17,7 +17,7 @@ class Face_Recognizer(object):
         for i in range(len(imgs)):
             img = imgs[i] 
 
-            # img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
+            img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
             img = img.astype(np.float)
             img = (img / 255. - 0.5) / 0.5
             img = img.transpose(2, 0, 1)

@@ -24,10 +24,16 @@ def preprocess(image, landmarks):
     warped = cv2.warpAffine(image, M, (112, 112), borderValue=0.0)
     return warped
 
+def get_face(image, bbox):
+    print(bbox)
+    face_area = image[bbox[1]:bbox[3], bbox[0]:bbox[2], ]
+    norm_image = cv2.cvtColor(face_area, cv2.COLOR_BGR2RGB).astype(np.float32)
+    norm_image = cv2.resize(norm_image, (112, 112))    
+    return norm_image
 
 def get_norm_face(image, landmarks):
     norm_image = preprocess(image, landmarks)
-    norm_image = cv2.cvtColor(norm_image, cv2.COLOR_BGR2RGB).astype(np.float32)
+    # norm_image = cv2.cvtColor(norm_image, cv2.COLOR_BGR2RGB).astype(np.float32)
     norm_image = cv2.resize(norm_image, (112, 112))
     # norm_image = norm_image.transpose((2, 0, 1))
     # norm_image = norm_image.transpose((1,2,0))
@@ -47,7 +53,7 @@ if __name__ == '__main__':
 
     face_detector = Face_Detector(det_face_model_path)
     face_recognizer = Face_Recognizer(reg_face_id_model_path)
-    for i in range(10):
+    for i in range(1):
         id_image = cv2.imread(id_image_path)
         life_image = cv2.imread(life_image_path)
          
@@ -55,13 +61,17 @@ if __name__ == '__main__':
         life_face_bboxes, life_face_landmarks, life_max_idx = face_detector.detect(life_image, face_det_thr)
         print(id_face_bboxes)
         print(life_face_bboxes)
+        # id_norm_image = get_face(id_image, id_face_bboxes[id_max_idx])
         id_norm_image = get_norm_face(id_image, id_face_landmarks[id_max_idx])
-        id_norm_image = np.transpose(id_norm_image, (2, 0, 1))
+        cv2.imwrite('results/face1.jpg', id_norm_image)
+        # id_norm_image = np.transpose(id_norm_image, (2, 0, 1))
         norm_images = [id_norm_image]
 
         for j in range(len(life_face_landmarks)):
             life_norm_image = get_norm_face(life_image, life_face_landmarks[j])
-            life_norm_image = np.transpose(life_norm_image, (2, 0, 1))
+            cv2.imwrite('results/face2.jpg', life_norm_image)
+            # life_norm_image = np.transpose(life_norm_image, (2, 0, 1))
+            # life_norm_image = get_face(life_image, life_face_bboxes[j])
             norm_images.append(life_norm_image)
 
         embeddings = face_recognizer.recognize(norm_images)