我有一个在图像上运行的脚本,可识别人脸并返回类似:
的列表[('Mike', (142, 464, 365, 241)),('Garry', (42, 364, 65, 141)),('unknown', (242, 564, 465, 341))]
第二个元组是识别出的面的边界框。我还有一个使用网络摄像头的脚本,可以识别帧中的面孔并将其显示在视频供稿中。我想在出现时自动在每个帧中保存带有“未知”标签的面孔。我的代码:
from stat_face_recog import runonimage
video_capture = cv2.VideoCapture(0)
while True:
ret, frame = video_capture.read()
cv2.imwrite("new.png",frame)
final_pred = runonimage(img_path = "new.png")
read_img = cv2.imread("new.png")
for name, (top, right, bottom, left) in final_pred:
cv2.rectangle(frame, (left, top), (right, bottom), (0, 0, 255), 2)
cv2.rectangle(frame, (left, bottom - 35), (right, bottom), (0, 0, 255), cv2.FILLED)
font = cv2.FONT_HERSHEY_DUPLEX
cv2.putText(frame, name, (left + 6, bottom - 6), font, 1.0, (255, 255, 255), 1)
unknown_counter = 0
if name == "unknown":
unknowns_name = "unknown" + str(unknown_counter) + ".png"
(new_top, new_right, new_bottom, new_left) = (int(0.8 * top), int(1.2* right), int(1.2*bottom), int(0.8*left))
cv2.imwrite(unknowns_name,read_img[new_top:new_bottom, new_left:new_right])
unknown_counter += 1
cv2.imshow('Video', frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
video_capture.release()
cv2.destroyAllWindows()
但是问题是,当它识别出未经训练的人物照片时,并没有保存那些未知的面孔。每次仅保存一个名为“ unknonw0.png”的图像。我的代码有什么问题?
答案 0 :(得分:2)
您正在将for循环中的unknown_counter
重置为零。因此,图像每次都被覆盖。只需将其移动到循环之外即可。