PaddleHub 视频口罩检测

 

临时代码：

处理带口罩的；

C:\Users\Administrator\.paddlehub\modules\pyramidbox_lite_mobile_mask\python\48913cc44e11a655acf3031cc10d7db8.py

处理不带口罩的：？

C:\Users\Administrator\.paddlehub\modules\pyramidbox_lite_mobile\python\450e1c94ad59d49eb3785afe840434d3.py

i7处理器，一次需要270ms，漏检，误检都有。
分辨率：
im_width = int(932 * 0.8)
im_height = int(611 * 0.8)

import time

import cv2

import paddlehub as hub

import paddlehub as hub
module = hub.Module(name="pyramidbox_lite_mobile_mask")
# 将模型保存在test_program文件夹之中
# module.processor.save_inference_model(dirname="test_program")

module = hub.Module(name="pyramidbox_lite_mobile_mask")


def detect_img():
    img_data= cv2.imread(r"test_program/6e1c74e85298b5120f9f884ad348c243.jpg")

    input_dict = {"data": [img_data]}


    for i in range(10):
        start=time.time()
        # execute predict and print the result
        results = module.face_detection(data=input_dict)
        for result in results:
            left=float(result['data']['left'])
            right=float(result['data']['right'])
            bottom=float(result['data']['bottom'])
            top=float(result['data']['top'])
            conf=float( result['data']['confidence'])
            color=(0, 0, 255)
            if result['data']['label'] != 'MASK':
                color = (255, 0, 0)
            cv2.rectangle(img_data, (int(left),int(top)),(int(right),int(bottom)), color, 1)
        print("time", time.time() - start)
        cv2.imshow("asdf",img_data)
        cv2.waitKey(1)

def detct_video():
    vc = cv2.VideoCapture(r"D:\project\models\capture\0.mp4")  # 读入视频文件

    index=0
    while True:  # 循环读取视频帧
        index+=1
        rval, img_o = vc.read()

        if img_o is None:
            break
        if index%2==0:
            continue
        input_dict = {"data": [img_o]}
        start = time.time()
        # execute predict and print the result
        results = module.face_detection(data=input_dict)
        for result in results:
            left = float(result['data']['left'])
            right = float(result['data']['right'])
            bottom = float(result['data']['bottom'])
            top = float(result['data']['top'])
            conf = float(result['data']['confidence'])
            color = (0, 255, 0)
            if result['data']['label'] != 'MASK':
                color = (0, 0, 255)
            cv2.rectangle(img_o, (int(left), int(top)), (int(right), int(bottom)), color, 1)
        print("time", time.time() - start)
        cv2.imshow("asdf", img_o)
        cv2.waitKey(1)

if __name__ == '__main__':
    detct_video()
# input_dict = {"image": test_img_path}

# execute predict and print the result
# results = module.face_detection(data=input_dict)
# for result in results:
#     print(result)
#
# # 预测结果展示
# img = mpimg.imread("detection_result/test_mask_detection.jpg")