python - 这个网络摄像头面部检测有什么问题？

Question

Dlib 有一个非常方便、快速和高效的对象检测例程，我想制作一个类似于示例 here 的很酷的面部跟踪示例.

得到广泛支持的 OpenCV 具有相当快的 VideoCapture 模块(五分之一秒可以拍摄快照，而调用一些程序来唤醒网络摄像头并获取图片则需要 1 秒或更多)。我将其添加到 Dlib 中的人脸检测器 Python 示例中。

如果您直接显示和处理 OpenCV VideoCapture 输出，它看起来很奇怪，因为显然 OpenCV 存储 BGR 而不是 RGB 顺序。调整后，它可以工作，但速度很慢:

from __future__ import division
import sys

import dlib
from skimage import io


detector = dlib.get_frontal_face_detector()
win = dlib.image_window()

if len( sys.argv[1:] ) == 0:
    from cv2 import VideoCapture
    from time import time

    cam = VideoCapture(0)  #set the port of the camera as before

    while True:
        start = time()
        retval, image = cam.read() #return a True bolean and and the image if all go right

        for row in image:
            for px in row:
                #rgb expected... but the array is bgr?
                r = px[2]
                px[2] = px[0]
                px[0] = r
        #import matplotlib.pyplot as plt
        #plt.imshow(image)
        #plt.show()

        print( "readimage: " + str( time() - start ) )

        start = time()
        dets = detector(image, 1)
        print "your faces: %f" % len(dets)
        for i, d in enumerate( dets ):
            print("Detection {}: Left: {} Top: {} Right: {} Bottom: {}".format(
                i, d.left(), d.top(), d.right(), d.bottom()))
            print("from left: {}".format( ( (d.left() + d.right()) / 2 ) / len(image[0]) ))
            print("from top: {}".format( ( (d.top() + d.bottom()) / 2 ) /len(image)) )
        print( "process: " + str( time() - start ) )

        start = time()
        win.clear_overlay()
        win.set_image(image)
        win.add_overlay(dets)

        print( "show: " + str( time() - start ) )
        #dlib.hit_enter_to_continue()



for f in sys.argv[1:]:
    print("Processing file: {}".format(f))
    img = io.imread(f)
    # The 1 in the second argument indicates that we should upsample the image
    # 1 time.  This will make everything bigger and allow us to detect more
    # faces.
    dets = detector(img, 1)
    print("Number of faces detected: {}".format(len(dets)))
    for i, d in enumerate(dets):
        print("Detection {}: Left: {} Top: {} Right: {} Bottom: {}".format(
            i, d.left(), d.top(), d.right(), d.bottom()))

    win.clear_overlay()
    win.set_image(img)
    win.add_overlay(dets)
    dlib.hit_enter_to_continue()


# Finally, if you really want to you can ask the detector to tell you the score
# for each detection.  The score is bigger for more confident detections.
# Also, the idx tells you which of the face sub-detectors matched.  This can be
# used to broadly identify faces in different orientations.
if (len(sys.argv[1:]) > 0):
    img = io.imread(sys.argv[1])
    dets, scores, idx = detector.run(img, 1)
    for i, d in enumerate(dets):
        print("Detection {}, score: {}, face_type:{}".format(
            d, scores[i], idx[i]))

从该程序的计时输出来看，似乎处理和抓取图片各花费五分之一秒，因此您会认为它应该每秒显示一到两次更新 - 但是，如果您举手它会在 5 秒左右后显示在网络摄像头 View 中!

是否有某种内部缓存阻止它获取最新的网络摄像头图像？我可以调整或多线程网络摄像头输入过程来修复延迟吗？这是在具有 16gb RAM 的 Intel i5 上。

更新

根据此处，它建议读取逐帧抓取视频。这可以解释它会抓取下一帧和下一帧，直到它最终 catch 它正在处理时抓取的所有帧。我想知道是否有一个选项可以设置帧速率或将其设置为丢帧，然后只需现在单击网络摄像头中的面部图片即可阅读？ http://docs.opencv.org/3.0-beta/doc/py_tutorials/py_gui/py_video_display/py_video_display.html#capture-video-from-camera

Answer 1

我感受到了你的痛苦。事实上，我最近使用了那个网络摄像头脚本(多次迭代；大量编辑)。我觉得我让它工作得很好，我想。为了让您看到我做了什么，我创建了一个包含详细信息的 GitHub Gist(代码；HTML 自述文件；示例输出):

https://gist.github.com/victoriastuart/8092a3dd7e97ab57ede7614251bf5cbd