Python 2.7/OpenCV 3.3：从立体相机捕获帧

Question

我正在使用 Python 2.7 和 OpenCV 3.3，并希望在每个定义的帧数或定义的秒数后保存来自立体相机的图像。就像每 5 帧或每 2 秒后的帧应该存储在不同的文件夹中，如 cam1folder 和 cam2folder。我能够从两个摄像头捕获视频，然后在预定义的秒数后使用这些视频捕获帧。下面给出了这些代码。

视频捕获代码：

import numpy as np
import cv2
cap = cv2.VideoCapture(0)
cap1 = cv2.VideoCapture(1)
fourcc = cv2.VideoWriter_fourcc(*'MJPG')
out = cv2.VideoWriter('output.avi',fourcc, 20.0, (640,480))
out1 = cv2.VideoWriter('output1.avi',fourcc, 20.0, (640,480))

while(True):
    ret, frame = cap.read()
    ret1, frame1 = cap1.read()
    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    out.write(frame)
    out1.write(frame1)
    cv2.imshow('frame',gray)
    cv2.imshow('frame1', frame1)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

cap.release()
out.release()
cap1.release()
out1.release()
cv2.destroyAllWindows()

保存视频后从两个摄像头捕获帧的代码：

import cv2
import math
import numpy as np
videoFile = "output.avi"
vidcap = cv2.VideoCapture(videoFile)
success,image = vidcap.read()

videoFile1 = "output1.avi"
vidcap1 = cv2.VideoCapture(videoFile1)
success1,image1 = vidcap1.read()

seconds = 3
fps = vidcap.get(cv2.CAP_PROP_FPS) # Gets the frames per second
multiplier = fps * seconds
fps1 = vidcap1.get(cv2.CAP_PROP_FPS) # Gets the frames per second
multiplier1 = fps1 * seconds


while success:
    frameId = int(round(vidcap.get(1))) #current frame number, rounded b/c sometimes you get frame intervals which aren't integers...this adds a little imprecision but is likely good enough
    success, image = vidcap.read()

    frameId1 = int(round(vidcap1.get(1)))  #current frame number
    success1, image1 = vidcap1.read()

    if frameId % multiplier == 0:
        cv2.imwrite("right/rframe%d.jpg" % frameId, image) # stored in right folder
    if frameId1 % multiplier1 == 0:
        cv2.imwrite("left/lframe%d.jpg" % frameId1, image1) # stored in left folder


vidcap.release()
vidcap1.release()
print "Complete"

但我想要的是直接从视频中实时捕获图像而不保存这些视频，而是将捕获的图像保存在两个不同的文件夹中。如何做到这一点？

谢谢。

Answer 1

您只需合并您自己编写的 2 个代码片段。

在您的第一个代码片段中，当您运行 while 循环时，您可以维护一个公共计数器，而不是为单独的相机维护单独的计数器。当您的计数器达到所需值（5 帧，根据您的问题）时，您直接对 frame 和 frame1 对象执行 imwrite。

您不需要与 VideoWriter 对象进行中介。您的最终代码如下所示：

cap = cv2.VideoCapture(0)
cap1 = cv2.VideoCapture(1)
multiplier = 5
frame_count = 0
while(True):
    frame_count += 1
    ret, frame = cap.read()
    ret1, frame1 = cap1.read()

    cv2.imshow('frame',frame)  #show all frames
    cv2.imshow('frame1', frame1)

    if frame_count % multiplier == 0:
        cv2.imwrite("right/rframe%d.jpg" % frame_count, frame) #write specific frames 
        cv2.imwrite("left/lframe%d.jpg" % frame_count, frame1)

    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

Python 2.7/OpenCV 3.3：从立体相机捕获帧

Python 2.7/OpenCV 3.3 : capture frames from stereo cameras

python-2.7

opencv3.0

camera-calibration

stereo-3d