OpenCV实现视频拼接 - 迷途小书童的Note迷途小书童的Note

文章目录

软硬件环境

anaconda with python 3.8
opencv 4.5

前言

为达到将2个视频拼接成一个的目的，本文使用了最简单的逐帧拼接的方法，要求输入视频的尺寸一致。

代码实现

import cv2
import numpy as np

# 分别读取2个视频
videoLeft = cv2.VideoCapture('left.mp4')
videoRight = cv2.VideoCapture('right.mp4')

# 获取fps、宽和高
fps = videoLeft.get(cv2.CAP_PROP_FPS)
width = (int(videoLeft.get(cv2.CAP_PROP_FRAME_WIDTH)))
height = (int(videoLeft.get(cv2.CAP_PROP_FRAME_HEIGHT)))

videoWriter = cv2.VideoWriter('output1.mp4', cv2.VideoWriter_fourcc('m', 'p', '4', 'v'), fps, (width, height))

# 分别从2个视频中读取第一帧的数据
successLeft,frameLeft = videoLeft.read()
successRight, frameRight = videoRight.read()

while successLeft and successRight:

    # resize操作，都是高度不变，宽度缩小一半，这样拼接后的就是原来的宽和高了
    frameLeft = cv2.resize(frameLeft, (int(width / 2), int(height)), interpolation=cv2.INTER_CUBIC)
    frameRight = cv2.resize(frameRight, (int(width / 2), int(height)), interpolation=cv2.INTER_CUBIC)

    # numpy操作，在水平方向堆叠数组ndarray，也就是图片堆叠，垂直方向上的操作是vstack
    frame = np.hstack((frameLeft, frameRight))

    # 保存处理后的视频帧
    videoWriter.write(frame)

    # 读取下一帧数据，循环处理
    successLeft,frameLeft = videoLeft.read()
    successRight, frameRight = videoRight.read()

# 释放资源
videoWriter.release()
videoLeft.release()
videoRight.release()

下面两张是原始视频中的图片

opencv concat