两个视频合并
sh
apt-get update && apt-get install -y ffmpeg
pip install opencv-python代码
py
import cv2
import numpy as np
def combine_video_with_image(config):
"""
将视频和图片合并为一个新的视频
Args:
config: 配置对象，包含以下字段：
video_path: 视频文件路径
image_path: 图片文件路径
output_path: 输出文件路径
position: 视频位置，可选值：left/right/top/bottom/center，默认为"right"
video_width: 视频宽度，默认为None（自动计算）
video_height: 视频高度，默认为None（自动计算）
video_size_ratio: 视频相对于图片的大小比例，默认0.3
x_ratio: 视频水平位置的比例（0-1），默认0.5
y_ratio: 视频垂直位置的比例（0-1），默认0.5
on_top: 视频是否在图片上层，默认True
quality: 输出视频质量（0-1），默认0.95
limit_fps: 限制输出视频帧率，默认None（使用原视频帧率）
"""
# 默认配置
default_config = {
"video_path": None, # 视频文件路径
"image_path": None, # 图片文件路径
"output_path": None, # 输出文件路径
"position": "right", # 视频位置：left/right/top/bottom/center
"video_width": None, # 视频宽度
"video_height": None, # 视频高度
"video_size_ratio": 0.3, # 视频大小比例
"x_ratio": 0.5, # 水平位置比例
"y_ratio": 0.5, # 垂直位置比例
"on_top": True, # 视频是否在图片上层
"quality": 0.95, # 输出视频质量
"limit_fps": None # 限制帧率
}
# 使用用户配置更新默认配置
if config is None:
config = default_config
else:
config = {**default_config, **config}
# 读取视频和图片
cap = cv2.VideoCapture(config['video_path'])
image = cv2.imread(config['image_path'])
# 获取视频的基本信息
fps = int(cap.get(cv2.CAP_PROP_FPS))
frame_width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
frame_height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
# 处理视频尺寸
if config['video_width'] and config['video_height']:
new_video_width = config['video_width']
new_video_height = config['video_height']
else:
# 如果设置了比例，根据图片尺寸计算视频大小
if config['video_size_ratio'] > 0:
new_video_width = int(image.shape[1] * config['video_size_ratio'])
new_video_height = int(frame_height * (new_video_width / frame_width))
else:
new_video_width = frame_width
new_video_height = frame_height
# 根据position确定输出视频的尺寸和视频位置
if config['position'] in ['left', 'right']:
output_width = image.shape[1] + new_video_width
output_height = max(image.shape[0], new_video_height)
elif config['position'] in ['top', 'bottom']:
output_width = max(image.shape[1], new_video_width)
output_height = image.shape[0] + new_video_height
else: # center
output_width = image.shape[1]
output_height = image.shape[0]
# 创建视频写入器
fourcc = cv2.VideoWriter_fourcc(*'mp4v')
out = cv2.VideoWriter(config['output_path'], fourcc, fps, (output_width, output_height))
while True:
ret, frame = cap.read()
if not ret:
break
# 调整视频帧大小
frame = cv2.resize(frame, (new_video_width, new_video_height))
# 创建输出帧
if config['position'] in ['left', 'right', 'top', 'bottom']:
# 创建空白背景
combined_frame = np.zeros((output_height, output_width, 3), dtype=np.uint8)
# 确定图片位置
if config['position'] == 'left':
combined_frame[0:image.shape[0], new_video_width:] = image
combined_frame[0:new_video_height, 0:new_video_width] = frame
elif config['position'] == 'right':
combined_frame[0:image.shape[0], 0:image.shape[1]] = image
combined_frame[0:new_video_height, image.shape[1]:] = frame
elif config['position'] == 'top':
combined_frame[new_video_height:, 0:image.shape[1]] = image
combined_frame[0:new_video_height, 0:new_video_width] = frame
elif config['position'] == 'bottom':
combined_frame[0:image.shape[0], 0:image.shape[1]] = image
combined_frame[image.shape[0]:, 0:new_video_width] = frame
else: # center
# 复制图片作为背景
combined_frame = image.copy()
# 计算视频的位置（根据比例）
x = int(config['x_ratio'] * (output_width - new_video_width))
y = int(config['y_ratio'] * (output_height - new_video_height))
# 确保位置在有效范围内
x = max(0, min(x, output_width - new_video_width))
y = max(0, min(y, output_height - new_video_height))
if config['on_top']:
# 视频在上层
combined_frame[y:y+new_video_height, x:x+new_video_width] = frame
else:
# 视频在下层
roi = image[y:y+new_video_height, x:x+new_video_width].copy()
combined_frame[y:y+new_video_height, x:x+new_video_width] = frame
# 覆盖回原始图片
combined_frame[y:y+new_video_height, x:x+new_video_width] = cv2.addWeighted(combined_frame[y:y+new_video_height, x:x+new_video_width], 0.3, roi, 0.7, 0)
# 写入合并后的帧
out.write(combined_frame)
# 释放资源
cap.release()
out.release()
# 使用 FFmpeg 添加原视频的音频到合并后的视频
try:
import subprocess
temp_output = config['output_path'] + '.temp.mp4'
import os
os.rename(config['output_path'], temp_output)
# 合并视频和音频
cmd = f'ffmpeg -i {temp_output} -i {config["video_path"]} -c:v copy -map 0:v:0 -map 1:a:0 -shortest {config["output_path"]}'
subprocess.run(cmd, shell=True)
# 删除临时文件
if os.path.exists(temp_output):
os.remove(temp_output)
except Exception as e:
print(f"添加音频失败: {str(e)}")
# 示例：合并视频和图片
combine_video_with_image({
'video_path': 'demo/shuziren_video.mp4',
'image_path': 'demo/demo.png',
'output_path': 'demo/merged_image_shuziren.mp4',
'position': 'right'
})
# 示例：在图片中央显示视频，视频位于图片右下角(比例0.8, 0.8)
combine_video_with_image({
'video_path': 'demo/shuziren_video.mp4',
'image_path': 'demo/demo.png',
'output_path': 'demo/merged_image_shuziren_1.mp4',
'position': 'center',
'video_width': 200,
'video_height': None,
'x_ratio': 0.1,
'y_ratio': 0.1
})