anbeime/skill / skills/video-frame-extractor/video-frame-extractor
video-frame-extractor
Unverified影片与视频编辑、自媒体创作者在需要拆解优秀视频、提取分镜参考或反推提示词时,使用本技能可自动抽取视频关键帧,调用视觉大模型批量分析画面,一键生成结构化描述与创作提示词,轻松搞定视频内容提取与场景分析。
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.0 KB |
| references/analysis-guide.md | markdown | 7.9 KB |
| scripts/coze_bot_client.py | script | 10.3 KB |
| scripts/video_frame_extractor.py | script | 6.4 KB |
| scripts/visual_analyzer.py | script | 9.2 KB |
scripts/video_frame_extractor.py
View on GitHub#!/usr/bin/env python3
"""
视频抽帧脚本
支持间隔抽帧、均匀采样,输出关键帧图片
"""
import os
import sys
import cv2
import argparse
from pathlib import Path
from typing import List, Dict, Any
def extract_frames(
video_path: str,
output_dir: str,
interval: float = 1.0,
max_frames: int = 10,
start_time: float = 0,
end_time: float = None,
resolution: tuple = None
) -> Dict[str, Any]:
"""
从视频中抽取关键帧
参数:
video_path: 视频文件路径
output_dir: 输出目录
interval: 抽帧间隔(秒)
max_frames: 最大抽帧数
start_time: 开始时间(秒)
end_time: 结束时间(秒),None表示到视频结尾
resolution: 输出分辨率,如(1920, 1080),None表示保持原分辨率
返回:
抽帧结果字典
"""
results = {
'success': True,
'total_frames': 0,
'output_files': [],
'errors': []
}
try:
# 创建输出目录
os.makedirs(output_dir, exist_ok=True)
# 打开视频文件
cap = cv2.VideoCapture(video_path)
if not cap.isOpened():
raise ValueError(f"无法打开视频文件: {video_path}")
# 获取视频信息
fps = cap.get(cv2.CAP_PROP_FPS)
total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
video_duration = total_frames / fps
print(f"视频信息:")
print(f" 帧率: {fps:.2f} fps")
print(f" 总帧数: {total_frames}")
print(f" 时长: {video_duration:.2f} 秒")
# 计算抽帧范围
if end_time is None:
end_time = video_duration
if start_time >= end_time:
raise ValueError(f"开始时间({start_time})必须小于结束时间({end_time})")
# 计算抽帧间隔(帧数)
frame_interval = int(interval * fps)
# 计算起始和结束帧号
start_frame = int(start_time * fps)
end_frame = int(end_time * fps)
# 设置起始帧
cap.set(cv2.CAP_PROP_POS_FRAMES, start_frame)
# 计算实际抽帧数
available_frames = (end_frame - start_frame) // frame_interval
actual_max_frames = min(max_frames, available_frames)
print(f"\n抽帧配置:")
print(f" 抽帧间隔: {interval} 秒 ({frame_interval} 帧)")
print(f" 抽帧范围: {start_time}s - {end_time}s")
print(f" 最大抽帧数: {max_frames}")
print(f" 实际抽帧数: {actual_max_frames}")
# 抽帧
frame_count = 0
extracted_count = 0
while extracted_count < actual_max_frames:
ret, frame = cap.read()
if not ret:
break
current_time = cap.get(cv2.CAP_PROP_POS_MSEC) / 1000.0
# 检查是否在时间范围内
if current_time > end_time:
break
# 按间隔抽帧
if frame_count % frame_interval == 0:
# 调整分辨率
if resolution:
frame = cv2.resize(frame, resolution)
# 生成文件名
frame_filename = f"frame_{extracted_count + 1:05d}.jpg"
frame_path = os.path.join(output_dir, frame_filename)
# 保存图片
cv2.imwrite(frame_path, frame, [cv2.IMWRITE_JPEG_QUALITY, 95])
# 记录结果
results['output_files'].append({
'filename': frame_filename,
'path': frame_path,
'timestamp': f"{int(current_time // 3600):02d}:{int((current_time % 3600) // 60):02d}:{int(current_time % 60):02d}",
'time_seconds': current_time
})
extracted_count += 1
print(f" 抽取帧 {extracted_count}/{actual_max_frames}: {frame_filename} (时间: {current_time:.2f}s)")
frame_count += 1
results['total_frames'] = extracted_count
# 释放资源
cap.release()
print(f"\n✅ 抽帧完成,共抽取 {extracted_count} 帧")
print(f"输出目录: {output_dir}")
except Exception as e:
results['success'] = False
results['errors'].append(str(e))
print(f"❌ 抽帧失败: {str(e)}")
return results
def main():
parser = argparse.ArgumentParser(description='视频抽帧工具')
parser.add_argument('--input', type=str, required=True, help='视频文件路径')
parser.add_argument('--output', type=str, required=True, help='输出目录')
parser.add_argument('--interval', type=float, default=1.0, help='抽帧间隔(秒),默认1秒')
parser.add_argument('--max_frames', type=int, default=10, help='最大抽帧数,默认10')
parser.add_argument('--start_time', type=float, default=0, help='开始时间(秒),默认0')
parser.add_argument('--end_time', type=float, default=None, help='结束时间(秒),默认视频结尾')
parser.add_argument('--resolution', type=str, default=None, help='输出分辨率,如1920x1080,默认保持原分辨率')
args = parser.parse_args()
# 解析分辨率
resolution = None
if args.resolution:
try:
width, height = map(int, args.resolution.split('x'))
resolution = (width, height)
except ValueError:
print(f"❌ 分辨率格式错误,应为'宽x高',如1920x1080")
sys.exit(1)
# 执行抽帧
results = extract_frames(
video_path=args.input,
output_dir=args.output,
interval=args.interval,
max_frames=args.max_frames,
start_time=args.start_time,
end_time=args.end_time,
resolution=resolution
)
# 输出结果
if results['success']:
print("\n抽帧文件列表:")
for frame_info in results['output_files']:
print(f" {frame_info['filename']} - {frame_info['timestamp']}")
# 保存结果JSON
import json
result_file = os.path.join(args.output, 'extraction_info.json')
with open(result_file, 'w', encoding='utf-8') as f:
json.dump(results, f, indent=2, ensure_ascii=False)
print(f"\n抽帧信息已保存到: {result_file}")
sys.exit(0)
else:
print("\n抽帧过程中出现错误:")
for error in results['errors']:
print(f" - {error}")
sys.exit(1)
if __name__ == '__main__':
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |