home / skills / dkyazzentwatwa / chatgpt-skills / video-thumbnail-extractor

video-thumbnail-extractor skill

safe

This skill extracts frames from videos at set times or intervals, detects best frames, and generates thumbnail grids for previews.

npx playbooks add skill dkyazzentwatwa/chatgpt-skills --skill video-thumbnail-extractor

Review the files below or copy the command above to add this skill to your agents.

Files (3)

SKILL.md

794 B

---
name: video-thumbnail-extractor
description: Extract frames from videos at specific timestamps or intervals, find best frames, and generate thumbnail grids for previews.
---

# Video Thumbnail Extractor

Extract frames and create thumbnails from videos.

## Features

- **Frame Extraction**: Extract at timestamps or intervals
- **Best Frame Detection**: Find sharpest/brightest frames
- **Grid Previews**: Contact sheet thumbnails
- **Batch Processing**: Process multiple videos
- **Multiple Formats**: PNG, JPG output

## CLI Usage

```bash
python video_thumbnail_extractor.py --input video.mp4 --time 00:01:30 --output thumb.jpg
python video_thumbnail_extractor.py --input video.mp4 --grid 4x4 --output preview.jpg
```

## Dependencies

- moviepy>=1.0.3
- pillow>=10.0.0
- numpy>=1.24.0

Overview

This skill extracts frames from videos at specific timestamps or regular intervals, identifies the best frames based on sharpness and brightness, and generates contact-sheet thumbnail grids for quick previews. It supports batch processing and common image formats (PNG, JPG), making visual summaries fast and scriptable. The tool is implemented in Python and integrates with common media libraries for reliable results.

How this skill works

Given an input video, the skill seeks to specified timestamps or samples frames at a fixed interval, decodes frames using a media backend, and converts them to image buffers. It evaluates each candidate frame using simple heuristics (edge/contrast metrics for sharpness and pixel intensity for brightness) to rank the best frames. It can output single thumbnails or compose a grid/contact sheet of multiple frames into one preview image.

When to use it

Create preview thumbnails for video libraries or CMS items.
Automatically select representative frames for video players or catalogs.
Generate visual summaries for long recordings or batch archives.
Produce contact-sheet previews for review and QA workflows.
Extract frames at precise timestamps for documentation or analysis.

Best practices

Supply exact timestamps for guaranteed frame capture when precision matters.
Use interval sampling for long videos to reduce processing time while still getting representative frames.
Adjust sharpness and brightness thresholds to match your content style and lighting.
Resize or crop frames to consistent aspect ratios before composing grids.
Batch process files during off-peak hours to minimize resource contention.

Example use cases

Generate a 4x4 contact sheet to preview a movie reel for an editor.
Extract a thumbnail at 00:01:30 for a video listing on a website.
Scan a folder of lecture recordings and pick the sharpest frame from each for a course catalog.
Create automated preview images for a video hosting platform using interval sampling.
Produce JPG thumbnails for social sharing and PNG grids for internal review.

FAQ

Which formats are supported for output?

The skill exports common image formats such as PNG and JPG; input relies on the video formats supported by the underlying media library.

Can it process multiple videos at once?

Yes, batch processing is supported so you can run the extractor across many files in a single operation.

How are the best frames chosen?

Frames are ranked using heuristics like edge-based sharpness (contrast) and average brightness; thresholds and weights are configurable for different content types.