# sherlockchou86/VideoPipe

A cross-platform video structuring (video analysis) framework based on CV models & mLLM.

Repository: https://github.com/sherlockchou86/VideoPipe
Canonical: https://ross.abutalabs.com/products/videopipe
Homepage: http://www.videopipe.cool
Language: C++
License: Apache-2.0
License Family: permissive
Topics: cv, deep-learning, object-detection, opencv, gstreamer, video-analysis, ai, behaviour-analysis, similarity-search, deepstream, face-recognition, feature-extraction, image-classification, image-segmentation, license-plate-recognition, reid, llm, openai, ollama, multimodal-large-language-models
Last push: 2026-02-25T02:50:57+00:00

## Health v2 (maintenance only)
Score: 54/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 69, release rhythm 8, longevity 100
- inputs: {"age_days": 1468, "days_push": 189, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2931, forks 462 (observed 2026-08-28T04:07:30.679175+00:00)

## What it is
VideoPipe is a C++ framework for video analysis and structuring that works like a pipeline of independent, composable nodes. It integrates CV models and multimodal LLMs for tasks like face recognition, license plate recognition, image search, and traffic behavior analysis, with cross-platform deployment and visual pipeline debugging.

## Use cases
- build a face recognition system from RTSP camera streams
- detect traffic violations like illegal parking or running red lights
- recognize license plates from IP cameras and report structured data
- search images by similarity across video streams
- analyze human behavior in security footage
- integrate YOLO or custom CV models into a video pipeline
- debug video pipeline performance bottlenecks with a visual UI
- run multimodal LLM inference on video frames

## When to choose
- you need a portable, open-source alternative to NVIDIA DeepStream or Huawei mxVision
- you want a low learning curve and minimal dependencies for video analysis in C++
- you need to combine multiple CV models and LLM APIs in one pipeline
- you deploy across diverse hardware platforms (edge devices, servers, different SDKs)
- you want built-in plugins for stream reading, decoding, inference, tracking, and message pushing

## When to avoid
- you need maximum performance on NVIDIA GPUs and are fine with vendor lock-in (DeepStream is faster)
- you want a deep learning training framework rather than an inference/integration pipeline
- your project is Python-first and you prefer Python tooling
- you need a turnkey commercial product rather than a framework to build on

## Facets
- artifact type: framework
- maturity: active
- function: video-processing, computer-vision, image-processing, machine-learning, deep-learning, llm-inference, search-engine, plugin-system, streaming
- domain: computer-vision, image-processing, artificial-intelligence, deep-learning, security, cross-platform
- platform: cross-platform, cpp
- tags: video-structuring, pipeline-framework, object-detection, face-recognition, license-plate-recognition, reid, gstreamer-alternative, deepstream-alternative, multimodal-llm, traffic-analysis, opencv-dnn, tensorrt, onnxruntime, rknn, video, linux, gpu

## Member repositories
- sherlockchou86/VideoPipe (main) score 54

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:30.679175+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:33:25.066707+00:00, confidence not recorded.
  - readme: https://github.com/sherlockchou86/VideoPipe (fetched 2026-08-28T04:07:30.679175+00:00, sha fb7232479321)
  - homepage: http://www.videopipe.cool (fetched 2026-08-29T09:48:40.753385+00:00, sha 4d58872da0a4)
  - site_page: http://www.videopipe.cool/index.php/about (fetched 2026-08-29T09:48:40.762816+00:00, sha 5765f162ca54)
- Data as of 2026-08-30T08:39:29.467469+00:00.