Ei kuvausta

Your Name 4ed3c4ac1d 修改: source/common/kafka_handler/kafkahandler_default.cpp 2 vuotta sitten
.github b18dd2e801 3.0.2 2 vuotta sitten
3rdparty 656c018f53 version 3.5.2 msg: 加入夜间过滤器的开关 2 vuotta sitten
cmake b18dd2e801 3.0.2 2 vuotta sitten
data 30d2e1a9c2 version 3.5.7 msg: 加入异类过滤器 2 vuotta sitten
docker b18dd2e801 3.0.2 2 vuotta sitten
easydk 60cdcaae45 version 3.4.2 msg: 优化cnstream运行流程、优化推理资源分配、提高稳定性 2 vuotta sitten
framework 60cdcaae45 version 3.4.2 msg: 优化cnstream运行流程、优化推理资源分配、提高稳定性 2 vuotta sitten
modules dc8b7c0a86 要提交的变更: 2 vuotta sitten
modules_contrib b18dd2e801 3.0.2 2 vuotta sitten
python b18dd2e801 3.0.2 2 vuotta sitten
source 4ed3c4ac1d 修改: source/common/kafka_handler/kafkahandler_default.cpp 2 vuotta sitten
tools bbdf8559ac version 修改流地址文件读取位置 msg: 优化cnstream运行流程、优化推理资源分配、提高稳定性 2 vuotta sitten
.dockerignore b18dd2e801 3.0.2 2 vuotta sitten
.gitignore b18dd2e801 3.0.2 2 vuotta sitten
.gitmodules b18dd2e801 3.0.2 2 vuotta sitten
CMakeLists.txt b18dd2e801 3.0.2 2 vuotta sitten
CPPLINT.cfg b18dd2e801 3.0.2 2 vuotta sitten
README.md b18dd2e801 3.0.2 2 vuotta sitten

README.md

Cambricon CNStream

CNStream is a streaming framework with plug-ins. It is used to connect other modules, includes basic functionalities, libraries, and essential elements.

CNStream provides the following built-in modules:

  • source: Support RTSP, video file, images and elementary stream in memory (H.264, H.265, and JPEG decoding).
  • inference: MLU-based inference accelerator for detection and classification.
  • inference2: Based on infer server to run inference, preprocessing and postprocessing.
  • osd (On-screen display): Module for highlighting objects and text overlay.
  • encode: Encode videos or images.
  • display: Display the video on screen.
  • tracker: Multi-object tracking.
  • rtsp_sink:Push RTSP stream to internet.

Getting started

To start using CNStream, please refer to the chapter of quick start in the document of Cambricon-CNStream-User-Guild-CN.pdf.

Samples

Classification Object Detection
Classification Object Detection
Object Tracking License plate recognization
Object Tracking License plate recognization
Body Pose Vehicle Detection
Body Pose Vehicle Detection

Best Practices

How to build a classic classification or detection application based on CNStream?

You should find a sample from samples/simple_run_pipeline/simple_run_pipeline.cpp that helps developers easily understand how to develop a classic classification or detection application based on CNStream pipeline.

This sample supports typical classification and detection neural networks like vgg resnet ssd fasterrcnn yolo-vx and so on.

This sample supports images or video file as input.

How to change the input video file?

Modify the files.list_video file, which is under the samples directory, to replace the video path. Each line represents one stream. It is recommended to use an absolute path or use a relative path relative to the executor path.

Documentation

Cambricon Forum Docs

Check out the Examples page for tutorials on how to use CNStream. Concepts page for basic definitions.

Community forum

Discuss - General community discussion around CNStream.