ISPLab

An open-source image signal processing (ISP) pipeline implemented by C++

Introduction

ISPLab is an open-source and platform-free ISP pipeline converting 10-bit RAW data to 8-bit RGB/YUV image without any 3rdparty libs (If you want to show the result and use deep learning, then OpenCV is required).

Result

| Raw input | RGB output | |-----------|------------| |||

Methods

The pipeline of ISPLab includes following methods:

• RAW data processing
  • Dead pixel correction (DPC)
  • Black level correction (BLC)
  • Lens shading correction (LSC)
  • Anti-aliasing filtering (AAF)
  • Auto white balance (AWB)
  • Chroma noise filtering (CNF)
• Demosaicing
• RGB image processing
  • Color correction (CCM)
  • Gamma correction (GC)
• Color space convertion
• YUV image processing
  • Noise reduction (NR)
  • Edge enhancement (EE)
  • Hue & saturatiton control (HSC)
  • Brightness & contrast control (BCC)

You can easily change the parameters of different methods in the config.csv, or change the Config.cpp file.

Contact me: liuyuqing@mail.dlut.edu.cn