Highly Efficient Lossless Coding for High Dynamic Range Red, Clear, Clear, Clear Image Sensors

Area of Science:

• Computer Vision
• Image Processing
• Automotive Engineering

Background:

• High dynamic range (HDR) image sensors are crucial for advanced driver-assistance systems (ADAS) and autonomous driving systems (ADS).
• Efficient data compression is necessary for real-time video processing during algorithm development and testing (e.g., software-in-the-loop and hardware-in-the-loop).

Purpose of the Study:

• To present a highly efficient coding procedure for HDR RCCC image sensors.
• To achieve state-of-the-art lossless compression ratios and real-time feasibility for video data.

Main Methods:

• Development of a specialized coding procedure for HDR RCCC image data.
• Utilization of the FFV1 lossless codec for testing compression and decompression performance.
• Performance evaluation on a single Intel i7 CPU.

Main Results:

• The coding procedure demonstrated high efficiency for lossless data reduction.
• Achieved compression speeds of up to 81 frames per second (fps).
• Achieved decompression speeds of up to 87 fps.

Conclusions:

• The proposed coding procedure is effective for lossless data volume reduction in HDR imaging for ADAS and ADS.
• The procedure offers both high compression ratios and real-time performance, suitable for algorithm development and testing.