A Decompression System Designed For Removing Dark Current

Compared with CCD image sensor, the CMOS image sensor has low power consumption, and high integration. However, because of higher random noise and greater fixed pattern noise and dark current noise, the dynamic range and SNR of CMOS image sensor cannot reach the level of the CCD image sensor, which limits the CMOS image sensor in the field of application.In this paper, the inconsistencies of the CMOS image sensor data are demonstrated firstly, and the solution is made to eliminate dark current one by one pixel. Then a digital decompression system, based on built based on the SPI bus, is designed to eliminate dark current in the CMOS image sensor. The decompression system, based on DPCM-Huffman compression algorithm, compresses data and stores in Flash, and then the CMOS image sensor read the compressed data from Flash and decoded by Huffman-DPCM. Finally the dark current data is removed from the original image data and the CMOS dark current is removed. In the decompression system, the most important thing is the degree of compression and accuracy problems. In the data decompression system, the compression ratio increased to 3.22. And the MSE (mean square error) reduces to 15-20, which is close to the original data significantly. At the same time relative to the DCT compression algorithm and wavelet compression algorithm, the PSNR (peak signal to noise ratio) and SNR (signal to noise ratio) signals to improve the accuracy of the extracted 4-5 and 0.4-0.5. The design process is through advancing of algorithm and verification, Verilog design and verification. Finally, it downloaded to the FPGA after logic synthesis. The system is stable and efficient to achieve the stated objectives and finish the CMOS image sensor data compression and the removing of dark current.