Research Of Control System And Key Techniques On High Performance CMOS Image Sensor

With a great improvement on noise elimination, dark current reduction and photosensitivity enhancement, CMOS image sensors have represented the most advanced imaging technology in the field of image collecting and indicate a comprehensive application perspective. To achieve smaller area, higher stability and lower power, the integration of CMOS image sensor, controller system and image signal processing on a single-chip is the trend of CMOS image sensors development.This paper implemented an efficient controller system and image signal processing fuctions of motion detection and edges sharpening and provided a good foundation for further research. We have achieved the follwing results. 1. Designed a special I2C controller and SDRAM controller as well as a synchronous USB controller, which guaranteed a stable operation of the system and provided an efficient data transmission capability. 2. Designed a general SRAM controller with synchronous interface and programmable ability. 3. Presented an on-chip low power real-time motion detection algorithm. This algorithm uses only 2.5‰image data for the motion detection. With a programmable ability of precision, an adaptive detection-bits and threshold architecture, high light adaptive ability, high precision, low power and antinoise characteristics could be achieved. 4. Based on the 3×3 Laplacian sharpening filter, the output image delayed just one-row-long clock periods. Futher experiments suggested that distinct edges of video image could be reached by this real-time algorithm.The controllers and image signal processing modules were verified on the FPGA platform with HV7131GP CMOS image sensor. The experiments suggested that the system was stable and efficient to display the video image in low power mode, and had good image processing effects.