Real-time Dsp-based Low Light Level Video Image Processing System Software Design,

It is an important disquisitive direction that applying video processing technology, which is centered about high-speed digital signal processor (DSP), to improve low-light-level (LLL) and infrared video systems' image quality and systems' vision distance. It has significant military value. And it also enlarges the applying domain of DSP and digital signal processing technology. So it has great significance. This dissertation comparative roundly introduces the international and internal developmental status, and summarizes the hardware-realizing scheme of LLL video processor centered about DSP. Based on them, the system software of LLL video processor centered about DSP is studied and designed. The main content of the dissertation is designing and realizing the total scheme and frame of system software, and the technical realizing method is brought forward that system software is divided into two relatively independent part: one is system control software, the other is image sampling and processing software. In the part of system control software design, the principium of I2C bus and SPI bus is studied, and the data transmission of I2C bus and SPI bus is simulated by MCU software. Consequently the whole system's effective management by MCU is carried out. In the part of video data transmission and storage, the dissertation analyzes the bottleneck problem of video data transmission during video sampling and data output. And the optimization resolvent of Ping-Pong dual buffer and data combination is brought forward. And in the course of data transfer the conversion of video format is also implemented. In the last part, the elements of some typical LLL image processing arithmetic and some key problems such as data storage in system is also probed into On the ground, the software realization method of LLL image processing arithmetic in high speed DSP image processing system is designed, and the video collectivity processing mode "motion detection + noise filter + gray shifting" and its realization means is also studied.The system debugging result indicated that the system software design and image processing scheme are correct and valid.