基于帧间差分法的目标检测研究与FPGA实现

针对以往帧间差分算法的硬件实现架构中数据同步信号设计复杂,调试周期较长的特点,设计了一种新型的硬件算法架构。整个系统基于Altera的Cyclone Ⅳ E系列进行开发,采用镁光的SDRAM作为高速储存器件,豪威科技的OV5640作为图像采集器件。该系统对采集到的原始图像数据进行灰度化处理,利用硬件差分算法模块对灰度图像进行差分处理与二值化,实现对视频图像中动态目标的捕捉。基于Quratus Ⅱ与Modelsim联合开发平台对系统进行硬件设计和功能模块仿真。硬件测试结果表明,系统能够清晰地检测出视频中的动态目标,且具备较好的实时性。

Research on Target Detection and FPGA Implementation Based on Inter-frame Difference Method

Aiming at the characteristics of complex data synchronization signal design and long debugging period in the hardware implementation architecture of the inter-frame difference algorithm in the past, a new type of hardware algorithm architecture is designed. The entire system is developed based on Altera’s Cyclone Ⅳ E series, using magnesium SDRAM as a high-speed storage device, and Howe’s OV5640 as an image acquisition device. The system performs gray-scale processing on the collected original image data, and uses the hardware differential algorithm module to perform differential processing and binarization on the gray-scale image to realize the capture of dynamic targets in the video image. Based on the joint development platform of Quratus Ⅱ and Modelsim, the hardware design and function module simulation of the system are carried out. The hardware test results show that the system can clearly detect the dynamic target in the video and has good realtime performance.