摘要
针对传统动态目标检测实时性差、功耗高及精度低等问题,提出了一种基于FPGA芯片的高精度、低功耗动态目标检测与防御躲避系统。该系统采集OV5640摄像头像素信息,经DDR4高速缓存后,依次对图像进行多步预处理,包括灰度化、自适应快速中值滤波、改进型三帧间差分、形态学运算及包围盒检测等,在HDMI液晶屏上实时显示目标检测结果。系统根据计算出的动态目标中心坐标,驱动电机控制摄像头进行精准的对角躲避。硬件测试结果表明,该系统在50 MHz的时钟频率下,视频帧率达到60 FPS,具有较高的实时性和鲁棒性,能够清晰并准确检测出动态目标,且防御躲避设计可行,功耗仅为0.969 W。
To address the issues of poor real-time performance,high power consumption,and low accuracy in traditional dynamic object detection,this paper proposes a high-precision,low-power dynamic object detection and evasion system based on FPGA chips.The system begins by capturing pixel information from the OV5640 camera,which is stored in DDR4 high-speed cache.Subsequently,the image undergoes a series of preprocessing steps,including grayscale conversion,adaptive fast median filtering,improved three-frame difference,morphological operations,and bounding box detection.The real-time object detection results are displayed on an HDMI monitor.Additionally,the system utilizes the calculated dynamic object center coordinates to drive the motor and control the camera accurately for precise diagonal evasion.The hardware testing results indicate that the system achieves a video frame rate of 60 FPS at a clock frequency of 50 MHz,demonstrating high real-time performance and robustness.It can accurately and clearly detect dynamic targets,and the defense evasion design is feasible,and the power consumption is only 0.969 W.
作者
衡国奇
孙科学
魏洪健
张瑛
HENG Guoqi;SUN Kexue;WEI Hongjian;ZHANG Ying(College of Electronic and Optical Engineering,College of Flexible Electronics(Future Technology),Nanjing University of Posts and Telecommunications,Nanjing 210023,China;Nation-Local Joint Project Engineering Lab of RF Integration&Micropackage,Nanjing 210023,China)
出处
《电子设计工程》
2024年第19期150-154,159,共6页
Electronic Design Engineering
基金
江苏省研究生实践创新计划资助(SJCX21_0279,KYCX21_0711)。
关键词
FPGA
动态目标检测
自适应快速中值滤波
三帧间差分法
防御躲避
FPGA
dynamic object detection
adaptive fast median filtering
three-frame differencing method
defense avoidance
作者简介
衡国奇(1999-),男,江苏徐州人,硕士研究生。研究方向:智能信号处理;通信作者:孙科学,sunkx@njupt.edu.cn。
