基于现场可编程门阵列的水果识别系统设计
A fruit identification system design based on field programmable gate array
金梅 1曾欣 1张立国 1冯景瑞 2吴文哲1
作者信息
- 1. 燕山大学电气工程学院 秦皇岛 066004
- 2. 燕山大学信息科学与工程学院 秦皇岛 066004
- 折叠
摘要
针对目前大多数水果识别系统实时性差和成本较高的问题,本文提出一种基于现场可编程门阵列(FPGA)的水果识别系统.整体硬件设计包含图像采集、检测识别和显示模块.软件部分通过图像采集平台将水果图像存储至双倍数据率同步动态随机存储器(DDR3)控制模块,进行图像灰度化处理,并创造性地提出背景帧差法对水果图像进行分割.通过融合水果颜色和几何特征,实现对水果数量、颜色和种类的识别.整个水果识别系统在不同光照下对常见水果进行了测试.实验结果表明,水果平均识别准确率达到93.25%,识别速度约为16.67 ms,FPGA硬件资源消耗率仅占28.02%,可以满足实时性需求.
Abstract
To address problems of high latency and high cost in most fruit identification systems,this study proposes a fruit recognition system based on the field programmable gate array(FPGA).The overall hardware design includes image acquisition,detection and recognition and display modules.In the software part,fruit images are stored in the double-data-rate3(DDR3)controller module through the image acquisition platform,and undergo image grayscale processing.Furthermore,a creative approach is introduced,using the background and the frame differ-ence method for segmentation of fruit images.By integrating fruit color and geometric features,recognition of fruit quantity,color,and species was achieved.The fruit identification system was tested under different lighting condi-tions and intensities.The experimental results show that the method has an average accuracy of 93.25%for the recognition of fruits,a recognition speed of approximately 16.67 ms,and the FPGA hardware design consumption rate of only 28.02%,which can meet the real-time requirements of fruits recognition.
关键词
水果识别/背景帧差法/图像处理/现场可编程门阵列(FPGA)Key words
fruit identification/background frame difference method/image processing/field programmable gate array(FPGA)引用本文复制引用
基金项目
河北省中央引导地方专项(199477141G)
河北省科学技术研究与发展计划科技支撑计划(20310302D)
出版年
2024
国家科技期刊平台
NSTL
万方数据