基于FPGA的刀具图像预处理系统研究
|
摘要
针对刀具检测质量不断提高的要求,提出了一种基于FPGA的图像预处理系统。该系统将CMOS传感器作为图像的采集模块,DDR2 SDRAM作为高速存储模块,ALTERA公司的FPGA芯片作为主控中心和处理模块,从而实现了将刀具从快速采集、存储、预处理到显示的整个过程。对比分辨率为256×1040的图像在Open CV与FPGA上进行双边滤波处理的运算时间,发现在FPGA上进行刀具预处理的运算时间优于Open CV。这是由于FPGA采用了并行流水线的设计,其运算时间基本忽略不计。基于FPGA的刀具图像预处理系统具有获取图像分辨率高的特点,并实现了实时处理,有助于后续的缺陷检测和磨损分析。
Aiming at the improvement of the tool detection quality requirements,an image preprocessing system based on FPGA is proposed. The system takes a CMOS sensor as the image acquisition module,DDR2 SDRAM as the high-speed storage module,and the FPGA chip as a main control center and processing module. In this way,the entire process of quickly acquiring,storing,preprocessing and displaying the tool is realized. Comparing the operating time of bilateral filter processing with Open CV and FPGA on the same resolution image of 256 × 1040,it can be found that the image preprocessing operation time on FPGA is better than Open CV. The reason is that the FPGA uses a parallel pipeline design and its computational time is negligible. The tool image preprocessing system based on FPGA has the characteristics of high resolution of image acquisition and real-time processing,which is helpful for subsequent defect detection and wear analysis.
Aiming at the improvement of the tool detection quality requirements,an image preprocessing system based on FPGA is proposed. The system takes a CMOS sensor as the image acquisition module,DDR2 SDRAM as the high-speed storage module,and the FPGA chip as a main control center and processing module. In this way,the entire process of quickly acquiring,storing,preprocessing and displaying the tool is realized. Comparing the operating time of bilateral filter processing with Open CV and FPGA on the same resolution image of 256 × 1040,it can be found that the image preprocessing operation time on FPGA is better than Open CV. The reason is that the FPGA uses a parallel pipeline design and its computational time is negligible. The tool image preprocessing system based on FPGA has the characteristics of high resolution of image acquisition and real-time processing,which is helpful for subsequent defect detection and wear analysis.
引文
[1]刘智,夏春蕾,戴曙光.基于FPGA的图像锐化实现[J].微型机与应用,2012,31(3):40-41,45.
[2]李玲.刀具磨损便携式视觉在位检测技术研究[J].南方农机,2018,49(6):93.
[3]陈雷明,杨润泽,张治,等.刀具检测方法综述[J].机械制造与自动化,2011,40(1):49-50,144.
[4]谭延凯,段智敏,杨华栋,等.刀具磨损检测的图像处理技术研究[J].科技创新导报,2010(26):123.
[5]夏军,庞征斌,李小芳,等.一种高性能DDR2控制器的设计与实现[J].计算机工程与科学,2010,32(7):62-64,98
[6]张志强,王万玉.一种改进的双边滤波算法[J].中国图象图形学报,2009,14(3):443-447.
[7]张俊涛,王园伟,庞多.一种硬件加速Open CV的图像处理方法研究[J].微型机与应用,2015,34(22):41-43.
[2]李玲.刀具磨损便携式视觉在位检测技术研究[J].南方农机,2018,49(6):93.
[3]陈雷明,杨润泽,张治,等.刀具检测方法综述[J].机械制造与自动化,2011,40(1):49-50,144.
[4]谭延凯,段智敏,杨华栋,等.刀具磨损检测的图像处理技术研究[J].科技创新导报,2010(26):123.
[5]夏军,庞征斌,李小芳,等.一种高性能DDR2控制器的设计与实现[J].计算机工程与科学,2010,32(7):62-64,98
[6]张志强,王万玉.一种改进的双边滤波算法[J].中国图象图形学报,2009,14(3):443-447.
[7]张俊涛,王园伟,庞多.一种硬件加速Open CV的图像处理方法研究[J].微型机与应用,2015,34(22):41-43.