【摘要】 应用4片DSP+FPGA流水阵列结构,用现场可编程门阵列FPGA对采集的视频数字图像信息进行处理,实现了织物疵点自动检测。设计了以4片TMS320C62x为数字图像信息处理核心,由FPGA实现系统控制互连的织物疵点自动检测系统,设计了基于直方图统计和基于支撑矢量机的织物疵点分类识别算法。实验结果表明,当样布传送速度达到100、120 m/min时,该织物疵点自动检测系统对样布的疵点识别准确率分别达到80%和60%。更多还原

【Abstract】 By application of the four pieces DSP+FPGA architecture and processing the collected video digital images by the field programmable gate array(FPGA),the automatic fabric inspection is realized.It designs the four pieces TM320C62x as digital image processing core and the connecting of the fabric inspection system control is executed by FPGA.It also designs the classified identification arithmetic of fabric defects based on attributed relational histogram and support vector machine.The experiments show when the velocity of sample fabric is of 100(m/s) and 120(m/s),the automatic fabric inspection system has achieved the accuracy of fabric defect detection 80% and 60% respectively.更多还原