摘要
对于织物纬斜检测问题,提出了一种基于DFT与投影映射的纬斜检测算法;该算法将织物图像傅里叶变换的频谱图进行二值化和形态学滤波,以消除噪声点,然后在横轴上进行多角度投影映射得到织物纬斜量,克服了噪声对检测精度的干扰;同时,文章设计了基于DSP+FPGA的摄像式整纬器检测装置硬件系统,充分利用了DSP的高速数据处理能力及FPGA的强大逻辑控制功能,实现控制与运算的分离,并将文章所提出的纬斜检测算法在该硬件系统中验证;实验结果表明,该系统能够实时地对织物纬斜进行检测,测量误差在0.5度以内,纬斜检测精度高,具有一定的应用价值。
For fabric image skew detection, this paper proposes a novel method based on DFT and multi--projection mapping algorithm. This method performs binarization and morphological filtering on the DFT spectrum of the fabric images to eliminate noise points. Then make the multi--angle projection mapping on the horizontal axis to obtain the fabric skew amount which overcomes the noise interference on the detection accuracy. And this paper designs camera--type skew detection device based DSP and FPGA which takes full advantage of high--speed data processing capabilities of DSP and the powerful logic control function of FPGA which realizes the separation of control and computing, and verifies the proposed skew detection algorithm in the hardware system. The experimental results show that the system is capable of real --time fabric skew detection with the measurement error less than 0. 5 degrees, which has high accuracy and certain application value.
出处
《计算机测量与控制》
北大核心
2013年第5期1294-1296,1300,共4页
Computer Measurement &Control
基金
国家自然科学基金资助项目(61071175)
郑州市科技攻关项目(121PPTGG363-8)
