基于多阈值分析的织物图像倾斜检测方法

A Threshold Analysis Based on Multi-Fabric Image Skew Detection Method

研究织物图像优化检测问题,织物图像在扫描输入时难免存在倾斜,这种倾斜给后续的相关织物图像处理引起识别错误。为了有效校正识别误,提出一种新的基于多阈值分析的图像倾斜检测方法:首先利用Sobel梯度算子提取纬纱走向信息,然后通过行差运算和设定多个阈值的方法来提高算法的运算速度,利用分级Hough变换,同时采用消除伪极值、等权值均值运算、误差补偿的方法以提高其精确度检测得出纬纱的倾斜角,检测得出正确图像的倾斜角度。实验结果证明,方法不仅能够快速的检测出图像的倾斜角,而且检测结果较为精确。

Slant exists when fabric is scaning for input image,which leds to great difficulties for the subsequent image processing operations.This paper presents a new analysis based on multi-threshold image skew detection method.Firstly,using the weft Sobel gradient operator to extract information,and then through the line of difference operation,multiple threshold method is set to improve the algorithm＇s computational speed.Finally,using hierarchical Hough transform and by the methods of eliminating false extreme,computing the mean weight error and compensation,its accuracy is improved to detect the slant angle of weft,then draw the slant angle of the image.Experimental results show that this method can quickly detect the slant angle of the image,and the test results are more accurate.