基于边缘检测的半色调图像网点面积检测

Printed Dot Detection Based on Edge Detection

网点面积率直接影响印刷的质量,而网点面积的检测一直是印刷过程控制中的一个难点。针对此问题介绍了一种基于边缘检测计算网点面积率的方法,可以通过网点面积对印刷过程进行有效控制。使用显微摄影仪器拍摄出印刷品的网点放大图像,对网点放大图像进行高斯降噪,然后使用Canny边缘检测算法提取出网点放大图像的边缘,根据边缘信息对网点放大图像进行局部二值化,图像二值化之后通过黑白像素的比值来计算网点面积率。这种算法得到的网点面积率的值比较稳定,而且可以减少计算量。

Dot area has a direct influence on the qual ity of the presswork, and the detection of dot area has always been a difficulty in controlling the printing process. This paper introduces a method to measure the dot area of print matter based on edge detection. At first, it uses microphotography e-quipment to get the picture of printing dot. In order to reduce the influence of noise, it uses the Gaussian filter to smooth the image, then uses the Canny edge detector to abstract the edge of the dot image. The image is divided into a number of regions according to the edge information and make image binary in these region. The rate of dot area is computed by calculating the proportion of black pixels in total pixels in the binary image. Under this approach, an accurate binary image can be generated, and this method is very stable and efficient.