摘要
为解决规则目标物体特征参数的测量问题，提出了一种基于链码的目标物体图像特征参数测量方法。目标物体图像经过灰度变换、平滑去噪等一系列处理后，图像中往往还会存在一些伪边缘信息，针对图像中的这些伪边缘，对图像进行链码跟踪，利用各链码组曲率统计分析的方法剔除目标物体图像中的伪边缘，同时利用像素直线连接的方法将图像中分段的链码组组合连接；然后利用傅里叶变换高频滤波的方法平滑图像边界的曲折部分进一步对图像进行去噪；最后计算目标区域的面积和周长。经实验分析，该方法测量结果精度高，可以实现对规则目标物体特征参数的有效测量。
In order to measure the area and the volume of regular target object, a measurement based on chain code is processed. The image processed by a series of methods such as gray-level transformation and noise reduction always has the false edge. Aiming at this problem, an image processing method based on the chain code is presented. And chain codes are connected by using the method of line among pixels. High frequency filter of the image could be smooth by using the method of DTFT. Measurement of the area and perimeter of target object are realized. The experiments results show that this method can realize effective measurement of regular target object.
关键词
图像处理
链码
傅里叶
面积
周长
Image Processing
Chain Code
DTFT
Area
Perimeter
