基于GM(1,1)图像边缘检测改进算法研究

由于传统的边缘检测算法会产生信息漏检以及伪影等问题,为了更有效地检测出图像边缘信息,提出了改进的基于灰色预测模型的图像边缘检测方法.该方法首先在图像考察点邻域选择建模数据,对建模数据进行平移与对数变换的预处理,通过引入增强因子p的GM(1,1)模型对建模数据进行处理,得到原始图像的预测图像,通过对预测图像与原始图像做差得到边缘检测结果.为了统一标准,将图像库中图像转换为256*256的bmp格式图像,使用改进后的方法对转换后80幅图像进行边缘检测,其结果表明该算法可以清晰地检测出图像的边缘信息,同时图像的细节也很好地保留下来.说明该改进方法的性能较未改进方法有了大的提高.

An Improved Algorithm of Image Edge Detection Based on GM(1,1)Model

Aim.Conventional edge detection algorithms suffer from the shortcomings of undetected and artificial information.We now propose an improved method based on grey model(GM)for improving considerably the performance of image edge detection.In the full paper,we explain in some demih how to improve the grey model and use it to improve the effect of image edge detection.In this abstract,we just add some pertinent remarks to listing the four topics of explanation:(1)why grey prediction model Can be used for image edge detection;(2)how to realize image edge detection based on GM(1,1);(3)the improved algorithm and its application;(4)the results of analysis.In the second topic,we illustrate the complete edge detection process as shown in fig.2.In the third topic,we apply the multi-point modeling method to avoid the production of random noise,and add an enhancement parameter P in GM(1,1)to avoid information blur.Finally,we use the improved algorithm to process 80 standard images with bmp format and 256*256 resolution,and take an aircraft image for example to illustrate the result.The results,given in Figs.7 and 8,show preliminarily that the improved algorithm can yield more precise and useful edge information.In addition,the details of the image can be obtained.