基于灰色预测模型的图像边缘检测

简要地介绍了灰色系统理论和灰色预测模型GM(1，1)，并将该模型和图像边缘检测有机地结合在一起，提出了一种新的图像边缘检测算法，对提出的算法进行了相应的仿真实验。仿真结果表明，该算法能有效地检测出图像的边缘，尤其在检测细密的条纹方面有明显优势。

Image Edge Detection Based on Grey Prediction Model

The grey prediction model GM(1,1) is a nonlinear predictor which can be used to set up a prediction model with 4 data. We are the first to use GM(1,1) for image edge detection. To every element of an image, we use a few elements nearby to set up a GM(1,1) model and predict the element, and a prediction image is formed. The error between the original image and the prediction image forms an error image. The error image is then separated into two sub-images: the positive error sub-image and the negative error sub-image. If the error is relatively big, it is an edge element. So we can find out the edge elements by setting a right threshold. The simulation result shows that our method can detect the edge of image efficiently, and there are obvious advantages especially for detecting thin and dense stripes.