灰色数列模型在图像边缘趋势预测中的应用  

Application of gray sequence model in image edge trend prediction

作　　者：林涛[1] 吴镝[2] 

机构地区：[1]河北工业大学计算机科学与软件学院,天津300130 [2]河北工业大学控制科学与工程学院,天津300130

出　　处：《电子设计工程》2012年第3期134-136,共3页Electronic Design Engineering

基　　金：河北省科技计划重点项目(09963536D)

摘　　要：为了达到跟踪预测刀具图像边缘变化趋势的目的,采用了灰色数列GM(1,1)模型预测拟合直线段斜率数列的方法。在分析灰色数列预测模型的基础上,通过建立了图像边缘点拟合直线斜率数列预测模型,并用建立的模型对一帧刀具图像进行边缘趋势预测,将得到的预测结果与实际值进行比较的试验。并以最大误差分析刀具后续边缘点分布趋势,为图像采集平台进给方向提供依据。根据实验结果,得出采用灰色数列模型预测刀具边缘轮廓方法完全满足平台跟踪刀具边缘轮廓要求的结论。To reach the purpose of predicting and tracing the trend of tool image edge, adopted a method to predict the slope sequence of the fitting straight lines with Gray sequence GM （1,1） model Prediction. Analysed the Gray Sequence Prediction Model, finished a trial of building a prediction model of the slope sequence of the lines fitted by points on image edge, predicting the trend of a frame of image with established model, and comparing the prediction result with the real value. Then analysed distribution trend of the succession edge points under the maximum deviation to offer the basis for feeding direction of image capture platform. Based on the experiment result, drawn the conclusion of that it was completely satisfied to trace tool edge with The Gray Sequence Model Prediction.

关 键 词：灰度模型 数列预测 刀具图像 边缘变化趋势 图像跟踪 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]