序列图像编码的自动识别  

Automatic Recognition of Sequence Image Coding

作　　者：仲思东[1] 高智[1] 

机构地区：[1]武汉大学电子信息学院,湖北武汉430079

出　　处：《计算机应用研究》2005年第5期148-150,共3页Application Research of Computers

基　　金：武汉市科技计划资助项目(200310040504)

摘　　要：目标的识别与匹配一直是完成图像测量和跟踪技术的关键。用图像编码的方法,可实现多幅图像中的测量点的识别与匹配,并进一步实现整个测量过程的自动化。其核心思想是对测量点进行编码加载信息,然后在多幅图像中对编码信息进行提取与识别,在确定这些点的位置信息和属性信息后即可实现识别与匹配。实验表明:该编码能在复杂的背景中被有效地识别,其识别与匹配正确率高、速度快,并且该编码方法也可以应用在测量系统的自动标定、移动目标的快速跟踪等多种场合。Recognition and matching of target is a pivotal step in the photogrammetric technology. Based on the image coding, automatic recognition and matching of measure points are realized and all the procedure of measuring are automatized. The kernel of the idea is to code the measured point, then recognize the coded information in different digital images. The automatic recognition and matching of homonymy points are realized with it’s attribute and position information. The experiment shows that this type of coded information can be recognized effectively in complicated background, and more this precept not only adapt to image measure but also adapt to the calibration of measure system and track the moving targets.

关 键 词：图像测量 图像编码 自动识别 自动匹配 同名点 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]