基于多级定向执行长度编码快速识别物体特征  

Fast test and recognition of object characters based on multilevel orientation run length coding

在线阅读下载全文

作  者:王欢欢[1] 卞越新 王向军[2] 

机构地区:[1]天津科技大学机械工程学院,天津300222 [2]天津大学微光机电系统技术教育部重点实验室,天津300072

出  处:《光学精密工程》2014年第7期1976-1983,共8页Optics and Precision Engineering

基  金:国家教育部支撑项目(No.625010110);天津科技大学引进人才科研启动基金资助项目(No.201304003);微光机电系统技术教育部重点实验室(天津大学)开放基金资助项目(No.KF2K131001)

摘  要:提出了一种新的快速物体检测与识别方法,用于检测和识别在运动过程中图像尺寸不断变化的物体。该方法基于多级定向执行长度编码法(MORLC)生成一种新的多级字节编码和物体特征样本形式,其属于字符数据样本,具有数据量小、占用存储空间少、构造过程简单、可根据不同的应用需求选择不同级次的样本形式等特点。对MORLC方法中二级字节编码和样本的系统理论进行了分析,给出了利用MORLC二级字节编码和样本检测和识别运动物体的实验数据。对静态背景中的行人及位置进行了检测和识别实验,结果显示多级字节编码和物体特征样本形式匹配处理速度快,匹配和识别的鲁棒性好,不易产生错误匹配。该方法特别适用于场景中由于被探测物体运动产生投影图像比例发生变化的情况,即对图像大小不敏感。该方法的另一个突出优点是高的执行处理效率。A new rapid detection and recognition method was proposed for those objects whose image sizes would change in movement processing.Based on Multilevel Orientation Run Length Coding(MORLC)method,the method can generate a new multilevel byte code and object feature templet forms.It is a byte coding and templets and characterized by smaller data amounts,smaller storage addresses,easy constructing,and can choose different classes of sample forms.According to different application requirements,two level byte code in MORLC method and system theory of templets were analyzed,and the experimental data from detection and recognition of moving objects using MORLC two level byte code were given.A detection and recognition experiment on a pedestrian and its position in a static background was performed.The experimental results show that the multilevel bytecode and object feature matching templet form have higher processing speeds,robust matching and identification,and not easily get wrong matching.This method is especially suitable for changed images caused by detecting object movement in the scene projection,which means it is not sensitive to image sizes.Another advantage of this method is high processing efficiency.

关 键 词:产品分类 特征检测 特征识别 图像编码 

分 类 号:TP391[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象