摘要
提出了1个特征提取技术,解决识别的时间问题,同时,在新的特征技术中,识别率也有一定的提高。在模式识别系统中,识别的对象需要1个规范化的光栅图像,必须有足够的代表性、差异性等特点进行分类,以便达到很好的识别率。尤其在无线监测中,涉及的数据量比较大,响应比较慢。提出了在无线监测中快速响应的区域分块算法,并采取区域特征提取的方式来提取动、静态的图像,实验证明此方法可大大缩短了响应时间,优于以前的全局特征分析方式。对于动态图像特征的提取有一定的指导意义。
This paper proposes a feature extraction technique to solve the problem of recognition time.At the same time,the new features technology improves our recognition rate.Need for a standardized pixel raster image pattern recognition system,the object of identification,we must have adequate representation of differences in the characteristics of the classification,in order to achieve a good recognition rate.Especially in the wireless monitoring of data involved than the larger,slower response.This paper proposes to adopt a regional feature extraction,greatly reducing the response time,way better than the previous global features analysis.Some guiding significance for the dynamic image feature extraction.
出处
《电子测量技术》
2012年第9期64-66,共3页
Electronic Measurement Technology
关键词
模式识别
光栅图像
特征提取
pattern recognition
raster image
feature extraction
