计算机应用研究2016，Vol.33Issue(11)：3495-3498,3512,5.DOI:10.3969/j.issn.1001--3695.2016.11.065
基于Gentle AdaBoost算法的牛体检测
Gentle AdaBoost based cow detection
摘要
Abstract
To solve the problem of low detection rate when targets occur great changes in shape,appearance and light illumi-nation conditions,taking cow detection for example,this paper proposed a Gentle AdaBoost algorithm based cow body detec-tion method.Firstly,it created feature dictionary by improved BOF.Then the cow target was extracted by the dictionary.Fi-nally,it processed training and classifying for the feature vectors of the data set to obtain the classification model of cow object and scene.Experiments show that the detectors trained by the proposed algorithm can achieve reliable detection results even in non-uniform and deformed situations.关键词
牛体检测/BOF特征向量/特征提取/Gentle AdaBoost/训练分类器Key words
cow detection/BOF feature vector/feature extraction/Gentle AdaBoost/trained classification分类
信息技术与安全科学引用本文复制引用
高韬,李凤民,杨耀权,王菁..基于Gentle AdaBoost算法的牛体检测[J].计算机应用研究,2016,33(11):3495-3498,3512,5.基金项目
国家自然科学基金资助项目（51306058，71102174）；河北省科技支撑计划重点资助项目（15212204D）；中央高校基本科研业务费专项资金青年资助项目（2014QN46）；国家留学基金委高级研究学者及访问学者（含博士后）资助项目（201406735004）；天津市科技支撑重点计划资助项目 （）
