基于智能视觉的特定人员检索平台设计与实现  

Design and implementation of specific staff retrieval platform based on intelligent vision

作　　者：陈熔[1] 刘杰 CHEN Rong;LIU Jie(Jiangsu Agri-animal Husbandry Vocational College,Taizhou 225300,China;Yangzhou University,Yangzhou 225000,China)

机构地区：[1]江苏农牧科技职业学院,江苏泰州225300 [2]扬州大学,江苏扬州225000

出　　处：《现代电子技术》2017年第14期102-105,共4页Modern Electronics Technique

基　　金：国家自然科学基金(61175111)

摘　　要：以往设计出的特定人员检索平台均或多或少地存在一些问题,主要体现在检索准确率和检索效率都不高。面对此现象,设计基于智能视觉的特定人员检索平台。平台中的采集模块负责进行特定人员图像特征的采集工作。特征提取模块根据采集模块给出的特定人员图像特征,使用色彩对比度方法,将检索数据库中与其具有相似特征的候选图像筛选出来,并利用候选图像对智能视觉模块进行训练。训练成功后的智能视觉模块将依据特定人员图像特征,将候选图像中的最佳目标图像检索出来。经实验分析可知,所设计的平台拥有很高的检索准确率和检索效率,可较好地实现设计目标。The previous-designed specific staff retrieval platform has some problems more or less,and is mainly reflectedin the low retrieval accuracy and low retrieval efficiency,therefore a specific staff retrieval platform based on intelligent visionwas designed.The acquisition module in the platform is responsible for the acquisition of the specific staff image characteristics.The feature extraction module gives the image feature of the specific staff according to the image acquisition module.The colorcontrast method is used to screen out the candidate images with similar characteristic in retrieval database.The candidate imagesare adopted to train the intelligent vision module.The successfully-trained intelligent vision module is employed to retrieve theoptimal target image in candidate images according to the image feature of the specific staff.The experimental analysis resultshows that the designed platform has high retrieval accuracy and high retrieval efficiency,and can realize the design objectivebetter

关 键 词：智能视觉 目标图像检索 检索平台 候选图像 

分 类 号：TN915.5-34[电子电信—通信与信息系统] TP391.3[电子电信—信息与通信工程]