摘要
为解决在处理和计算视觉大数据中遇到的速度瓶颈,提出了一种模糊聚类局部保存投影算法用于视觉数据的特征提取应用中。首先,通过某种方法对图像进行分割;接着,将通过统计方法对图像进行特征描述得到相应的视觉数据;然后,通过提出的模糊聚类局部保存投影对视觉数据进行特征提取;最后,通过Ada Boost对提取后的特征进行识别分类。经在国际公开的UCAS-AOD和Flower-102数据集上进行大量实验,经实验对照,结果验证了模糊聚类局部保存投影算法在视觉数据特征提取中的有效性。
In order to solve the speed bottleneck of visual big data processing and computing speed problem,a fuzzy clustering locality preserving projections algorithm is proposed for feature extraction of visual data.Firstly,the image is segmented by some method.Then,the image is characterized by statistical methods to obtain the corresponding visual data.Then,the visual data is extracted by the fuzzy clustering locality preserving projections proposed.Finally,through AdaBoost,the extracted features are identified and classified.Through a large number of experiments on the internationally published UCAS-AOD and Flower-102 datasets,the experimental results verify the effectiveness of the fuzzy clustering locality preserving projections algorithm in visual data feature extraction.
作者
张乾
杨玉成
岳诗琴
邵定琴
王林
ZHANG Qian;YANG Yu-cheng;YUE Shi-qin;SHAO Ding-qin;WANG Lin(School of Data Science and Information Engineering,Guiyang 550025,China;School of Data Science and Academic Affairs Office,Guiyang 550025,China;School of Data Science and Guizhou Minzu University,Key Laboratory of Pattern Recognition and Intelligent Systems of Guizhou ,Guiyang 550025,China)
出处
《科学技术与工程》
北大核心
2019年第29期179-183,共5页
Science Technology and Engineering
基金
国家自然科学基金(61802082,61263034,61762020)
贵州省科技厅计划基金(黔科合J字[2014]2094号)
贵州省教育厅自然科学基金(黔教合KY字[2017]129)
贵州省教育厅创新群体重大研究项目(黔教合KY字[2018]018)
教育部产学合作协同育人基金(201702044007)资助
关键词
视觉数据
特征提取
模糊聚类
局部保存投影
visual data
feature extraction
fuzzy clustering
locality preserving projections