红外技术2016，Vol.38Issue(9)：765-769,5.
基于非完备字典的舰船特征提取和识别
Feature Extraction and Recognition of Ships by an Uncompleted Dictionary
许德海 1魏学明 1彭垚 1缪康 1任明艺1
作者信息
- 1. 四川长虹电子科技有限公司，四川 绵阳 621000
- 折叠
摘要
Abstract
This paper presents a novel approach of extracting ships feature by an uncompleted dictionary. First, we refer to the thought of sparse representation, and the uncompleted dictionary is constructed in terms of various ships shape in infrared images. Furthermore, using orthogonal matching pursuit algorithm to decompose the ship target signal, we obtain a ship response that describes the target feature. Finally, the classification result is decided by voting strategy. The results of experimentation indicate that the proposed approach has better performance than moments feature, compactness, rectangularity and so on.关键词
舰船目标/稀疏表示/特征提取/目标识别Key words
infrared image/image processing/feature extraction/pattern recognition分类
信息技术与安全科学引用本文复制引用
许德海,魏学明,彭垚,缪康,任明艺..基于非完备字典的舰船特征提取和识别[J].红外技术,2016,38(9):765-769,5.
