作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2011, Vol. 37 ›› Issue (9): 239-241,244. doi: 10.3969/j.issn.1000-3428.2011.09.084

• 图形图像处理 • 上一篇    下一篇

基于图像质量等级的自动目标识别效果评估

秦富童1,2,岳丽华1,万寿红1   

  1. (1. 中国科学技术大学计算机科学与技术学院,合肥 230027;2. 中国人民解放军63880部队,河南 洛阳 471003)
  • 出版日期:2011-05-05 发布日期:2011-05-12
  • 作者简介:秦富童(1985-),男,硕士研究生,主研方向:目标识别效果评估,图像处理;岳丽华,教授、博士生导师;万寿红,讲师

Performance Evaluation of Automatic Target Recognition Based on Image Quality Level

QIN Fu-tong  1,2, YUE Li-hua  1, WAN Shou-hong  1   

  1. (1. Institute of Computer Science and Technology, University of Science and Technology of China, Hefei 230027, China; 2. No.63880 Force of PLA, Luoyang 471003, China)
  • Online:2011-05-05 Published:2011-05-12

摘要: 通过对图像质量度量指标类型进行统计分析,提出一种简单的图像质量等级划分方法,在此基础上利用支持向量机对各图像子集进行目标识别,分析图像质量对目标识别效果的影响,结合传统的目标识别效果评估方法,给出一种基于图像质量等级的目标识别效果评估方法。实验结果证明,与传统方法相比,该方法的评估结果更准确。

关键词: 目标识别效果评估, 图像质量, 自动目标识别, 主成分分析, 支持向量机

Abstract: By statistically analyzing the type of image quality measure indexes, this paper proposes a simple method for dividing image quality into different levels. Support Vector Machine(SVM) is used to recognize targets of each image subset, and the effect of image quality to target recognition is analyzed. Combined with traditional performance evaluation method of Automatic Target Recognition(ATR), a new performance evaluation method based on image quality levels is presented. Experimental result proves that the new method evaluate more accurately.

Key words: performance evaluation of target recognition, image quality, Automatic Target Recognition(ATR), Principal Component Analysis (PCA), Support Vector Machine(SVM)

中图分类号: