阈值化分割算法及其在显著物体检测中的应用研究
毕业论文关键词： 阈值分割；物体检测；最大类间方差法；矩量保持法；算法效率
Thresholding Segmentation Algorithm and its application on Visual Object Detection
Abstract: Image Segmentation is the basic premise of Image Recognition and Image Analysis. The quality of Image Segmentation affects the effect of Image Processing directly, even determine the success or failure. Therefore, Image Segmentation plays a very important role in Digital Image Processing Technology. Object Detection is one of the focus areas of the Image Processing. It has a very wide range of applications, like aircraft aerial or satellite images road detection, surveillance video of the vehicle pedestrian detection, CT images of tumor detection, etc. This subject is on the basis of the research of Digital Image Segmentation, research a suitable Thresholding Segmentation Algorithm of Significant Object Detection. Focuses on classical thresholding method is applied to detect significant objects,
and the realization by MATLAB programming and debugging, combined with the complexity of the algorithm and experimental results, analysis algorithm performance. In short, there are many threshold segmentation method, each method has its own unique background almost advantages and practical applications. In practical applications, thresholding and other methods need to be used in conjunction with each other to get the best or satisfactory segmentation results.
Keywords: thresholding segmentation; object detection; Otsu; moment-preserving; efficiency of algorithm
目录
摘要 i
Abstract ii
目录 iii
1 绪论 1
1.1 研究背景与意义 1
1.2 国内外研究现状与水平 2
1.2.1 图像阈值分割的技术发展 2
1.2.2 图像阈值分割的研究现状与发展趋势 4
1.2.3 物体检测的技术发展和研究现状 4
1.3 本文的组织结构 5
2 阈值分割算法 6
2.1 Otsu最大类间方差法 7
2.2 Tsai矩量保持法 8
3 系统设计 10
3.1 运行环境 10
3.2 界面设计 11
4 结果分析 21
5 总结 23
5.1 工作总结 23
5.2 存在的问题及展望 23
致谢 24
参考文献 25
1 绪论