摘要图像识别是人工智能的一个重要领域。现今多数视觉处理的任务,都依赖于对物体、场景的识别、分类的能力。而图像识别已在诸如基于内容的图像搜索、视频信息的挖掘及移动机器人的识别方面都有广泛的运用,因此,能否做好图像识别对于这些问题而言至关重要。经过几十年的发展,图像识别的方法已经由原先的基于全局视觉特征转向为局部视觉特征,而后者现在几乎成了事实上的应用标准。全文分为四个部分:8208
第一部分为导言部分,介绍了图像识别的研究背景和意义,并对相应的研究现状进行简要评述。
第二部分对基于局部视觉特征的图像识别方法进行概述。
第三部分着重介绍了其中关于局部特征算子的匹配的一个算法——带核的局部敏感哈希算法(Kernelized Locality-Sensitive Hashing)
第四部分为总结,总结了对图像识别算法的研究方法及方向。
关键词 局部视觉特征 核方法 局部敏感哈希算法
毕业设计说明书(论文)外文摘要
Title Study of Algorithms of Object Recognition Based on Local Features
Abstract
Object Recognition is a very important field of Artificial Intelligence. Most any vision task fundamentally relies on the ability to recognize objects, scenes, and categories. And it has been widely used in applications such as content-based image search, video miming, or object identification for mobile robots, thus, whether we can handle these problems well has been considered of great importance. After decades of developments, People has turned from global image representations to local feature representations, which in the meantime, become the standard of specific object recognition tasks.
The article is pided into four sections:
The first section is an introduction of the backgrounds and meaning of study object recognition as well as short comments on the state of art.
The second section is an overview of current methods of object recognitions based on local feature.
The third section emphasize an algorithm to match local feature descriptors -- kernelized locality-sensitive hashing(KLSH).
The fourth section is to summarize. We talk about how to study object recognition and the direction of study.
Keywords local-feature Kernel-Methods Locality-sensitive-hashing
1. 绪论 1
1.1 研究背景及意义 1
1.2 图像识别的内容及方法概述 1
1.3 论文组织结构 2
2. 基于局部特征的图像识别 3
2.1 概述 3
2.2 局部特征的检测与表示 3
2.2.1 介绍 3
2.2.2 兴趣点及区域的检测 4
2.2.3 特征的描述 7
2.2.4 特征的提取与表示小结 8
2.3 特征算子的匹配 9
2.3.1 介绍 9
2.3.2 常用方法 9
2.4 匹配点几何关系的验证 10
2.4.1 几何模型的推测 10
2.4.2 异常值的处理 11
2.4.3 几何关系验证小结 12
3. 带核的局部敏感哈希算法(KLSH) 13
3.1 提出背景 13
3.2 核方法 13
3.3 局部敏感哈希算法 14
3.4 带核的局部敏感哈希算法 15
3.5 一些讨论 18
3.6 实验结果及分析 18
4. 结论 22 基于局部视觉特征的图像识别研究:http://www.751com.cn/jisuanji/lunwen_6448.html