基于svm的图像分类系统设计文档毕设论文.doc

LANZHOU UNIVERSITY OF TECHNOLOGY 毕业设计 题 目 基于SVM的图象分类系统 学生姓名 学 号 专业班级 计算机科学与技术3班 指导教师 学 院 计算机与通信学院 答辩日期 摘 要 支持向量机(SVM)方法是建立在统计学习理论基础之上的，克服了神经网络分类和传统统计分类方法的许多缺点，具有较高的泛化性能。但是，由于支持向量机尚处在发展阶段，很多方面尚不完善，现有成果多局限于理论分析，而应用显得较薄弱，因此研究和完善利用支持向量机进行图像分类对进一步推进支持向量机在图像分析领域的应用具有积极的推动作用。 本文通过支持向量机技术和图像特征提取技术实现了一个图像分类实验系统。文中首先引入了支持向量机概念，对支持向量机做了较全面的介绍；然后，讨论了图像特征的描述和提取方法，对图像的颜色矩特征做了详细的描述，对svm分类也做了详细的说明；最后讨论了由分类结果所表现的一些问题。测试结果表明，利用图像颜色矩特征的分类方法是可行的，并且推断出采用综合特征方法比采用单一特征方法进行分类得到的结果要更令人满意。 关键词：支持向量机 图像分类 特征提取 颜色矩 Abstract The support vector machine (SVM) method is based on statistical learning theory foundation, overcome the neural network classification and traditional statistical classification method of faults, and has high generalization performance. But, because the support vector machine (SVM) is still in the development stage, many still not perfect, the existing results more limited to the theoretical analysis, and the use of appear more weak and therefore study and improve the use of support vector machines to image classification support vector machine to further advance in the application of image analysis play a positive role in promoting. In this paper, support vector machine (SVM) technology and image feature extraction technology implements a image classification experiment system. This paper first introduces the concept of support vector machine (SVM), the support vector machine (SVM) made a more comprehensive introduction; Then, discussed the image characteristics of description and extraction method, the image color moment features described in detail, also made detailed instructions for the SVM classification; Finally discussed the classification results of some problems. Test results show that using the torque characteristics of the image color classification method is feasible, and deduce the comprehensive ch