面向情感特征的风景图像分类研究及其实现

视觉活动是认知外部世界的主要途径，视觉感知源于物体的外形、色彩及纹理。而颜色特征的不同对人类情感变化起着至关重要的作用。鉴于不同风景图像会产生不同的颜色特征和情感特点，如何建立图像颜色特征与用户评价之间的关系并加以鉴别具有重要的研究意义。该文在学习前人提出的各种颜色提取分类算法的基础上，结合图像的颜色直方图对风景图像的颜色特征进行提取，并利用支持向量机的方法对图像进行分类。实验表明，该方法取得了较好的准确率。

Landscape Images Research and Its Application Oriented to Emotional Features

Visual activity is an important path for people to know and understand the outside world. As we known, people ’s visu-al perception come from the color, shape, and texture of objects, and among them, color features plays a vital role for the change of human emotions. In view of the different scenery image will produce different color characteristics and emotional characteris-tics, how to effectively extract image color features and built the mapping and identification for user evaluation has important re-search significance. Based on the previous studies proposed a variety of color extraction and classification algorithms, we apply an image color histogram method to extract the image features for the actual landscape images, and then use a classification method based on support vector machine （SVM） to identify emotional features. The experiment proves that it has obtained the high accu-racy.