图像中的文本字符存在于杂乱的背景之中,拍摄视角的不同使得文本具有较大的几何变形,再加上存在光照变化、字符颜色不统一等现象会导致背景分离和文本识别困难.为此提出一种基于图像文本区域的图像聚类方法.该方法首先对自然场景图像中已定位的文本区域提取局部特征描述,并使用随机投影方法将局部特征矢量集映射为固定维的特征向量,然后对包含图像文本区域的图像进行聚类.这种方法避免了由图像分割与字符识别带来的困难.实验结果表明,该方法可以对包含文字的自然场景图像有效地进行聚类,聚类的准确率能达到86.66％.%Text characters in the images are always in a complex background and the diflerent film perspective causes the text characters with large geomelrir deformation. The illuminulion or the character color is always not uniform, leading to difficul-ties in the background separating and text recognition. This paper proposed a method which was based on the image text area for image clustering. Firstly, this method extracted local feature description of the targeted text area in the image and used ran-dom projection method to map the local feature vectors into a vector which had a specific dimension. Then to cluster these ima-ges which contained text characters. The method avoided the difficulties of image segmentation and character recognition. Ex-perimental results show that the method can be effective clustering for the natural scene images that contain text characters and the accuracy is 86. 66% .
展开▼
