基于全局显著性检测的手绘草图图像检索

Hand-drawn Sketches Image Retrieval Based on Global Saliency Detection

提出一种基于全局特征图像显著性的手绘草图图像检索算法。首先利用图像分割方法获得若干分区,再计算各分区的彩色稀疏直方图,然后利用直方图计算各个分区的颜色对比度。通过剔除对比度低的若干分区从而获取显著目标,将显著目标所在区域按比例分块,保存每个分块的平均灰度和空间关系,作为待搜索的特征数据。搜索过程是将手绘轮廓填充后提取特征数据与保存的特征数据按分块计算灰度的差异,并结合空间权值进行累加。最后将结果排序,值越小则相似度越高。

A hand-drawn sketches image retrieval algorithm based on global saliency detection was proposed.First, the image segmentation method was used to obtain a number of partitions, and then the partition color sparse histogram was calculated, and the histogram was used to calculate each partition color contrast.Saliency target was obtained by eliminating low contrast partitions and was divided into equal blocks .The average gray value of each block and spatial relationships were feature data for searching.Search process was filling hand draw contour and extracting the feature data and comparing the feature data on disk, then the difference was combined with the spatial weights.Finally the results were sorted, and the pictures were similar if the difference was small.This method can extract saliency targets well and can obtain accurate result.