基于特征融合的行人重识别方法

Person re-identification based on feature fusion

针对行人重识别中已有方法难以解决行人图像光照、视角变化大的问题,提出了一种基于特征融合的行人重识别方法。首先利用Retinex变换对图像进行预处理;然后将CN特征与原有的颜色和纹理特征融合,并通过区域和块划分的方式提取直方图获得图像特征;最后采用不同的距离学习方法在4个数据集上进行行人重识别。实验结果表明,融合后的特征对行人图像具有更好的表述能力,实现了重识别精度的较大提升,验证了方法的有效性。

Existing methods in person re-identification cannot describe the pedestrian well for significant intra-class variationsin illumination and viewpoint.In order to solve this problem,a method based on feature fusion is proposed.Firstly,the image is pre-processed using the Rentiex algorithm.Then the CN feature is combined with the color and texture features.Histograms are extracted through regional and block partition,and then the feature of image is obtained.Finally,personre-identification is performed on four public datasets with different metric learning methods.The experimental resultsshow that the feature combined with CN achieves a stronger representation and increases the accuracy obviously.