Target Detection Using Sparse Representation With Element and Construction Combination Feature

In this paper, we propose a target detection method using sparse representation with element and construction combination (ECC) feature. The proposed method consists of the following main steps. First, the dense scale-invariant feature transform descriptors of source image are extracted as the element features and correlations between each patch in the image are computed as the construction features. The two kinds of features are combined to represent the image. Then, the ECC feature is coded as sparse vector through a trained dictionary, and a feature histogram of sparse vector is computed based on spatial pyramid. Finally, the feature histogram is fed into support vector machine classifier. The targets are detected in the activation map which is generated from the classifier. Experimental results demonstrate that the proposed method can detect targets with high performance.