A fast image detection method using variance and variance covariance matrix

In this paper a new simple efficient algorithm for image descriptor using variance covariance matrix and variance calculations is presented. Also an approach to use this descriptor for image recognition is described. The recognition proceeds by matching individual features to a database of features from known images using a fast nearest-neighbor algorithm. We apply the full search algorithm to detect the matched image from a database which consists of 300 images. To reduce the search space, we use the variance constraint in our search. Feature matching is a simple nearest neighbor search under the distance metric and performs extremely rapidly using the variance and variance covariance matrix. The algorithm allows significant acceleration in the matching process. Experimental results are presented on image database. These results confirm the effectiveness of the proposed algorithm.