Learn More
Many computer vision applications, such as object recognition and content-based image retrieval could function more reliably and effectively if regions of interest were isolated from their background. A new method for regions of interest extraction from color image based on visual saliency in HSV color space is proposed in this paper. Color saliency is(More)
—With the development of content-based multimedia systems, there is a need for automatic extraction of central object from natural color images. A new method for automatic extraction of central object is presented in this paper. First, a criterion of homogeneity based on both the global and the local information for HSV color images is proposed, and is used(More)
Aimed at the application requirements of content-based image retrieval technology on the Internet, firstly, some key techniques and application ways are researched, then the principles and methods how to reduce the “gap” between low-level visual features and high-level semantic description of image are analyzed for improving the efficiency(More)
Multi-sensors fusion to construct an accurate 3D map about unknown indoor scene and outdoor in the simultaneous localization and mapping (SLAM) area is becoming increasingly popular. In this paper our methods pay an attention to how to optimize the 2D map from the indoor or outdoor unknown large-scale environment and how to make the 3D map intuitive and(More)
The algorithm based on SIFT feature matching and Kalman filter has been used for digital video stabilization, it is efficient in many applications. However, video obtained by the method is still not stable. An improved scheme in motion filtering is proposed in this paper. The scheme is that global motion vector estimated by Kalman filter is filtered by an(More)
Target tracking has always been a hot research topic in the field of computer vision. Tracking-Learning-Detection (TLD) is a new algorithm for online learning tracking proposed by Zdenek Kalal. In the algorithm, the computation consuming of detection module is relatively large. To solve this problem and improve the algorithm, we proposed an online learning(More)
  • 1