Image corner location algorithm based on sub-blocking with dynamic threshold and lifting wavelet transform

An image corner location algorithm based on lifting wavelet and sub-blocking with dynamic two-thresholds is presented. According to the gray similarity around pixels and the properties of the corner in an image, the lifting wavelet transform theory is used to obtain the high-frequency region of the image. The high-frequency image is plotted regions and dynamic two-threshold algorithm is used to locate corners precisely in every region. The results of the experiment show that the corners are precisely located by the proposed algorithm and the algorithm runs much faster.