摘要 |
A method for surveillance is disclosed. Firstly, video of an area is captured. The video includes a first image and a second. Each image is subdivided into a number of subdivisions. A contrast value matrix is then calculated for each image by calculating a contrast value for each subdivision. A compared value matrix of the second image is calculated by comparing each element of the contrast value matrix associated with the current image with a corresponding element of the contrast value matrix associated with the prior image. Next, the compared value matrix is weighted using a preset weight matrix to locate the movement in the second image. Finally, the aim of capturing is adjusted based upon the location of movement. |