| 摘要 |
The object boundary detection on the image consists in projection processed images onto matrix array camera, an intensity of each image point to a code are conversed, a code value is determined corresponding to a background level, exceeded background level value in determined and derived data between neighboring cells are compared. For each matrix array camera cell is formed the neighborhood of neighboring cells of this matrix. The intensity value in each cell is determined, a mean intensity value for each is determined and its neighborhood. The mean value derived with control point is compared and if control point intensity value is more derived neighborhood mean value relates it to the cell belonging to a gray-level image boundary, if control point intensity value is less than derived neighborhood mean value, intensity of the given cell is decreased to a background intensity value. |
