| 摘要 |
An image processing method includes the steps of setting a predetermined concentric circular area of first and second image data as a workpiece image, dividing a plurality of pixels constituting the workpiece image in the concentric circular area into a concentric circuit arrangement and extracting the plurality of pixels, generating the plurality of pixels in the divided concentric circuit arrangement as a function of the central angular position of the concentric circle, computing the initial term resulting from subjecting the luminance of the plurality of pixels in the concentric circular arrangement to Fourier expansion together with the central angular position, computing the central angular position where the initial term of Fourier expansion is maximized as a function of a plurality of concentric radii, and comparing the function with the first and second image data so as to recognize the arrangement of an object to be detected. |
