| 摘要 |
PURPOSE:To make generation of algorithm easy which recognizes object shapes by changing the amount of data taken in from picture information obtained from an inspection object according to the size of the object and speed up the inspection. CONSTITUTION:In a compression factor setting part 1 of a size standardization device 10, actual inspection object size is read out of picture signal after A/D conversion (7) and the reference inspection object size out of a data base 14, and it is decided how many times (alpha) the inspection object is larger than the reference one. Based on the multiple alpha, thinning-out process part 2 does thinning with the use of land width of picture signal pattern and sizes in the longitudinal directions, for example, the picture data just as one of pixels is extracted and converted into standardization size pattern. Here, the pattern size becomes 1/alpha of the original. A reference value correcting part 3, based on the multiple alpha, corrects actual numbers of pixel of parts width and land width, etc. With these thinning and standardization, the process time for inclination code generation is shortened, for easier generation of recognition algorithm. |
