摘要 |
<P>PROBLEM TO BE SOLVED: To provide a learning-type defect detecting apparatus, a learning-type defect detecting method and a learning-type defect detecting program for quickly and accurately detecting a defect. Ž<P>SOLUTION: In the case in which a plurality of partial areas BLK<SP>LN</SP>s are set along with an A-A line of a learning image LNIMG, a determination unit determines that the objects (candidate image vectors), present between a lower limit threshold ThL and an upper limit threshold ThH from among distances corresponding to such partial areas BLKs, are learning objects. Meanwhile, the objects with the distances not longer than the lower limit threshold ThL, and the objects with the distances not shorter than the upper limit threshold ThH are excluded from the learning objects. Ž<P>COPYRIGHT: (C)2010,JPO&INPIT Ž
|