處理規則排列的金屬線經過加工(修整)後的切割寬度，本論文提出一種透過一維離散傅立葉轉換以及反離散傅立葉轉換，搭配影像統計法的處理方式，利用工業用CCD取得之影像，達到高速、準確的判斷加工結果(是否切斷)以及回報加工寬度的資訊。透過研究中所提之模糊度評價函數的使用，可針對某些因素(乘載平台機構的不穩定、CCD模組的晃動)導致的影像模糊情況加以剔除，進而提高處理影像的正確性以及可靠度。搭配雷射修整設備，可以即時反應修整結果，免除因修整錯誤而導致生產成本增加。進而提升雷射修整設備競爭力。 For the laser trimming width of thick film resistor (circuit), the image processing method combined with one dimension discrete Fourier transform and with image statistics is proposed. Via captured image from industrial CCD, the trimmed result and trimmed width are inspected with high speed and high accuracy. With the blur evaluation function, the blurred images are inspected and eliminated from normal cases, and the validity and accuracy can be increased. The blurred image is caused by unstable loading mechanism or vibration of CCD module. With the inspection method introduced in this research, the trimming error could be avoided, and the competitiveness of the laser trimmer could be significantly improved.
