| Summary: | 碩士 === 國立屏東科技大學 === 機械工程系所 === 95 === The popularity of liquid crystal displays (LCDs) with increasingly delicate color performance and sophisticated functionality has stimulated the advance in technical levels for their peripheral products. In particular, due to the demands for more leads, finer pitch and more compact size on the driven integrated circuits (IC) for LCDs, the technology of chip on film (COF) has also progressed markedly.
In this study, an image processing method utilizing fast Fourier transformation (FFT) and inverse FFT has been developed to inspect and identify the surface defects of COF. This method can reduce the misjudgments caused by human inspections, and hence improve the yield during the fabrication processes.
The developed vision inspection system in this study is dedicated for inspecting surface morphology and deficiency of COF, which is based on a machine vision system combined with the techniques of optical imaging, image processing and automatic control. This system is capable of identifying surface dents, indentations, foreign materials, discolorations, and printed defects of COF, and could be used to take the place of human inspection in order to meet the demands for high-accuracy inspections in production lines.
|
|---|
