Osamu Hirose

Quantitative Inspection of Color Shading Defects on the Anti-Reflection Coating (2nd Report)

Evaluation of color shading defects on anti-reflection coating is described. Color shading defects are caused by disturbance of coating thickness. The defect is observed as slight color difference. The relation between thickness disturbance and color difference is shown by using a simulation method. MacAdams ellipsoid is effective to judge the quality of anti-reflection coating to pass or fail. Variation of color difference with thickness and observation angle is evaluated. Designing method of coating thickness, in which color shading is maintained a low profile, is suggested.

Visual Inspection of Shading Defects on Grid Structure Films by Diffraction Pattern.

This paper describes a new visual inspection method for grid structure films used for elements of display unit. The main target of inspection is SHADING DEFECT, which is caused by unevenness of grid line width. Degree of shading defects defies measurement in numerical terms, so in many cases SENSORY INSPECTION by human operators are adopted in manufacturing sites. The new method is based on a diffraction effect of laser beam. The diffraction pattern varies due to grids characteristic. Variations of the diffraction pattern with the defects have been proven by simulations and experiments. The defect detection method by image processing with CCD area sensor is proposed. The method is applied on several defective sample films. The results proved that the defects could be detected by the intensity profiles of the images. The effectiveness of this method at the manufacturing sites is described. Quantitative evaluation of shading defects is proposed.

Detection of Small Convex and Concave Defects on Optical Films by Patterned Illumination.

A new detective method for small convex and concave defects, which are observed in surfaces of optical films, have been proposed. The method is based on a phenomenon that regular patterns in a reflected image is blurred by the existences of their defects under an illumination having regular patterns. An algorithm to extract only blurred zones as defective zones is shown. The method is applied to defects on several sample films. The results show that only defects are successfully detected even if the films are curved and warped dynamically. Also the effectiveness of the method in manufacturing sites is shown by the robustness against the fluctuation of intensity of the illumination and executing time.