부산대학교 도서관

Large-scale product inspection is an important aspect in thin film industry to identify defects with a high precision. Although vision line scan camera (VLSC)-based inspection has been frequently implemented, it is limited to surface inspections. Therefore, to overcome the conventional drawbacks, there is a need to extend inspection capabilities to internal structures. Considering that VLSC systems have access to rich information, such as color and texture, high-resolution real-time multimodal optical synchronization between VLSC and dual spectral domain optical coherence tomography (SD-OCT) systems was developed with a laboratory customized in-built automated defect-tracking algorithm for optical thin films (OTFs). Distinguishable differences in the color and texture of the bezel area were precisely determined by the VLSC. Detailed OCT assessments were conducted to verify the detection of previously unobtainable border regions and micrometer-range sub-surface defects. To enhance the accuracy of the method, VLSC images were aided for the precise surface defect identification using OCT and the image acquisition, signal processing, image analysis, and classification of both techniques were simultaneously processed in real-time for industrial applicability. The results demonstrate that the proposed method is capable of detecting and enumerating total number of defects in OTF samples with exceptional resolution and in a cost-effective manner facilitating wide area inspection for OTF samples.