Dust defect is always a critical issue to degrade image qualities in kinds of imaging capturing modules, such as compact camera modules (CCMs) and digital single lens cameras (DSLRs). Dust particle may be deposited on the surface of lens or sensor in the step of assembly and even in the step of changing lens module by users. There have been lots of studies about image defects produced by dust particles on the surface of lens or sensor and the image compensation for dust defect on images, but the studies about the direct analysis of dust defects are still deficient. In this study, geometrical optics was used to analyze the image defects due to dust in the imaging capturing module. A defocused optical system with a planar light source with Lambertian emission was developed to analyze dust defect with image process and the concept of just noticeable difference (JND) of human vision. The proposed defocused optical system provides a way to help manufacturers to improve their production line efficiently.
Dust is one of the most critical issues in assembly of Compact Camera Module (CCM) for mobile phones. Defect due to dust entry or dust deposit severely degrades image quality. There have been lots of literatures about the compensating of dust defect on images by image processing, but the discussion about where the dust locates is still deficient. Dust may sneak in the CCM in any step of packaging process, so the analysis of the dust location may be useful for improving of the production line. This work develops an optical inspection algorithm to detect the location of dust inside CCM based on imaging optics. A planar light source with uniformly emission is designed as the capture target. A series of defocused images are then taken and analyzed. According to the dependence of the image defect on the capture distance, the location of the dust can be well defined. This inspection algorithm provides an easy and efficient way to help manufacturers improve their packaging process.