Abstract
While the computer vision algorithms perform better in terms of object recognition, these algorithms fail to detect the occluded objects. This research study performs object recognition through 2D image processing and aims to tackle the primary causes of occlusion, specifically by identifying suitable illumination, extracting objects and features from various perspectives of a camera. Numerous illumination sources such as Fluorescent, Quartz Halogen – Fiber Optics, LED, Mercury, Xenon, High Pressure Sodium, etc. exist. This study considers LED lighting due to its extended lifespan and cost efficiency. An analysis of diverse lighting techniques is underway in this research. Additionally, the proposed approach evaluates and enhances the image quality acquired through various lighting techniques, presenting the recommended ones. To address occlusion, a new algorithm has been proposed for the selected lighting techniques, focusing on handling the occluded objects. The outcomes obtained from this algorithm are compared and analyzed against existing methods.