The advances in solar technologies has lead higher device conversion efficiency and lower production costs. Ensuring the quality of the solar cells, however, remains a challenge and automation of defect identification in solar cells can potentially make the process efficient. Light Beam Induced Current (LBIC) mapping is a high-resolution imaging technique that allows researchers to see defects in solar cell. LBIC maps show the spatial distribution of the short-circuit current of a solar cell. In this study, surface and sub-surface defects in an LBIC image were identified using watershed segmentation, image dilation and inpainting. The image processing model on the image presented resulted in an Intersection over Union score of 41.33%. The model was able to highlight features on the LBIC map that are potential defects and further investigation can be made on these areas to understand the underlying cause of these defects.