Multi Object Detection and Shadow Identification in Images: State-of-Art

Abstract

When an image is classified, a class label is attached to it, and when an object is located, a bounding box is drawn around the object or objects in the image. There are two jobs involved in object detection: drawing a box around each object in the image and assigning it a name. Video surveillance, scene interpretation, and object recognition all benefit from shadow detection and removal. There may be major issues with object merging and object loss as well as object misinterpretation and form alternation in various visual applications such as segmentation, scene analysis, and tracking when shadows are ignored. It has been presented in the literature that shadow identification and removal can be accomplished using a variety of techniques. Object detection is a technology that uses moving images or digital images to recognise objects in the real world. The object is a member of a class that includes things like vehicles and people. There are several industries where object detection is critical to the safety and productivity of the business. Image recovery, safety approaches, research purposes, machine system assessments, computerised vehicle structures are only some of the many uses for object detection. A shadow is a visual phenomenon that occurs when an area of a scene is occluded by the primary source of light. Shadow detection and removal is a complex process that relies heavily on texture analysis and colour information. Machine learning-based object detection is a fundamental application of machine learning because of the powerful capabilities of feature learning and feature representation. It is a difficult problem to solve when it comes to pattern recognition and computer systems. This research presents a detailed review of shadow detection and removal techniques with object detection in images.