Object Detection In Medical Images Using Yolo

Abstract

In real-life scenario, the task of object detection is used in many applications like medical image systems, pedestrian detection and etc. Object detection is the method of identifying objects in the real world using pre-defined object detection approaches. It is a very important task in computer vision tasks. The object detection approach comprises a model database, robust hypothesis, feature detection block, and hypothesis verifier. Feature detection blocks identify the features from the input image. The hypothesis takes the features and searches for candidate objects. The candidate objects are checked against the hypothesis for object class generation. Hypothesis and hypothesis verification blocks link to the model database, which contains the pre-labelled class corresponding objects related to the respective object detection method. The development of an object detection model based on deep learning for medical and non-medical images is the objective. The object detection model was developed based on deep learning. The model was initially tested on non-medical images and subsequently used on medical images. Automated feature selection, extraction, and fusion are essential building blocks for the object detection model, thus making the model appropriate for the detection of medical and non-medical images.