Real Time Object Detection

Abstract

REAL TIME OBJECT DETECTION is a desktop based application that will detect multiple objects in an image and draw the bounding boxes around them. It comes under the field of computer vision. It will detect the activity performed by different entities and can be helpful in security and surveillance. The language used for this project will be PYTHON. Mainly we will use OS, tensorflow, keras, colorsys and yolo3. The framework used in this project is SLIMYOLOv3. SlimYOLOv3 with fewer trainable parameters and floating point operations (FLOP) compared to the original YOLOv3 as a promising solution for real-time object detection in UAV. The front-end of this app is developed with tkinter framework which helps us to build an easy and feasible desktop application. As machine learning is evolving day by day new frameworks, models and datasets are releasing which will help us to give more accuracy. That’s why we use SlimYOLOv3 algorithm and COCO dataset which helps us to give result with the help of fps.