Counting Individuals in an Image using Machine Learning Technique

Abstract

In video surveillance system, the most complex thing is to detect individuals. In recent years, the research is done using deep learning technique, which gives powerful individuals detection results. A model in Deep Learning i.e. YOLO (You Only Look Once) has explored in individual detection in all the angles of a individuals. The model is tested and trained on viewing person dataset. Further, counting individuals has done through information of classified bounding box. The trained model is going to verify by giving several testable data set and takes two datasets for training and testing. Trained model is tested rigorously to find out the accuracy of a model. This methodology gives efficient results for counting individual in real world.