Published Paper Details:

According to research people have usually loved to have a seat for themselves assigned beforehand, it maybe because people might feel overwhelmed about not having a seat during an important conference though it is often neglected it is true that a person does go through a bit of stress when in between a conference or be it a ballet, therefore taking this as a problem statement this thesis is focused on the study of seating occupancy. With Convolution neural network being a beneficent recognizer and processor of the image, this thesis of Seating Occupancy was made possible. Seat occupancy monitoring involves techniques to sense the presence of a human occupant and the algorithm to relay the information and display the seat is occupied or unoccupied. We used COCO dataset along with YOLOv3 which has already been trained with weights for the thesis. COCO dataset is having more 380k images of different objects of world. From this it is easy to find a perfect dataset for persons and chair. The dataset has image with names of object labeled in Coco.names file. Also YOLO makes the idea of detecting object according to need easy in terms of speed and accuracy. As our goal is to detect unoccupied chair or seats in a frame, we need to find the chairs in the image which is not overlapped by any person or other object, For that it is very important to have perfect camera angle with best image quality so noise in image should not disturb the goal.