/Real-Time-Object-Counting

This is a computer vision project which can detect objects in real time using webcam. It can count different types of object present in the frame.

Primary LanguageJupyter Notebook

Object-Count-using-YOLOV3

Demo:

object.detection.count.In.realtime.example.mp4

Description:

1.This project can detect object in real time by analysis every frame.

2.YoloV3 have been used for object detection which is trained on COCO dataset with 80 classes.

2.The yolo3.weights can be find in the followwing link: https://drive.google.com/file/d/1aQn8Lh1wVhIKMkn0wWxvRt12p1-4EBlO/view?usp=sharing

3.In "object count In realtime example.mp4" I have recorded the output of real time object count.

4.In "example1" and "example2" the output of object count from image is shown.

How to Run:

  1. For installing all dependdecy use requirements.txt file
  2. For detecting object from image run Count object from image(YOLOV3).ipynb file
  3. For detecting object from real time run Real time object count (YOLOV3).ipynb

Direction:

  1. coco.names, yolov3.cfg, yolo3.weights file have to be in the same folder
  2. Press "q" in real time object detection for breaking the loop.