ConveyorVision is an innovative real-time system designed to automate the counting and tracking of cement bags on conveyor belts. Utilizing cutting-edge deep learning techniques like YOLOv8 for object detection and Byte tracker for precise tracking, ConveyorVision accurately monitors cement bags as they traverse the conveyor belt. Its seamless integration, reliable counting at the referee line, and robust performance in complex environments make it a valuable tool for optimizing industrial processes and enhancing productivity.
- NVIDIA Driver (Official Download Link)
- CUDA Toolkit (Official Link)
- Miniconda (Official Link)
- PyTorch (Official Link)
- Ultralytics YOLOv8 (Official Link)
- ByteTracker (Official Link)
- Supervision (Official Link)
- Onemetric (Official Link)
- Create conda env.
- Install dependencies into env.
- Annotate your datasets of cement bags. A good online data annotation tool is Roboflow or VGG Image Annotator. A
data.yaml
file must get created along withtrain
,valid
andtest
folders containing the images and labels. - Follow Official Link to train network and generate
yolo8.pt
file with your network architecture of choice, along with your dataset.
- Update the
video file
and.pt
file paths incounter.py
in themain()
function. - Run
python counter.py
inside your conda env.
A brief REPORT can be read to better understand the algorithm.
See the LICENSE file for details.