- Developed an object detection deep learning model to identify and count the number of tree logs in every truck image from the Georgia Pacific warehouse.
- Made an object segmentation tool to identify, segment and manually label the tree logs from the images to develop a labelled dataset using OpenCV and KL Divergence in python.
- Defined a deep learning model using U-Net architecture to predict the number of tree logs in over 11000 images with 78% accuracy.