The aim of this project is to detect an Object in an Image using a Mask RCNN with a ResNet101 backbone. The model segments the images creating masks for each object detected. The implementation is a customized implementation of a Pysource's public repository using Python 3 and a COCO pre-built model.
The repository includes:
-
Source code of Radio Tower Detection with pre-trained weights.
-
Radio Tower Dataset with annotations.
The code is documented and designed to be easy to extend. If you use it in your research, please consider citing this repository (bibtex below).
The libraries needed can be found directly in the Radio_Tower_Detection_Mask_RCNN.ipynb notebook, they will be installed before training.
Using Google Colab is recommended for this project.
- Radio_Tower_Detection_Mask_RCNN.ipynb Is the easiest way to start. It shows a step by step example of using a Mask R-CNN pre-trained model for Object Detection.