CoastLineObjectDetection

In this project we detected common objects on sea shore, using pre-trained Caffe model.

Structure

• coastlineObjectDetection.py : The main program.
• IntersectionOverUnion.py : The code for calculate the Intersection over Union (IoU) for results of object detection.
• 7grandTruth.py : The same as IntersectionOverUnion.py.
• video folder : The videos for test.
• grandtruth folder : The grand truth images .
  • iouResults : The results of grand truth
• MobileNetSSD_deploy.caffemodel : The trained weigths are saved in this file.
• MobileNetSSD_deploy.prototxt.txt : The network structure is defined in this file.

Model

We used Caffe framework for desing model. The trained weigths are saved in MobileNetSSD_deploy.caffemodel and the network structure is defined in MobileNetSSD_deploy.prototxt.txt.

Dataset

• train : Coco dataset.
• test : Videos captured with iphon7 form Gilan sea shores which is one of the province of Iran.

Results

  

Thanks to https://www.pyimagesearch.com/