Pixel segmentation of Bangalore

[elboyran]

• [x] Make a script for classifying each 10th pixel for Banaglore ROIs test_classifyEachPixelFromTile_Bangalore based on test_classifyEachPixelFromTile_Kalyan

• Classify each 10th pixel for all ROIs (warning- it takes hours!) :

• [x] ROI1

• [x] ROI2

• [x] ROI3

• [x] ROI4

• [x] ROI5

• [x] Create config_params_Banagalore script containing all parameter settings.

• [x] Edit existing (and use in new below) scripts to use config_params_Banagalore.

• [x] Make a script for filling the labels of missing pixels test_fillMissingPixels_Bangalore based on test_fillMissingPixels_Kalyan .

• Run missing pixels filling to obtain full segmentation :

• [x] ROI1

• [x] ROI2

• [x] ROI3

• [x] ROI4

• [x] ROI5

• [x] Create figure demonstrating the pixel segmentation steps and include it in the paper

• [x] Make a script for interpolating via majority filter test_majorityFilterSegmentation_Bangalore based on test_majorityFilterSegmentation_Kalyan .

• Run majority filter to obtain smoother segmentation. :

• [x] ROI1

• [x] ROI2

• [x] ROI3

• [x] ROI4

• [x] ROI5

• [x] Script to display the ground truth- overlaid on the image

• [x] Script to display the ground truth only (oneclass_masks2multiclass_mask_Bangalore)

• [x] Script to display the results only and overlaid on the image (show_image_overlaid_segmentation_Bangalore)

• [x] Have versions of the overlay scripts with some specific rectangles embedded (for now hard coded!)

• [x] Generate figures and include them in the paper.