TBA4560 - Geomatics, Specialization Project

split_into_quadrants.py

Requires the following libraries:

• rasterio

The program takes in a GeoTIFF to split into quadrants and outputs the four quadrants as GeoTIFFs.

The code is run with the following command:

python split_into_quadrants.py input_filepath.tif

• input_filepath.tif: The filepath for the GeoTIFF to split into quadrants. Replace "input_filepath" with the filepath of the .tif file, for example to_split.tif

retrieve_placenames.py

The program takes in a list of placenames and a list of approved name object types, and outputs a list of approved placenames.

The code is run with the following command:

python retrieve_placenames.py gml_filepath.txt approved_filepath.txt csv_filepath.txt

• gml_filepath.txt: The filepath for the GML placename data. Replace "gml_filepath" with the filepath of the .txt file, for example placenames_gml.txt
• approved_filepath.txt: The filepath for the list of approved name object types. Replace "approved_filepath" with the filepath of the .txt file, for example approved_types.txt
• csv_filepath.csv: The filepath for the CSV output file. Replace "csv_filepath" with the filepath of the .csv file, for example approved_placenames.csv

preprocessing.py

Requires the following libraries:

• numpy
• opencv-python
• rasterio

The program takes in a GeoTIFF to be preprocessed and outputs a preprocessed GeoTIFF.

The code is run with the following command:

python preprocessing.py input_file.txt threshold_value template_window_size search_window_size kernel_iterations

• input_file.tif: The filepath for the GeoTIFF to be preprocessed. Replace "input_file" with the filepath of the .tif file, for example for_preprocessing.tif
• threshold_value: The threshold used to binarize the image. Must be an int between 0 and 255. By default it is 125. Example: 125
• template_window_size: The template window size used in denoising. Must be an int. By default it is 7. Enter 0 for no denoising
• search_window_size: The search window size used in denoising. Must be an int. By default it is 21. Enter 0 for no denoising
• kernel_iterations: The number of iterations used in the thinning process. Must be an int. By default it is 1. Type 0 for no thinning

gcv.py

Requires the following libraries:

• google-cloud-vision
• pillow
• rasterio

The program takes in a GeoTIFF for text detection and outputs the detected text along with its bounding boxes.

The code is run with the following command:

python gcv.py input_file.tif output_folder

• input_file.tif: The filepath for the GeoTIFF to be used in text detection. Replace "input_file" with the filepath of the .tif file to be used, for example for_detection.tif
• output_folder: The folder in which to put the CSV output file. Replace "output_folder" with the filepath of the folder, for example detected_text

Note: When run, this program creates a temporary file dummy.jpg. This file is created and deleted by the program. Do not attempt to modify this file while the program is running.

postprocessing.py

Requires the following libraries:

• pyproj
• Levenshtein
• numpy

The program takes in a folder containing lists of detected placenames, a list of abbreviations and their expansions, and a list of relevant placenames, and gives some statistics about the unidentified, identified, and georeferenced placenames.

The code is run with the following command:

python postprocessing.py detected_folder abbreviations_file.csv placenames_file.csv levenshtein_threshold

• detected_folder: The folder containing CSVs of detected placenames. Replace "detected_folder" with the filepath of the folder file to be used, for example amtskart
• abbreviations_file.csv: The filepath for the CSV containing the list of abbreviations and their expansions. Replace "abbreviations_file" with the filepath of the .csv file to be used, for example abbreviations.csv
• placenames_file.csv: The filepath for the CSV containing the list of placenames from the given dataset. Replace "placenames_file" with the filepath of the .csv file to be used, for example placenames.csv
• levenshtein_threshold: The Levenshtein threshold used in the correlation analysis. Must be a float. By default it is 0.8

placenames.csv

Example of placenames data. First column is the placename, second is longitude, third is latitude. Coordinates given in EPSG:25832.

abbreviations.csv

Example of abbreviations data. First column is the abbreviation, second is the archaic spelling of the lengthened form, third and final is the modern spelling.