The handwritten-writer-identification from muhammad-sayed-mahdy

muhammad-sayed-mahdy / handwritten-writer-identification Goto Github PK

View Code? Open in Web Editor NEW

handwritten writer identification using classical machine learning pattern recognition techniques

License: GNU General Public License v3.0

Python 99.32% Shell 0.68%

handwritten-writer-identification's People

Contributors

Stargazers

Watchers

handwritten-writer-identification's Issues

Dataset

Hello there! I was going through your project and was trying to run it. I am unable to understand how you have assembled data. Can you please explain it a little? I am using a dataset from kaggle which has folders named like 000 001 and so on. It has total of 571 folders with varying number of images. Can you please elaborate a little. Would be of great help.
Hoping to receive a response from your side!

Error

Thanks for your work. I wanted to try your work, but I get an error.

D:\WriterCSharper\handwritten-writer-identification-main>python src/main.py
Traceback (most recent call last):
  File "D:\WriterCSharper\handwritten-writer-identification-main\src\main.py", line 25, in <module>
    res,time = pipeline.pipe(feature='cslbcop', clf='svm',_verbose=VERBOSE, _mode=MODE, test_folder=test_folder)
  File "D:\WriterCSharper\handwritten-writer-identification-main\src\pipeline.py", line 139, in pipe
    X_tune,y_tune = step_1(train_images,_verbose, feat=feature)
  File "D:\WriterCSharper\handwritten-writer-identification-main\src\pipeline.py", line 56, in step_1
    list_images = preprocessing.preprocess(image)
  File "D:\WriterCSharper\handwritten-writer-identification-main\src\preprocessing.py", line 14, in preprocess
    xy_coord_lines = get_lines_coord(thresh)
  File "D:\WriterCSharper\handwritten-writer-identification-main\src\preprocessing.py", line 37, in get_lines_coord
    y_start = lines_indexes[ np.argmax(lines_indexes > (lines_indexes[0]+150) ) ]
IndexError: index 0 is out of bounds for axis 0 with size 0

How can I fix this?

Recommend Projects