From Classic to Cutting-Edge: A Near-Perfect Global Thresholding Approach with Machine Learning
Image binarization is an important process in many computer-vision applications. This transforms the color space of the original image into black and white. Global thresholding is a quick and reliable way to achieve binarization, but it is inherently limited by image noise and uneven lighting. This...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-07-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/15/14/8096 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849419212048564224 |
|---|---|
| author | Nicolae Tarbă Costin-Anton Boiangiu Mihai-Lucian Voncilă |
| author_facet | Nicolae Tarbă Costin-Anton Boiangiu Mihai-Lucian Voncilă |
| author_sort | Nicolae Tarbă |
| collection | DOAJ |
| description | Image binarization is an important process in many computer-vision applications. This transforms the color space of the original image into black and white. Global thresholding is a quick and reliable way to achieve binarization, but it is inherently limited by image noise and uneven lighting. This paper introduces a global thresholding method that uses the results of classical global thresholding algorithms and other global image features to train a regression model via machine learning. We prove through nested cross-validation that the model can predict the best possible global threshold with an average F-measure of 90.86% and a confidence of 0.79%. We apply our approach to a popular computer vision problem, document image binarization, and compare popular metrics with the best possible values achievable through global thresholding and with the values obtained through the algorithms we used to train our model. Our results show a significant improvement over these classical global thresholding algorithms, achieving near-perfect scores on all the computed metrics. We also compared our results with state-of-the-art binarization algorithms and outperformed them on certain datasets. The global threshold obtained through our method closely approximates the ideal global threshold and could be used in a mixed local-global approach for better results. |
| format | Article |
| id | doaj-art-46948a5e47804f1faa4068ba2f30786e |
| institution | Kabale University |
| issn | 2076-3417 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj-art-46948a5e47804f1faa4068ba2f30786e2025-08-20T03:32:12ZengMDPI AGApplied Sciences2076-34172025-07-011514809610.3390/app15148096From Classic to Cutting-Edge: A Near-Perfect Global Thresholding Approach with Machine LearningNicolae Tarbă0Costin-Anton Boiangiu1Mihai-Lucian Voncilă2Computer Science and Engineering Department, Faculty of Automatic Control and Computers, National University of Science and Technology Politehnica Bucharest, 060042 Bucharest, RomaniaComputer Science and Engineering Department, Faculty of Automatic Control and Computers, National University of Science and Technology Politehnica Bucharest, 060042 Bucharest, RomaniaComputer Science and Engineering Department, Faculty of Automatic Control and Computers, National University of Science and Technology Politehnica Bucharest, 060042 Bucharest, RomaniaImage binarization is an important process in many computer-vision applications. This transforms the color space of the original image into black and white. Global thresholding is a quick and reliable way to achieve binarization, but it is inherently limited by image noise and uneven lighting. This paper introduces a global thresholding method that uses the results of classical global thresholding algorithms and other global image features to train a regression model via machine learning. We prove through nested cross-validation that the model can predict the best possible global threshold with an average F-measure of 90.86% and a confidence of 0.79%. We apply our approach to a popular computer vision problem, document image binarization, and compare popular metrics with the best possible values achievable through global thresholding and with the values obtained through the algorithms we used to train our model. Our results show a significant improvement over these classical global thresholding algorithms, achieving near-perfect scores on all the computed metrics. We also compared our results with state-of-the-art binarization algorithms and outperformed them on certain datasets. The global threshold obtained through our method closely approximates the ideal global threshold and could be used in a mixed local-global approach for better results.https://www.mdpi.com/2076-3417/15/14/8096document image binarizationglobal thresholdingmachine learningimage preprocessingideal global thresholdensemble learning |
| spellingShingle | Nicolae Tarbă Costin-Anton Boiangiu Mihai-Lucian Voncilă From Classic to Cutting-Edge: A Near-Perfect Global Thresholding Approach with Machine Learning Applied Sciences document image binarization global thresholding machine learning image preprocessing ideal global threshold ensemble learning |
| title | From Classic to Cutting-Edge: A Near-Perfect Global Thresholding Approach with Machine Learning |
| title_full | From Classic to Cutting-Edge: A Near-Perfect Global Thresholding Approach with Machine Learning |
| title_fullStr | From Classic to Cutting-Edge: A Near-Perfect Global Thresholding Approach with Machine Learning |
| title_full_unstemmed | From Classic to Cutting-Edge: A Near-Perfect Global Thresholding Approach with Machine Learning |
| title_short | From Classic to Cutting-Edge: A Near-Perfect Global Thresholding Approach with Machine Learning |
| title_sort | from classic to cutting edge a near perfect global thresholding approach with machine learning |
| topic | document image binarization global thresholding machine learning image preprocessing ideal global threshold ensemble learning |
| url | https://www.mdpi.com/2076-3417/15/14/8096 |
| work_keys_str_mv | AT nicolaetarba fromclassictocuttingedgeanearperfectglobalthresholdingapproachwithmachinelearning AT costinantonboiangiu fromclassictocuttingedgeanearperfectglobalthresholdingapproachwithmachinelearning AT mihailucianvoncila fromclassictocuttingedgeanearperfectglobalthresholdingapproachwithmachinelearning |