In the field of image processing and computer vision, many people have tackle the problem of text recognition and classification. Many of these literature however tackle the
problem of detecting other languages other than Thai. For example, Chinese characters.
Some literature that works with Thai language requires sophisticated algorithm that requires machine learning, such as neural networks to tackle the classification problem. However, in this report, we propose and analyze more simply methods that will only require image processing techniques to classify Thai characters. This will include using the SIFT and RANSAC algorithm, the SVD and Principle Component Analysis technique, and XOR template matching method. Not only will the report concentrate on the classification part of the problem, we will also propose a preprocessing pipeline that will clean up documents with Thai languages by utilizing the structure of the language to detect where texts are within an image.