The following repository is a result of the lectures
- Image processing
- Data Mining and Neurocomputing
in winter term 2021/2022 at Alpen-Adria-Universität Klagenfurt.
It contains Python implementations of basic and more advanced image processing techniques and how those techniques can be used and combined.
Moreover it contains two small projects which ended up being bigger than first expected. Both projects aim to classify certain objects, more precise
-
circuit-recognition aims to classify simple electrical components in handwritten circuit sketches
-
signature-verification aims to classify validity of handwritten signatures.
Beside data preparation the exercises and both projects use/combine image processing and data mining techniques and procedures like
- Image preprocessing: Filtering, Thresholding, Thinning, Segmentation, Closing, etc.
- Feature extraction: HOG features, SIFT features, Edge/Line/Circle detection, etc.
- Machine Learning: k-means clustering, visual bag of words, support vector machine and others.
For a detailed explanation what is happening in the code, we refer to the reports and to the literature listed in the reports.
We also refer to the homepages
https://scikit-learn.org/stable/
since these libraries were used throughout all examples and both projects.
Have fun with exploring and testing!