Project information
- Category: WPF.NET App
- Language: C#
- Project date: January, 2022
-
Image Source:
https://github.gatech.edu/omscs6601/assignment_5
https://omscs.gatech.edu/cs-6601-artificial-intelligence
Automatic Image Processing
The goal of this project was to perform automatic image compression and point cloud segmentation, using k-means clustering and Gaussian mixture models. First, a mixture of multivariate Gaussians (GMM) was trained to represent an image using the raw image data. Then the model was segmented into a fixed number of components as a simplified representation of the image. Finally, the image was regenerated from the segmented data and the GMM. To optimize this process, Bayes Information Criterion was calculated and used to determine the number of components that resulted in the most accurate posterior image regeneration. Vectorization, by use of the Numpy library, was used throughout the program while working with the day to improve program efficiency.