Texture Classification is a fundamental issue in computer vision and image processing, playing a significant role in many applications such as medical image analysis, remote sensing, object recognition, document analysis, environment modeling, content-based image retrieval and many more.
In this work, we propose a very simple deep learning network for image classification which comprises only the very basic data processing components: cascaded principal component analysis (PCA), binary hashing, and block-wise histograms.
Ranked #16 on Image Classification on MNIST
Thereby, we rely on the freely available LUCAS topsoil dataset.
We propose a novel CNN architecture, wavelet CNNs, which integrates a spectral analysis into CNNs.
Its architecture is indeed well suited to object analysis by learning and classifying complex (deep) features that represent parts of an object or the object itself.
For that reason, transforming these diagrams in a way that is compatible with machine learning is an important topic currently researched in TDA.