Image Segmentation, Compression and Reconstruction from Edge Distribution Estimation with Random Field and Random Cluster Theories

TL;DR

This paper introduces a novel image processing framework combining random field and cluster theories to achieve segmentation, compression, and reconstruction of images with minimal pixel data, supported by mathematical analysis and deep learning methods.

Contribution

It develops a new mathematical approach for image segmentation and compression using random theories, enabling efficient image reconstruction from limited pixel information.

Findings

Image can be reconstructed with just over half the pixels in worst case.

A deep belief network enables unsupervised segmentation and object detection.

Mathematical analysis provides bounds and properties of pixel intensity distributions.

Abstract

Random field and random cluster theory are used to describe certain mathematical results concerning the probability distribution of image pixel intensities characterized as generic $2D$ integer arrays. The size of the smallest bounded region within an image is estimated for segmenting an image, from which, the equilibrium distribution of intensities can be recovered. From the estimated bounded regions, properties of the sub-optimal and equilibrium distributions of intensities are derived, which leads to an image compression methodology whereby only slightly more than half of all pixels are required for a worst-case reconstruction of the original image. A custom deep belief network and heuristic allows for the unsupervised segmentation, detection and localization of objects in an image. An example illustrates the mathematical results.