2018/06/21 by Kelvin C. K. Chan, Raymond H. Chan, Chan, Kelvin C. K. +3
Computer Science · Engineering · #Advanced Neural Network Applications #Advancements in Photolithography Techniques #FOS: Electrical engineering #Image and Video Processing (eess.IV) #Medical Image Segmentation Techniques #Sparse and Compressive Sensing Techniques #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1806.08101
openalex publication_date 2018/06/21 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
The goal of edge-histogram specification is to find an image whose edge image\nhas a histogram that matches a given edge-histogram as much as possible.\nMignotte has proposed a non-convex model for the problem [M. Mignotte. An\nenergy-based model for the image edge-histogram specification problem. IEEE\nTransactions on Image Processing, 21(1):379--386, 2012]. In his work, edge\nmagnitudes of an input image are first modified by histogram specification to\nmatch the given edge-histogram. Then, a non-convex model is minimized to find\nan output image whose edge-histogram matches the modified edge-histogram. The\nnon-convexity of the model hinders the computations and the inclusion of useful\nconstraints such as the dynamic range constraint. In this paper, instead of\nconsidering edge magnitudes, we directly consider the image gradients and\npropose a convex model based on them. Furthermore, we include additional\nconstraints in our model based on different applications. The convexity of our\nmodel allows us to compute the output image efficiently using either\nAlternating Direction Method of Multipliers or Fast Iterative\nShrinkage-Thresholding Algorithm. We consider several applications in\nedge-preserving smoothing including image abstraction, edge extraction, details\nexaggeration, and documents scan-through removal. Numerical results are given\nto illustrate that our method successfully produces decent results efficiently.\n