Adaptive histogram equalization and its variations
Computer Vision, Graphics, and Image Processing
Automatic, adaptive, brightness independent contrast enhancement
Signal Processing
Discrete-time signal processing (2nd ed.)
Discrete-time signal processing (2nd ed.)
Introduction to algorithms
Digital Image Processing: A Practical Introduction Using Java (with CD-ROM)
Digital Image Processing: A Practical Introduction Using Java (with CD-ROM)
Digital Image Processing
A discrete model for histogram shaping
Smarandache notions
Beginning GIMP: From Novice to Professional
Beginning GIMP: From Novice to Professional
Contrast enhancement of images using Partitioned Iterated Function Systems
Image and Vision Computing
A histogram smoothing method for digital subtraction radiography
ADVIS'04 Proceedings of the Third international conference on Advances in Information Systems
Hi-index | 0.00 |
Contrast and brightness matching are often required in many medical imaging applications, especially when comparing medical data acquired over different time periods, due to dissimilarities in the acquisition process. Numerous methods have been proposed in this field, ranging from simple correction filters to more complicated recursive techniques. This paper presents a comprehensive comparison of five methods for matching the contrast and brightness of medical image pairs, namely, Contrast Stretching, Ruttimann's Robust Film Correction, Boxcar Filtering, Least-Squares Approximation and Histogram Registration. The five methods were applied to a total of 100 image pairs, divided into five sets, in order to evaluate the performance of the compared methods on images with different levels of contrast, brightness and combinational contrast and brightness variations. Qualitative evaluation was performed by means of visual assessment on the corrected images as well as on digitally subtracted images, in order to estimate the deviations relative to the reference data. Quantitative evaluation was performed by pair-wise statistical evaluation on all image pairs in terms of specific features of merit based on widely used metrics. Following qualitative and quantitative analysis, it was deduced that the Histogram Registration method systematically outperformed the other four methods in comparison in most cases on average.