Multiscale color invariants based on the human visual system

Abstract

This paper proposes a new representation for color texture using a set of multiscale illuminant invariant features. The approach was specifically developed to investigate the feasibility of using machine vision to automatically monitor populations of animal species in ecologically sensitive regions, such as the Amazon Forest. The approach uses a combination of Finlayson's (1994) color angle idea and Gabor multichannel filters and was inspired by the multichannel model of the human visual system (HVS). Using a database of color textures from three species of Amazonian monkey, and also a previously published reference database of color regions, we show that the approach performs better than methods based on color angles or Gabor filters alone. The Monkey database was compiled from texture segments extracted from a video of the Amazon Forest using a spatio-temporal segmentation algorithm. The approach is evaluated by applying two different classification tests in order to measure the quality of the recognition features root mean square (RMS) analysis and receiver operating characteristic (ROC) analysis.

Texture-based computer-assisted diagnosis for fiberscopic images

“…Wanderley et al [10] published an illuminationinvariant version of color Gabor filters which compute features appropriate to the current scale. Together with this algorithm we used four different feature sets for the experimental evaluation.…”

“…All of these methods rely more or less on the extraction of textural features such as co-occurrence matrices [6], sumand difference-histograms (SDH) [7], wavelet features [8], local binary patterns [9], or Gabor filters [10]. However, to our knowledge no one ever tried to apply such algorithms to fiberscopic images which are very difficult to analyse due to the unavoidable artifacts imposed by the construction.…”

A Robust Digital Watermarking Scheme Based on Bandelet and Image Fusion

“…By taking use of frequency contrast sensitivity matrix, visual sensitivity of an image block can be calculated for embedding watermarks and then decomposition principle of sub-tree gets clear. Specifically, it can be computed as [14]: (2) is Fourier transform of an image block and s reflects visual sensitivity of human eyes to this image block. The larger the value of s is, the less sensitive human eyes are to this image block.…”

Machine learning and multiscale methods in the identification of bivalve larvae

“…The use of colorangle based features in multiscale representation of natural images has been further explored in [21]. The use of HSV (Hue, Saturation, and Value) space to split an image into chromatic and achromatic channels is explored in [9].…”