Heterogeneous Pixels Research Articles

The area proportion distribution—relationship with the internal structure of mixels and its application to image classification

An image pixel, generally assumed to be the smallest element of imagery, may actually be a mixture of multiple constituents. To establish a probabilistic model for such a heterogeneous pixel called a mixel, we propose a novel probabilistic model, the area proportion distribution, which is the distribution of the area proportions for a population of mixels. Next we investigate the properties of this model in association with the internal structure of mixels. This model has often been implicitly assumed to be the uniform distribution; however, based on the results of theoretical distributions derived from figure models and of empirical distributions derived from fractal synthetic images, we propose the Beta distribution as the generalized model of the area proportion distribution. We then incorporate this model into the finite mixture density model, and apply it to the classification of satellite images. Finally the performance of our proposed model is shown to be superior using the information criterion. © 2000 Scripta Technica, Syst Comp Jpn, 31(5): 57–76, 2000

Characterisation of ecotones using membership degrees computed with a fuzzy classifier

Ecotones zones lie between homogeneous ecological systems. They are characterised on images by heterogeneous pixels in a specific neighbourhood. Fuzzy classifiers output membership degrees that better represent the heterogeneity within a pixel and can be further processed within the context of a local neighbourhood. This Letter formalises these notations. Two coral reef systems examples are presented. They illustrate the use of possibility measurement to characterise ecotones, and the use of information on well-known ecotones to increase the accuracy of image classification.

A spatialized APAR for heterogeneous pixels

Abstract The importance of APAR models for the study of biomass production from space requires an analysis of their performances on natural mixed pixels. This is the objective of present study which gives quantitative values for the role of heterogeneity on the APAR assessment from space. These results show that the accuracy of the APAR as currently computed from satellite data cannot be considered as acceptable. Wegive a new APAR model adapted to heterogeneous pixels.

Fractal analysis of renal cortical perfusion.

Contrast-enhanced computed tomography (CT) images of canine renal cortex were analyzed to determine if the heterogeneous pixel intensity patterns met the requirements for fractal analysis and if the heterogeneity could be quantified by a fractal dimension, Df. Contrast-enhanced CT images were obtained after injection of iodipamide ethyl ester (IDE), a vascular marker, or iohexol, a freely filtered interstitial marker, into the catheterized renal artery of an anesthetized dog. Images were mounted on a graphics workstation for analysis. A computer program was written to determine the fractal dimension of the pixel-intensity pattern within selected regions. All regions of renal cortex examined met the requirements for fractal analysis. Three seconds after injection of IDE, the mean fractal dimension decreased significantly from 1.21 +/- 0.05 to 1.12 +/- 0.06 (P < .05). Although the mean fractal dimensions were not significantly different, the variation in the fractal dimension around the renal cortex was significantly different with IDE as compared with iohexol at 3 seconds (P < .05). Differences in the change in fractal dimension over time were also observed with IDE as compared with iohexol. Fractal dimension measurement provides a new means to examine the in-vivo organization of renal vascular perfusion by quantifying pixel heterogeneity in contrast-enhanced CT. This may prove useful in understanding and quantifying the pathophysiologic changes in renal disease.