User talk:83.85.11.84

Regarding the deleted first statement "Histogram equalization is one of the best methods for image enhancements.": This is Trump-like rhetoric. There are several 10k scientific publications over a time span of about 50-60 years on image enhancement. It well known that there is no agreed definition of image quality, rather the notion of quality depends on the image analysis or further use of the image. Also note that quality apart from technical measures also covers perceptual measures ("beautiful"). I therefore hold that "best method" is a meaningless statement, as the criterion for comparison is not established (and as far as is accepted in the field of image analysis and computer vision also never well be established). The statement "It provides better quality of images without loss of any information." is unclear: better than what? Than any other method? Regarding what criterion, see above? The statement "without loss of any information" OBVIOUSLY is wrong: as can be seen already from the two histograms before and after histogram equalization, histogram equalization changes pixel-values that were originally different to values that are the same. So pixels that were originally distinguishable by their pixel values after equalization are no longer distinguishable. So evidently there is a loss in information. The cited article is not of the right calibre to be cited on wikipedia. It presently has 7(!) citations according to scholar.google.com. Note that well known articles in image processing, e.g. Otsu method for thresholding have several 10k citations. Already from its abstract it is apparent that it merely concerns some selected algorithms (of which the authors know) with FPGA implementations. (I also note that the statements made in the article, in the generality in which they are made, are mostly false!) If no better article can be found better refer to a chapter of a book on general image processing, e.g. the one from R.C. Gonzalez and Woods.