Image Processing Algorithms Visualiser

Abstract

A highly visual way to understand and develop computer vision algorithms is proposed. Image processing realization commonly sustains of successive function calls over the output image from the previous function. In the later phases (i.e. contour analysis, operations over segments) when the output argument of a function is not always an image, there is often a way to visualize the arguments by an image. The developed program described in this article is capable of executing basic computer vision algorithms over the input image sequentially and visualize individual steps outputs as image queue. The initialization of functions chain is possible with simple text input of individual function names with an easy option to program new algorithms.