Commonly available software does not seem to properly detect edges in transition area under strong shadows. This blow explores alternative clustering algorithms and other methods to improve edge detection on a field of items under strong shadows. A combination of open source Linux packages will be used to generate statistical animations of pixel classification under adverse lighting conditions. Predictive and/or Bayesian Markov chains will be examined for their possible value in this process.
walkandfish 100000K4VA Identificações:  opencv ggplot2 detection gnuplot edge clustering r c++ 844 Visitas