(3) To evaluate the efficiency of the image classification scheme, once the computational module receives the entire image successfully, the classification will be computed in the cloud
to show the degree of population density of the chickens. Here, the main metric is accuracy measured over ten times for various EVAP farm images, and the average was plotted, including computational time complexity compared to three other well-known image classification techniques, i.e., K – Means clustering, Fuzzy C – Means, and Mean-Shift (Aggarwal and Reddy, 2013). Here, K is two (black and white), and with Fuzzy C-Means and Mean-Shift, maximum
iterations were set up to ten and fourteen, respectively, based on the experiment when the classification results moved toward a steady stage