Determination of the wrinkles in persimmon dried by osmosis-microwave combination method using image processing and modeling with artificial intelligence

In this method, image processing (fractal dimension) to predict the amount of wrinkles persimmon mixture osmosis-microwave drying process was used. Image processing software used for image j. Drying combination osmos- microwave persimmon was solutions of sucrose at a concentration of 40, 50 and 60%(w/w),10,20 and 30 minutes and at 30,45 and 60 ° C and the 300and 450 watt microwave. For prediction of wrinkles feedforward artificial neural network MLP with 4 inputs (the process) by neuro solution 6 was used. 9 neurons in the hidden layer network with the lowest mean square error, mean square error and mean absolute error for the predicted normalized fractal dimension of the intersection between the line and the background was dried persimmon (respectively 1/33, 0/49 and 0/002). Also, this model is able to predict the percentage of wrinkles with a correlation coefficient of 0/85 percent.