In this study, coupled mechanical-biospeckle imaging method was performed on potato tubers subjected to creep and uniaxial compression tests. Using this method, we analyzed the characteristics of biospeckle images in relation to the structural changes of potato tissue on loading. During mechanical tests, the speckle pattern images were recorded. The biospeckle activity indices including entropy of the wavelet transform, contrast of successive correlations (CSC), inertia moment, and absolute value of differences were calculated to extract features from the captured videos. The results revealed that the entropy could be an acceptable parameter to assess structural changes and locate rupture point of tuber in the compression test. Analysis of the biospeckle images also showed that the CSC has a decreasing trend during the creep test. Prediction models of mechanical properties were developed by fuzzy c-means clustering algorithm and were optimized using the ANFIS and genetic algorithm. The results can provide the confidence to use this method in drying, frying, and all processes that cannot be taken out samples of it.
