Identification of material surfaces using grey level co-occurrence matrix and Elman neural network

Material type absorption coefficient is one of the important parameter that used for acoustic room calculation. Currently, absorption coefficient is obtained by using impedance tube or resonance tube. Both techniques need long learning good skills, high cost equipment, and time consuming to conduct. This paper proposed a system distinguished absorption coefficient thru the material surface identification from digital images. The system was built by applying Grey Level Co-occurrence Matrices (GLCM) and Elman Neural Network (ENN). Result for the best mean squared error (MSE) was 4.62e-9 for training phase and 0.5084 for testing phase. Overall, the system is able to identify the material surfaces and thus directly obtain the absorption coefficient of the material without using any physical equipment as oppose to the current techniques.