Skip to main content
ARS Home » Southeast Area » Dawson, Georgia » National Peanut Research Laboratory » Research » Publications at this Location » Publication #246975
Title: Application of discrete wavelet analysis for moisture content estimation of in-shell nuts nondestructively with a capacitance sensor

Author
item Kandala, Chari
item Sundaram, Jaya
item GOVINDARAJAN, K - University Of Nebraska
item Butts, Christopher - Chris
item SUBBIAH, JEYAM - University Of Nebraska

Submitted to: Proceedings of the American Society of Agricultural and Biological Engineers International (ASABE)
Publication Type: Abstract Only
Publication Acceptance Date: 5/5/2009
Publication Date: 6/11/2009
Citation: Kandala, C., Sundaram, J., Govindarajan, K.N., Butts, C.L., Subbiah, J. 2009. Application of discrete wavelet analysis for moisture content estimation of in-shell nuts nondestructively with a capacitance sensor. Proceedings of the American Society of Agricultural and Biological Engineers International (ASABE).

Interpretive Summary: none required

Technical Abstract: Moisture content is an important quality factors often measured and monitored in the processing and storage of food products such as corn and peanuts. For estimating this parameter for peanuts nondestructively a parallel-plate capacitance sensor was used in conjunction with an impedance analyzer. Impedance and phase angle were measured for the parallel-plate system, holding the in-shell peanut samples between its plates, at frequencies ranging between 0.5MHz and 30 MHz in intervals of 0.5 MHz. The acquired signals were analyzed with discrete wavelet analysis. The signals were decomposed to 4 levels using Daubechies mother wavelet. The decomposition coefficients of the fourth level were passed onto a stepwise variable selection routine to select significant variables. A linear regression was developed using only the significant variables to predict the moisture content of peanut pods (in-shell peanuts) from the impedance measurements. The wavelet analysis yielded similar R2 values with fewer variables as compared to multiple linear and partial least squares regressions. The estimated values were found to be in good agreement with the standard values for the samples tested. Ability to estimate the moisture content in peanuts without shelling them will be of considerable help to the peanut industry.