Wavelet textural features from ultrasonic elastograms for meat quality prediction

Ultrasonic elastography was used for quantitative imaging of strain and elastic modulus distribution in meat samples. This article proposes the application of wavelet wavelet analysis for textural feature extraction from these images. In this article, the wavelet analysis is briefly introduced, the application of elastography for imaging meat muscles is reviewed, and a procedure of generating a wavelet transform of meat elastogram for the extraction of textural features is suggested. Wavelet analysis is then applied on beef samples for elastogram feature extraction compared to prior work based on Haralick's method. For beef tenderness prediction wavelet features produced significantly higher R-2 values (0.7-0.9) in linear statistical models than Haralick's features (0.1-0.8).