Weight prediction of pork cuts and tissue composition using spectral graph wavelet

Abstract The manual quality assessment of a pork half-carcass using cutting and dissection process is time-consuming, financially expensive, and requires expert butchers. Recently, no technology is available for the automatic quality assessment of pork carcasses to maximize their market value through studying pig feeding cycle, changes in genetic lines, and feeding methods. In this paper, a novel approach for the quality assessment of pork carcasses using 3D shape analysis is proposed. First, a pork half-carcass is precisely modeled by harnessing the power of the spectral graph wavelets, called SpectralWeight. Then, the SpectralWeight is exploited as a predictive model to weigh different cuts and tissue composition of pork. Our study on employing SpectralWeight for weight prediction of a variety of pork cuts including primal and commercial cuts as well as tissue composition of pork demonstrates high accuracy of prediction for most cuts. Although in this work the performance of SpectralWeight for the weight prediction of pork cuts is evaluated, our framework is fairly general and enables new ways to estimate the quality and economical value of carcasses of different animals.