A Portable Device for Rapid Nondestructive Detection of Fresh Meat Quality

Wan Lin,Yankun Peng
DOI: https://doi.org/10.1117/12.2050142
2014-01-01
Abstract:Quality attributes of fresh meat influence nutritional value and consumers' purchasing power. In order to meet the demand of inspection department for portable device, a rapid and nondestructive detection device for fresh meat quality based on ARM (Advanced RISC Machines) processor and VIS/NIR technology was designed. Working principal, hardware composition, software system and functional test were introduced. Hardware system consisted of ARM processing unit, light source unit, detection probe unit, spectral data acquisition unit, LCD (Liquid Crystal Display) touch screen display unit, power unit and the cooling unit. Linux operating system and quality parameters acquisition processing application were designed. This system has realized collecting spectral signal, storing, displaying and processing as integration with the weight of 3.5 kg. 40 pieces of beef were used in experiment to validate the stability and reliability. The results indicated that prediction model developed using PLSR method using SNV as pre-processing method had good performance, with the correlation coefficient of 0.90 and root mean square error of 1.56 for validation set for L*, 0.95 and 1.74 for a*, 0.94 and 0.59 for b*, 0.88 and 0.13 for pH, 0.79 and 12.46 for tenderness, 0.89 and 0.91 for water content, respectively. The experimental result shows that this device can be a useful tool for detecting quality of meat.
What problem does this paper attempt to address?