跳到主要內容區

Using the Multispectral imaging for Predicting the Water Status in 2 Mushroom and Moisture Content and Water Holding Capacity in 3 Cooked Pork Sausages

學生姓名: 廖頌賢
指導教授: 蕭心怡
學期: 110上
摘  要: Moisture as general component of food, it can maintain the texture and appearance of food, but it will also become a condition for the growth of microorganisms. Therefore, moisture form and content are used to the detection indicators of factories. In traditional
technology, drying method and LF-NMR are mostly used. But these methods are very time-consuming, so the purpose of this research is to use multispectral imaging technology (wavelength range from 405 to 970nm) to detect moisture content, status and
water holding capacity. The results show that the best predictive model for detecting the moisture form of mushrooms during drying is BPNN, with R2 20 P of 0.9369 and RMSEP of 0.2291%. In addition, for cooked pork sausages, the best models for monitoring moisture content and water retention are PLSR combining spectrum and texture data of PC1, and PLSR combining spectrum and texture data of PC2, respectively. Its rp is 0.948 and 0.832. In conclusion, The prediction model established by the use of multispectral image technology can accurately predict the moisture content, status and water holding capacity.
瀏覽數: