The Research Of Technique In Non-destructive Detection By On-line VIS/NIR Spectroscopy

China is a big fruit producing country since ancient times; the fruit make a veryimportant rule in international fruit market. Recently years, the total number of fruitexport increased year by year, but less than5%of the world total output. The cause is theadded value of our China’s post-harvest fruit is very low, the low level ofcommercialization of our fruit industry, lead to the lack of market competitiveness. As akey process of fruit commercialization, the online non-destructive detection of fruitexternal and internal quality is very important. The development and application ofonline non-destructive detection equipment not only can promote the fruit industry inChina, and increase the farmer’s income, improve people’s lives level, also is a win-winpolicy, open the high quality fruit market, enhance our fruit industry competitiveness inthe international market.Near infrared (NIR) spectrum detection technology have been widely applied tonondestructive detection in various agriculture products, the NIR detection technologyhave made a primary development on basis research in recently years, also do somepositive research in fruit internal quality, but the online non-destructive detection devicebased on NIR spectrum technology still in the preliminary stage, especially theequipment applied to industrialization.This paper used thick peel fruit (‘Gannan’ navel orange) and thin peel fruit(‘Cuiguan’ pear and ‘Nanshui’ pear) for research object, based on NIR spectrumnon-destructive technology,and combined with mechanical and electrical, computertechnology. Designed an online equipment to detect fruit sugar content of fruit,completed the software control system, and made different online detection experiments,established the online non-destructive detection prediction model.Established a best ‘Gannan’ navel orange prediction model with r_c~2=0.9046,RMSEC=0.38%, r_p~2=0.8649, RMSEP=0.45%.Established a best Cuiguan’ pear prediction model with r_c~2=0.9364, RMSEC=0.29%,r_p~2=0.8782, RMSEP=0.44%.Established a best ‘Nanshui’ pear prediction model with r_c~2=0.9359,RMSEC=0.32%, r_p~2=0.8634, RMSEP=0.42%.The result show that it is feasible to use NIR spectrum technology to detect fruitsugar content online, and offer some theory foundation and references for onlinepractices.