Nondestructive Detection of Internal Deterioration of Mango Based on Near Infrared Spectroscopy

Long Li,Yongyu Li,Yankun Peng

DOI: https://doi.org/10.13031/aim.201800106

2018-01-01

Abstract:The internal deterioration of mango is an essential factor affecting its quality. A new method for detecting the internal quality of mango is necessary. To achieve rapid, accurate and nondestructive detection of mango internal deterioration, a visible/near infrared spectrum detection system based on the diffuse transmission principle has been set up. This system utilizes the ocean optical USB2000+ spectrometer to collect the mango diffuse spectrum in the range of 350 - 1000 nm. The experimental samples are Kate mango 80, picked in Panzhihua. The collected spectra are processed by the S-G polynomial smoothing. In the first case of deterioration, the internal deterioration of mango is in the later stage, the deterioration area is larger, the blackening is severe, the flesh tissue is spongy, 765nm and 815nm are the characteristic spectra of the first case, and the slope of two spectral points is utilized as the characteristic information of the first case deterioration detection. For the deterioration second situation, mango internal deterioration also reaches the later stage, and the deterioration area is relatively large. Nonetheless, different from the first situation, the color of the deterioration flesh does not turn black, the flesh color is lighter than the normal fruit, the moisture content is higher than that of the normal fruit, 815nm is the characteristic spectrum of the second case, and the transmittance spectral intensity at 815nm is utilized as the characteristic information to detect the second deterioration situation. As for the third situation of deterioration, internal deterioration is in the early stage, the deterioration area is small, and the color of the flesh is similar to the normal flesh. Nonetheless, with the change of the internal components of mango, the absorption peak of deteriorated flesh is mainly concentrated in 860 - 870nm. The ratio of the mean transmittance (mean) and the peak intensity of 815 nm in the wavelength range of 860 - 870nm are utilized to detect the third deterioration situation. Based on the three different cases of deterioration, a unified prediction model is established with three pieces of feature information. The results show that the correct rate of mango is 95%. The research result provides a new method for the detection of the internal deterioration of mango, as well as offers a theoretical basis for the development of the internal deterioration detection device of visible near infrared mango.