柑橘成熟度机器视觉无损检测方法研究

水果内部品质无损检测技术是确定水果最佳采收期和按成熟度进行准确分级的关键。本研究以表面色泽与固酸比为柑橘成熟度指标，建立了用于柑橘成熟度检测的机器视觉系统，确定了适宜的背景颜色，进行了柑橘的分光反射试验，发现绿色柑橘表面与桔黄色表面的反射率在700 nm时反射率相差最大，约达53%，且各自的反射率都较大，700 nm是获得高质量的柑橘图像的较佳中心波长。建立了利用协方差矩阵和样本属于桔黄色和绿色的概率来判断柑橘成熟度的判别分析法，并以实测的固酸比值作为对照，对72枚柑橘样本进行了试验，柑橘成熟度的判别准确率达到91.67%。这表明柑橘果实的表皮颜色与成熟度之间具有相关性。

Methodology for nondestructive inspection of citrus maturity with machine vision

Non-destructive maturity evaluation of fruits is the key to harvesting and sorting of fruits according to their maturity. In order to explore a methodology for the maturity inspection of citrus with machine vision technology, the surface color information and the ratio of total soluble solid to titratable acid (TSS/TA) were used as maturity indexes of citrus. The spectral reflectance characteristics of different color citrus were determined by use of a UV-240 ultraviolet and visible spectrophotometer. The results stated that at the wavelength of 700 nm, the green surface and saffron surface of citrus were of higher spectral reflection, the difference between them reached the maximum, about 53%, and the image acquired at this wavelength could be of much color information for the maturity inspection. At last, the maturity of 72 citrus was evaluated by the discriminate analysis and covariance matrix. The test results showed that the identification accuracy was 91.67%. It was concluded that the maturity of citrus was related to their surface color information, and it is feasible to nondestructively evaluate the maturity of citrus with the machine vision system.