Studies On Detection And Grading Of Apple's Internal Quality Based On Biological Speckle Technique And Instrument Development

Our country was the largest producer and seller of apples in the world.According to the statistical data of the United Nations Food and Agriculture organization,it was mainly low-end market on apple export in our country and the unit price was low.Analysis its reason,it was that Chinese apple post-processing techniques lagged behind.The overall quality of apple was uneven without grading.As a result,apple was unable to enter the high-end market because of the less competitive power of apple in the international market.Currently,nondestructive testing classification research was relatively mature about the external quality of apple.But consumer demand for apple was not only satisfied with the appearance quality,such as size,shape and colour but also paid more attention to the taste and the sweetness of the internal quality of apple.While the nondestructive detection and classification of research was still in a stage about apple's internal quality.Therefore,this thesis was based on biological speckle technique to apple's internal quality grading test research,which improved the comprehensive utilization of apple,and it had the vital significance to the promotion of the competitiveness of the apple.This paper selected apple(c.v.Gala and c.v.Red Fuji)as the research object.It was based on the biological speckle technique to build the speckle measurement device.The apple speckle images and postharvest quality relationship was studied,which established the forecast model of apple postharvest quality indicators.In addition,a set of apple quality nondestructive testing system was developed.1.It was investigated the influence of laser parameters on the speckle image,this paper chose the wavelength of 650nm,20mw power for semiconductor laser as light source.Under the condition of this parameter,the speckle effect was the best.With the parameters of the other components was determined,biological speckle measurement device system was constructed.2.Different image preprocessing method was studied for the speckle image of apple.The gray level transformation,median filtering and threshold segmentation of bimodal method was finally adopted to process speckle image.Outer ring pixel area of S3 was extracted as the characteristic parameters of the speckle image.3.The medium-well apple was select as test object,which grown at commercial farm in Jiangsu.During the storage,speckle images of apple were collected and the quality indexes(hardness,soluble solid content)were determined accordingly.Through the analysis to data pertinence,the forecast model between image parameters and quality indexes of apple was established by using regression analysis.The fit indicators of the model were acceptable and R-squared of each model were above 0.8,which reached the significant level.The results show that the model predicted value was consistent with the actual value and error was smaller by selecting samples to verify this model,4.According to the established forecast model,a set of simple,portable,low cost system to nondestructively detect apple's internal quality was developed,including the design of the apparatus and the development of the software.Based on the biological speckle technique,this paper established the prediction model of the apple quality,and a set of nondestructive testing on apple quality system is developed.Under the situation of in badly need of inspection to frut quality,to improve the quality of fruit has a very important role.To promote high tech of fruit classification in our country and to improve the fruit economic value and competitiveness is of great significance.