Research On Nondenstructive Detection Of Apple Mould Core Disease Based On Density Character

Apple’s mould core is a common fruit disease. It leads to a series of problems, such as the fruit abscission during the early stages of development, deterioration in the quality of apples during the mature stage, being decayed in the store process and even infected other apples sometimes. Meanwhile, it will cause serious economic losses. Currently, there are some apples external quality detection grading methods, but there is little research that on the methods of apples’ internal quality detections. Especially for apple’s mould core detection and discriminant are rare. The mildew diseases affect food safety of apple fruit and apple production, and it has an impact on the apple’s market value and hinders the development of China’s agricultural economy seriously.In order to achieve the detection of apple’s mould core quickly and nondestructive, this research machine vision technology in combination with fruit physiological characteristics to propose apple’s mould core detection method based on density character. The main research contents as follows.(1) A quality measuring apparatus controlled by Arduino UNO was used to achieve the accurate determination of apple’s quality. Then quality measuring devices communicate with the PC through the serial port assistant, the state of serial port and the data of quality could be got in MATLAB.(2) The image acquisition device was used to capture apple images from different angles, and the adaptive threshold algorithm(Otsu) and K-means clustering algorithm combining with digital image processing algorithm of machine vision technology in order to achieve the accurate extraction of fruit area;Then the relationship model between the volume calculated by machine vision algorithm and real volume of apple were built in order to achieve the purpose of volume precise calculation of apple. The experimental result showed that by using this method to get the apples of volume error is 3.12%.(3) Explore the change law of apple density was studied and the critical value between mould core apples and normal apples was confirmed. The density of apple was calculated by using the acquired quality and volume values, and the density calculated was utilized to judge whether the apple was mould core apple or not. The experimental result shows that the right identification rate up to 88.46% or more for the apple’s mould core.(4) The MATLAB 7.6 compile GUI toolbox was used to design the apple’s mould core nondestructive detection system, which bases on density character. The system contains modular of building volume relationship model of mould core apples, the quality measure modular, the volume measure modular and density calculation modular. The experiment results show that the system could meet the basic requirement of identification and has the characteristic of interface beauty and operate convenience.