Detection Device For Moldy Core In Apples Based On Characteristic Waveband And Apple Diameter

The moldy core in apples is a kind of bacteria invasion of disease, which begins the heart of apples and expands to fruits appearance slowly. Now it has become one of the main diseases to endanger apple internal quality and can not be identified and sorted out through the appearance features.Spectrum technology has been widely used to detect sugar content,acidity, VC content and some internal diseases in many detection systems because of its many virtues such as efficient, fast,sample undestroyed, and online, etc. However, most of these systems the way of spectrumreflection and diffuse transmission, it has a higher accuracy for detecting the content of fruit surface or subcutaneous flesh, but it hard to be applied to detect the content of fruit core. Studies have shown that the way of transmission is superior toreflection to detect the content of fruit core. These systems using spectrumtransmissionto analysis fruit internal quality usually be composed of spectrometer, computer, halogen light source. Although these systems have better accuracy in quality recognition, it still have the disadvantages of high-cost, heavy energy-consuming light source, complex analysis process and redundancy spectral information in wide waveband.So in this paper, a portable and low cost nondestructive detection device of moldy core in apples is designed by using a narrowband light source LED and the photodiode detector.Atest platform was built to obtain the transmission spectrum data by using the tungsten-halogen bulbs, spectrometer and the computer. 700nm-720 nm waveband was confirmed as the characteristic spectrum after using correlation analysis for transmission spectrum data. According to the narrow band of 700nm-720 nm, the LED was chose as light source. the information collection device was designed to obtain the apple diameter and the intensity of transmission spectra. The device was consisted of the core processing unit, the driver circuit of LED, the conversion and detection circuit of photoelectric and the measure module of apple diameter. The MSP430 was used as the core processing unit to call other function modules and process data. The PT4115 was used as the driver chip to control the LED. The linear guide and limit sensor were combined to measure apple diameter online.A function test was done to test the apple diameter measurement accuracy and the acquisition of intensity of transmission when the base station of detection device was finished. The result shows that the mean absolute errorof apple diameter measurement module is 0.56 mm, the maximum absolute error ofintensity of transmission obtain module is 8.33 mV.The 120 Fuji apples produced in Shaanxi Qianxian were selected to do the experiment bythe information collection device and 80% of them were used as the test samples to build discrimination model. The model input was the apples’ diameter(D) and the spectral transmission intensity(V). The model output was the apples’ healthy status(M) and it was recorded when apple was cut in stalk direction, when M is equal to 1, the apple is a healthy apple. Otherwise, when M is equal to-1, it is a sick one. The statistical data show that 24 of samples are moldy core apples and the other 72 are healthy apple. The BP network was used to set up a perfect discriminant model after analyzing the distribution diagram of samples.Finally, the remaining samples for one fifth of total apple samples were used as test sets to estimate the designed detection device with the embedded intelligent discriminant model. The result is that one in six sick apples was detected wrongly and all of 18 healthy apples were detected correctly. Meanwhile the detection time and the rate of non-trouble work of detection device were tested, and the results indicate that the minimumdetection timeis 1.5 seconds and the maximum time is 4 seconds, only onetime breakdownoccursin the continuous316 times work.The study indicates that the designed detection device of moldy core in apples is reliable,high discrimination rate, and has the promising potential for application.