Research And Development Of Integrative Portable Testing Device For Multi-quality Parameters In Apple Based On Visible/near Infrared Spectroscopy

At present,consumers’demand for apples has gradually changed from“quantity type”to“quality type”.People pay more attention to the internal quality of apples.The indicators for measuring the internal quality of apples mainly include sugar value,acidity value and hardness value.Near-infrared spectroscopy has the characteristics of non-destruction,easy operation and high efficiency.It has been widely used in the field of fruit quality.However,most of the near-infrared spectroscopy techniques use wide-band detection methods,which have high application cost,high energy consumption and redundant spectral information.And the complexity of the analysis process.In this paper,a combination of photodiode and characteristic narrow-band LED is proposed,and an apple multi-quality integrated non-destructive testing device based on visible/near-infrared spectroscopy is developed.Achieve rapid non-destructive testing of apple’s internal quality.It is of great significance to guide apple harvesting,delivery time and sales price,and it is helpful to promote the healthy development of the apple industry.The research contents and conclusions of this paper are as follows.(1)Construct a diffuse reflection detection platform and extract characteristic wavelengths.Using a tungsten halogen lamp,a spectrometer and a computer to form an internal diffuse reflectance spectrum acquisition platform for apples,the diffuse reflectance spectrum information of the sample was collected,and the sugar content and acidity of the apple were measured by physical and chemical properties,and the hardness was measured by a texture analyzer.Firstly,Monte carlo sampling method is used to detect abnormal data and cull.Then,using competitive adaptive weighting algorithm,moving window algorithm,monroco no information elimination algorithm and random leapfrog algorithm,three indicators of internal quality of apple are extracted respectively.The characteristic wavelength of the value,the acidity value,and the hardness value is finally optimized to the 10 characteristic wavelengths most relevant to the three indexes,namely 420,445,480,510,600,690,730,900,970,and 1035 nm,respectively.(2)Hardware design of detection equipment based on the principle of diffuse reflectance spectroscopy.Using 10 features narrow-band LED and photoelectric sensor combined detection method,with Raspberry Pi 3b+as the core processor,designed diffuse reflection detection optical path,narrow-band LED ring light source,detection probe,light source drive,spectrum acquisition,device shape,user Hardware systems such as interactive,QR code storage and printing functions,control circuits,etc.(3)Software design of detection equipment based on Linux system,design and development control procedures,QR code storage of quality information,cloud Platform records and human-computer interaction interface.According to the development requirements of embedded software system,the software and hardware development platform based on Linux operating system is constructed,the apple quality information based on 16 encoding is realized,the quality information is transmitted based on TTL level,the quality information cloud platform record based on HTTP protocol,and the user interaction interface based on QT design,then complete the software design of the system.(4)The internal quality prediction model of apple was constructed,and then the embedded internal quality model of apple was validated and tested.Using the spectral information collected by the equipment and the quality information measured,an internal quality prediction model of apple based on MLR was constructed and transplanted to raspberry pie.Finally,the test accuracy and stability of the equipment are verified and tested.The results showed that the determinant coefficients of three internal quality indexes were 0.8085,0.7592 and 0.7651,the root mean square errors were 0.5875~°Brix,0.0389%and 0.4125 kg/cm~2,and the maximum variation coefficients were 0.0112,0.0115 and0.0061,respectively,which could meet the normal detection requirements and the stability of the equipment was good.Device power test shows that it can work continuously for 7.5hours at maximum power consumption to meet the normal dosage requirements of the user.