Development Of Multi-Spectral Intelligent Sensor Fof Crop Growth

The precise management of crop production is an important part of precision agriculture, the core thought is positioning control, precision investment with the least amount of production inputs to produce to maximize production efficiency for crop group. Crop access to information is the basis of precise crop management, which plays a decisive role to crop growth, quality, and yield. Based on canopy reflectance spectra, the information non-destructive monitoring techniques of crop growth is a key technology for the current precision management. Sensors as the perception of information equipment is the ways and means of access to the crop canopy spectral information.However, we have not the practical sensing device to access the crop information in real time. So, we urgent need to study smart detection method and carrier to solve the bottleneck of fast non-destructive monitoring of field crops, and to provide technical and equipment support for crop cultivation and management.Based on the physical and chemical mechanism nondestructive monitoring of crop growth information, the multi-specral intelligent sensor was developed, using sunlight as light source with filter spectrophotometry. Field of view of the senson was determined by crop canopy characteristics and photovoltaic systems. Based on crop canopy characteristics and operating environment in the field, the suitable structure parameters of detection lens was designed, streamlining the sensor optical light path, and ensuring that sensitivity and resolution effect of multi-spectral sensor. The sensor can measure weak signals of incident and reflected radiance of vegetation by the eight photodetector with characteristics of special spectral response at the four characteristic bands of560nm,710nm,720nm and810nm respectively. And then the weak spectral information amplifying circuit was built by T-type resistance network to process the weak signals. Based on embedded development technology, the multi-spectral information wsa processed.Static performance parameters experiments of multi-spectral sensors were carried out indoors. Results showed that nonlinearity error were5.72%,3.67%,3.54%and5.51% respectively in560nm,710nm,720nm and810nm, hysteresis error of2.85%,3.15%,3.15%and4.2%. The sensor has the characteristics of good linearity and high accuracy. Through wheat experiments, we established the calibration equation between the multispectral crop intelligent sensor and Fieldspec Pro FR2500spectrometer. The coefficient of determination was0.8028,0.8068,0.8185,0.89, the model in a good correlation. Through experiments in the different plots, the average measurement error of this model is5.6%,4.6%and1.4%,4.5%. Through experiments in the different years, the relative root mean square error (RRMSE) of this model was7.87%,4.76%,5.48%and8.61%.The sensor can realize real-time online tests of rice and wheat canopy reflectance spectra, providing strong support for the development of the crop growth monitoring equipment. The sensor with a portable design, in small size, light weight, can access crop information quickly in real-time.