Dynamic high-resolution images-based monitoring of corn temperature in granary
In order to keep track of temperature of corn stored in granary and provide theoretical reference for subsequent storage conditions,this paper proposed a temperature monitoring method for stored corn with dynamic high-resolution images as the basis.Fundamental data were first collected from corn granary using thermal imaging sensors and then processed and stored in H3C R4900 G3 server.Absolute calibration method was employed to determine interference rate and eliminate interference from sensors before generation of high-resolution images.The images received further geometric correction.To extract differen-tial characteristics,image segmentation was applied to divide images into several regions for extracting textural features and de-termining monitoring scope.After multi-scale segmentation and regional integration,image spots were obtained.Image spot inte-gration conditions were figured out by solving formulas with image spot differential parameter.Finally,dynamic high-resolution monitoring was achieved on temperature change within corn granary under multidimensional features.By computing convolution filtering center and convolution layer characteristics,image spot features were summarized for successfully monitoring tempera-ture change.The experimental results indicated the proposed method could effectively keep track of the temperature within corn granary and provide valuable reference for future warehousing.
万方数据
维普
