Chinese Journal on Internet of Things ›› 2020, Vol. 4 ›› Issue (4): 51-61.doi: 10.11959/j.issn.2096-3750.2020.00186

• Topic:Smart Agriculture • Previous Articles     Next Articles

Wi-Pest:a method for detecting stored grain pests based on CSI

Shaowei SHAN1,2,Weidong YANG1,2(),Le XIAO1,2,Ke WANG1,2   

  1. 1 College of Information Science and Engineering,Henan University of Technology,Zhengzhou 450066,China
    2 Key Laboratory of Grain Information Processing and Control (Henan University of Technology),Ministry of Education,Zhengzhou 450066,China
  • Revised:2020-03-03 Online:2020-12-30 Published:2020-12-14
  • Supported by:
    The National Key R&D Program of China(2017YFD0401001);The National Natural Science Foundation of China(61741107);The Fundamental Research Funds for the Henan Provincial Colleges and Universities(2018RCJH12);The Open Fund for Henan University of Technology(KFJJ-2016-202)

Abstract:

The environmental and biological factors that affect the food security during the food storage,such as the food temperature,environment humidity,moisture,gas,mildew,pests and others pose a threat to the food storage security,among which the pest is an important factor threatening food storage security.Therefore,a fast and effective detection method is needed to detect stored grain pests.Some of the existing methods are time consuming,using expensive equipment,potentially harmful to health and inefficient.A non-contact,fast and low-cost detection method for stored grain pests based on the amplitude of the channel state information (CSI) was proposed,namely,wireless-pest (Wi-Pest).The feasibility of the pest detection in the stored grain was verified by using CSI amplitude data.On this basis,a Wi-Pest detection method was designed.Firstly,the amplitude data of CSI was preprocessed by outliers removal,data normalization and noise elimination.Then the principal component analysis (PCA) was used to compress the data and extract the main feature components.Finally,random forest (RF) classification method was used to detect stored grain pests.Experiments show that the abnormal density of live pests in grain heaps can be detected under the line of sight (LOS) scenario,and the detection accuracy of the proposed method can reach 97% on average.

Key words: CSI, stored grain pest detection, amplitude, random forest classification

CLC Number: 

No Suggested Reading articles found!