Research On Detecting Technology Of Cotton Bale Quantity In Warehouse Based On GPR

Our cotton reserves are huge,and the production,sale and storage of cotton are closely related to the lives of the people,while cotton is the second largest crop after food and has an important role in the development of the national economy.Due to the high market price of cotton,theft of cotton sometimes occurs during its storage in the warehouse.In recent years,there have been a number of cases of theft of bales from inside the pallets or the replacement of cotton with foreign materials,which have put pressure on the security of the cotton reserves.To this end,the China Cotton Reserve Management Corporation(CRMC)conducts comprehensive inspections twice a year at the country’s cotton storage warehouses to ensure that the quantity of bales stored in the warehouses is safe.Because of the large number and wide distribution of cotton warehouses in the country,manual verification of bale numbers is slow and requires a large number of inspectors.Therefore,the development of a set of effective rapid non-destructive testing methods and information technology to carry out the detection of the number of bales of cotton stored in the warehouse,to ensure the accurate verification of the national stockpile of cotton has far-reaching significance.Based on the above issues,a study on quantity detection techniques for the storage of cotton bales in earth-penetrating radar-based warehouses was carried out.First of all,this paper provides an in-depth understanding of the current status of verification of the number of bales stored in warehouses and the methods used in the field of nondestructive testing.On the basis of a comprehensive analysis of the structure of the pallets and their characteristics,the analysis illustrates the technical advantages of using ground-penetrating radar for palletization,and analyzes and elaborates the main problems to be solved in the processing of the radar detection data of the palletization.Secondly,the working principle of using ground-penetrating radar for pallet detection is introduced,and the data form of ground-penetrating radar echo is described;based on the understanding of the general warehouse storage pallet structure,the pallet structure with two abnormal conditions of cavity and foreign fill inside the pallet is discussed;based on the structure of the pallet and the electromagnetic parameters of the media inside the pallet,the time domain finite difference method is used to simulate the radar echo values of normal and abnormal pallet,and the radar echo characteristics of the two abnormal types are analyzed according to the data results of the simulation,which prepares the algorithm processing study of the radar echo data of the pallet.Next,from the simulation results,it was found that the strong reflected waves generated by the strapping wire of the cotton bales had a large impact on the detection of the anomalous region,and in this paper,the image was pre-processed using an offset algorithm with strong noise resistance.As the medium of the pallet is placed for a long time will be affected by moist air and so on,the magnetic properties of the pallet interior and the pallet surface electricity are different,therefore,this paper proposes an improved imaging technique of the offset algorithm to adjust the wave speed parameter of the offset algorithm according to the change of the electromagnetic properties of the pallet.The use of improved offset algorithm imaging technology can effectively suppress the noise,reduce the cotton bale strapping steel wire on the effect of anomalous spatial interface reflected waves,can better restore the position of the galvanized steel wire inside the pallet,making the detection data imaging more clear,greatly improving the quality of the radar detection echo image of the pallet.Then,this paper proposes a new method for the extraction of the target region of interest combined with the spectral residual method to localize the anomalous region of the pallets.The algorithm analyzes the radar return characteristics of the cavity pallets and the foreign-filled pallets,first adopts the multi-resolution single-performance wavelet extraction algorithm to obtain the multi-scale amplitude component of the radar detection map of the anomalous pallets,selects the component extraction map with obvious amplitude,and then processes it using the spectral residual method to obtain the localization results of the anomalous region.The method is able to quickly extract and locate anomalous areas in the radar images of anomalous pallets without any a priori knowledge,greatly improving the accuracy of the radar detection data decoding of the pallets.Finally,this paper validates the effectiveness of the method of detecting the number of bales stored in a warehouse using real-world data from radar detection of cotton pallets,and integrates the studied algorithm and develops a radar detection data processing software for cotton pallets,reducing the difficulty of using the algorithm.The main innovation points in this paper are the first study on the detection of bale quantity in warehouse storage using ground-penetrating radar;an improved offset algorithm imaging technique is proposed,which effectively handles the strong reflected waves of a large number of strapping wires inside the cotton pallet;a new method for the extraction of the target area of interest combined with spectral residuals is proposed,which achieves the clear localization of anomalous areas of the cotton pallet.