Research On Sea Ice Concentration Detection Based On Radar Video Accumulation

Sea ice concentration detection has always been the focus of marine disaster monitoring.New and higher requirements are put forward for sea ice concentration detection with the increasing prosperity of marine transportation industry and the exploitation of global marine oil and gas resources potential.In order to improve the accuracy of sea ice concentration detection and better protect the safety of marine transportation and production,this thesis proposes a method of sea ice concentration detection based on radar video accumulation,which improves the detection accuracy of sea ice concentration in real time.The traditional visual and instrumental inspection method not only consume a lot of human resources,but also have low inspection efficiency.In response to these shortcomings,remote sensing measurement technology is gradually developed in sea ice concentration detection.Compared with satellite remote sensing and aerial remote sensing,remote sensing technology using X-band marine radar sensor has the advantages such as real-time detection.This study retains the sensor unit of electromagnetic wave receiving and transmitting part of marine radar and innovatively designs the radar signal processing unit.This study leverages the target detection capability of marine radar to complete a task on real-time and refined detection of sea ice concentration by ice detection radar system.By analyzing the reflection characteristics of sea ice,this thesis optimizes the radar beam model with half-power point horizontal beam width as absolute boundary from three aspects:radar erection height,earth curvature and beam power.The optimized model uses information contained in the antenna pattern more accurately to improve the angular detection accuracy of radar beam.The angular error compensation correction algorithm proposed in this thesis can better eliminate the angular spreading of radar images to obtain the inverse sea ice concentration.In this thesis,wind and ice flow field data is processed by bilinear interpolation method.Additionally,this thesis uses linear fitting method to obtain two main forces of wind and current that affect the sea ice motion velocity,and applies Video Accumulation Algorithm to the sea ice image processing for the first time based on this outcome.Based on the weak echo characteristics of sea ice,the Mean Filtering Algorithm in spatial domain cannot effectively remove the noise signals from sea ice radar images.This thesis combines the Video Accumulation Algorithm in time domain with the Mean Filtering Algorithm in spatial domain and proposes the Mean Filtering based Video Accumulation Algorithm(MFVA).After comparison experiments,the improved algorithm has better denoising effect and provides more accurate data for the calculation of sea ice concentration.Finally,this thesis conducts experiments relies on the vehicle-mounted mobile radar monitoring station.Sea ice radar echo images in Liaodong Bay are collected,analyzed and processed.The ice detection radar system built in this thesis is a breakthrough in the field of sea ice concentration after the special radars such as wave radar,oil spill radar and bathymetric radar.The ice detection radar system described in this thesis only retains the original scanning transceiver sensor unit of the marine radar,while the signal processing unit is innovatively designed to complete the detection of sea ice concentration in a more continuous and accurate way.Comparing calculation accuracy of the algorithm proposed and other existing algorithms,the calculation error of the proposed algorithm is within a reasonable range,and the accuracy of sea ice concentration detection meets the actual engineering requirements.