Semantic Segmentation of Radar Data with Deep Learning
MetadataShow full item record
Autonomous ships have the potential to redefine the maritime industry, providing substantialimprovements in safety, economics and fuel efficiency, while creating new previouslyunimagined services. Radar is a key part of the solution to the navigation andcollision avoidance challenge for autonomous vessels, due to its high range and accuracy,as well as robustness against harsh weather conditions. A semantic segmentationsimplifies the radar data into homogeneous clusters of defined targets of interest, allowingthe autonomous agent to understand its environment at a resolution suitablefor high-level reasoning and planning. A large database of maritime radar envelope data was gathered from two S-bandradar systems along the Norwegian coastline, and ground truth labels were constructedfrom Automatic Identification System (AIS) data and chart data. A Machine Learning(ML) detector, in the form of a Convolutional Neural Network (CNN), was trained in asupervised manner to segment the radar envelope image into four classes: background(sea), vessels, land and unknown (i.e. regions that the radar s signals can not reach).Several modifications to CNN architectures and augmentations to radar envelope dataare suggested and evaluated, with downsampling of the envelope data increasing thereceptive field of the CNN units and replacing batch normalization layers with groupnormalization layers, alleviating the adverse effects of limited memory, having the mostsignificant contributions. The ML detector exhibits great results on the test set: correctly classifying over 99%of the navigable area, an overall 69.11 mean Intersection Over Union (IoU) score anda mean of 75.55 on the Boundary Jaccard (BJ) index. The vessel detection capability iscompared to a traditional approach to radar detection in the form of a Constant FalseAlarm Rate (CFAR) detector supplied by Kongsberg Seatex. The ML detector achieves ahigher probability of detection for all false alarm rates, with both detectors managinga 0.9 probability of detection for a false alarm rate around 2 · 10−2. Below this point,the ML detector outperforms the CFAR detector by several orders of magnitude onthe false alarm rate. The ML detector also exhibits superior streak performance; theaverage and maximum period that vessels go undetected is significantly shorter, especiallyat lower false alarm rates. This performance discrepancy can in part be explainedby lower-than-expected performance from a CFAR detector, owing to a difficult, multimodaldata distribution and issues with the vessel labels. The master project and all experiments are conducted on regular consumer hardware.The ML detector processes data at a fraction of the data collection rate of radarsystems (0.3 vs. 2.4 seconds). The developed detector system could therefore be deployedas is, but more research on the effects of adversarial techniques should be conductedbefore it is employed for safety-critical tasks.