Real-Time Ship Tracking under Challenges of Scale Variation and Different Visibility Weather Conditions

Visual ship tracking provides crucial kinematic traffic information to maritime participants, which helps accurately predict traveling behaviors in the near future. Traditional models obtain a satisfactory performance by exploiting distinct features from images, may fail when scale varies image sequences. Moreover, previous frameworks have not paid much attention weather condition interferences (e.g., visibility). To address this challenge, we propose scale-adaptive framework with help of kernelized correlation filter (KCF) and log-polar transformation operation. First, proposed tracker employs conventional KCF model raw position current image. Second, both step output training sample are transformed into coordinate system, further processed determine factor suppress negative influence conditions. We verify on three typical scenarios under navigational conditions (i.e., sunny, fog). The findings study can participants efficiently situation awareness videos, real time, different visibility