.png){kind=icon}
.png){kind=icon}
Return
Marine radar-based oil spill detection method utilizing particle swarm optimization algorithm
.png){kind=icon}
PRE
AI
DOI：10.1080/01431161.2025.2567020.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
As maritime transportation operations become increasingly frequent, the associated risks of oil spills have risen substantially. Consequently, offshore oil spill detection technology stands as a fundamental pillar of disaster response and environmental preservation initiatives. A marine radar-based approach was introduced to detect oil spills, which employs an enhanced particle swarm optimization (PSO) algorithm. First, the row vector and grey threshold were used to extract co-frequency interferences in Cartesian coordinate system. Then, a mean filter was used to suppress co-frequency interferences. After that, a dual-threshold and a median filter were selected to remove speckle noise. Fourth, a greyscale correction matrix was introduced to adjust the grey distribution of the denoised image. Afterwards, the Contrast Limited Adaptive Histogram Equalization (CLAHE) method was used to strengthen the features of oil film targets. Subsequently, the PSO method was improved to calculate the oil film segmentation threshold with the Quasi Newton, the Levenberg Marquardt, the Trust Region Dogleg and the Trust Region Reflection. The experimental results show that oil film targets can be effectively extracted by all four improved PSO algorithms. For complex data, a 31% speed improvement is achieved by the Trust Region Dogleg algorithm compared to other methods. It was more accurate than the other three improvement methods. This method will provide scientific basis and technical assistance for responding to offshore oil spills.
Keywords:
Marine radar
oil spill
particle swarm optimization
Journal
IF:
2.6
Papers:
1.2W
Citations:
2.7W