Ocean Eddy Recognition in SAR Images With Adaptive Weighted Feature Fusion

Abstract

Automatic recognition of ocean eddies has become one of the hotspots in the field of physical oceanography. Traditional methods based on either physical parameters or geometry features require manual parameter adjustment, and cannot adapt to the dynamic changes of ocean eddies caused by complicated ocean environments. To address these issues, we propose a new eddy recognition method in SAR images with adaptive weighted multi-feature fusion. Specially, to better characterize eddies, we first extract texture, shape and corner features using global Gray Level Co-occurrence Matrix (GLCM), detailed Fourier Descriptor (FD) and local salient Harris features respectively. Secondly, considering the different importance of features for eddy recognition, we propose an adaptive weighted feature fusion method with multiple kernel learning (MKL). Here, a combined kernel is derived to fuse three selected kernels for the three types of features with the weights trained by MKL. Finally, we design a SVM classifier with the combined kernel to realize the eddy recognition. The experimental results show that: 1) our proposed method can reach 93.42% of eddy recognition accuracy, which is much higher than the methods with only one single feature; 2) adaptive weighted fusion plays an important role in improving the accuracy. Our proposed method with MKL gains a 8.36% accuracy increase than the method without MKL. Through adaptive weighted fusion, our method avoids the manual parameter adjustment and is more robust and general. Experimental results have proven that our method is effective and applicable to recognize eddies.INDEX TERMS Multi-feature fusion, adaptive weighted fusion, multiple kernel learning, ocean eddies, image recognition, SAR images.

Directional Wave Scattering Distribution Modes Analysis and Synthesis of Random Ocean Media Roughness for SAR Electromagnetic Interactions Using Feature Fusion in Dynamic Sea States: A Survey

“…This MTD fusion scheme can refine texture and preserve wave pattern, addressing challenges in multi-mode ROM fusion [150]. The upcoming section describes the proposed CNN roughness reconstruction [129], the LGE strategy [126], pyramidal fusion [127], and implementation scenario [150]. However, design details exceed the scope of this survey, and further instructions on architecture and training scenarios are available in [125], [150].…”

“…Given the dependency of directional WFS formulation on both spreading function and its boundary conditions, which includes the effects of sea states [84], [108], [112], no ROM model generated thus far outperforms others in terms of representing reference roughness pattern and texture characteristics [45], [49], [56], [63], [65], [77], [88], [89], [93], [99], [103], [108], [113], [117], [120]. To address this limitation, a multi-scale transform domain (MTD) fusion method is proposed [124]- [127], incorporating a convolutional neural network (CNN) [128], [129], enabling the reconstruction of a fused roughness model independent of spreading functions as a reference ROM [50], [80], [103], [129], and facilitating the generation of an optimized SAR raw data [59], [60], [73], [74], [77], [80], [83]- [87], [93], [106], [107], [119], [121].…”

Submesoscale oceanic eddy detection in SAR images using context and edge association network

“…The polarization modes of IW data are VH and VV. According to previous research (Du et al, 2019a;Du et al, 2019b;Yan et al, 2019;Zhang et al, 2020), the VV polarization image is used for oceanic eddies recognition. Then, some indispensable preprocessing is performed on raw data, such as the application of orbit file, noise removal, terrain correction, and dB transformation.…”

“…However, numerous studies confirm the location, diameter, vorticity, and manifestation types of oceanic eddies by visually inspecting at full resolution, and further analyzing them from the oceanographic point of view (Munk et al, 2000;Kozlov et al, 2019;Stuhlmacher and Gade, 2020;Ji et al, 2021). Apart from manual interpretation, the algorithms of eddy detection are mainly divided into image transformation-based (Karimova, 2017), handcrafted feature-based (Chen et al, 2019;Du et al, 2019a), and machine learning-based (Du et al, 2019a). Karimova (Karimova, 2017) proposed an algorithm incorporating various image transformations to detect black eddies, which are reflected by surfactant films.…”

“…Karimova (Karimova, 2017) proposed an algorithm incorporating various image transformations to detect black eddies, which are reflected by surfactant films. Du et al (Du et al, 2019a) designed an adaptive weighted multi-feature fusion algorithm for automatic mesoscale ocean eddies recognition. The recognition method can meet the requirements of marine complex and changeable environment, with the optimal accuracy reaching 93.42%.…”