Abstract:Remote Sensing plays a fundamental role in acquiring crucial information about the Earth's surface from a distance, especially through fully polarimetric data, which offers a rich source of information for diverse applications. However, extracting meaningful insights from this intricate data necessitates sophisticated techniques. In addressing this challenge, one predominant trend that has emerged is known as target decomposition techniques. These techniques can be broadly classified into coherent and non-coherent methods. Each of these methods provides high-quality information using different procedures. In this context, this paper introduces innovative feature fusion techniques, amalgamating coherent and non-coherent information. While coherent techniques excel in detailed exploration and specific feature extraction, non-coherent methods offer a broader perspective. Our feature fusion techniques aim to harness the strengths of both approaches, providing a comprehensive and high-quality fusion of information. In the first approach, features derived from Pauli coherent decomposition, Freeman–Durden non-coherent technique, and the Symmetry criterion from Cameron's stepwise algorithm are combined to construct a sophisticated feature vector. This fusion is achieved using the well-established Fisher Linear Discriminant Analysis algorithm. In the second approach, the Symmetry criterion serves as the basis for fusing coherent and non-coherent coefficients, resulting in the creation of a new feature vector. Both approaches aim to exploit information simultaneously extracted from coherent and non-coherent methods in feature extraction from Remote Sensing data through fusion at the feature level. To evaluate the effectiveness of the feature generated by the proposed fusion techniques, we employ a land cover classification procedure. This involves utilizing a basic classifier, achieving overall accuracies of approximately 82% and 86% for each of the two proposed techniques. Furthermore, the accuracy in individual classes surpasses 92%. The evaluation aims to gauge the effectiveness of the fusion methods in enhancing feature extraction from fully polarimetric data and opens avenues for further exploration in the integration of coherent and non-coherent features for remote sensing applications.

Combining the FCM Classifier with Various Kernels to Handle Non-linearity of Class Boundaries

Semi-supervised kernel FCM algorithm for remote sensing image classification

Robust Semi-supervised Kernel-FCM Algorithm Incorporating Local Spatial Information for Remote Sensing Image Classification

Hybrid fuzzy multiple SVM classifier through feature fusion based on convolution neural networks and its practical applications

Fusing Deep Features by Kernel Collaborative Representation for Remote Sensing Scene Classification

Learning to Combine Kernels for Object Categorization

Kernel collaborative face recognition

Multi-kernel Collaboration-Induced Fuzzy Local Information C-Means Algorithm for Image Segmentation.

Combinational approach of object oriented land cover image classification

Kernel generalized fuzzy c-means clustering with spatial information for image segmentation

Higher-level Feature Combination Via Multiple Kernel Learning for Image Classification

Optimal Double-Kernel Combination for Classification

Reducing Boundary Effects in a Kernel-Based Classifier

Object-oriented Land Cover Classification of HJ-1B CCD Image Through Multiple Classifier Fusion.

Algorithm of Image Segmentation Based on KFCM Using Hybrid Kernels with Pairwise Constraints

Fusion of Coherent and Non-Coherent Pol-SAR Features for Land Cover Classification

A Kernel Elm Classifier For High-Resolution Remotely Sensed Imagery Based On Multiple Features

Object-based image retrieval with kernel on adjacency matrix and local combined features

Type-2 Kernelized Fuzzy C-Means Algorithm Based on the Uncertain Width of Gaussian Kernel with Applications in Mr Image Segmentation

Kernel-based Fuzzy-Rough Nearest Neighbour Classification.

Object Classification Via Feature Fusion Based Marginalized Kernels