Knowledge-Based Morphological Deep Transparent Neural Networks for Remote Sensing Image Classification

Abstract

Land use/land cover classification of remote sensing images provide information to take efficient decisions related to resource monitoring. There exists several algorithms for remote sensing image classification. In the recent years, Deep learning models like convolution neural networks (CNNs) are widely used for remote sensing image classification. The learning and generalization ability of CNN, results in better performance in comparison with similar type of models. The functional behavior of CNNs is unexplainable because of its multiple layers of convolution and pooling operations. This results in black box characteristics of CNNs. Motivated with this factor, a CNN model with functional transparency is proposed in the present study. The model is named as Knowledge Based Morphological Deep Transparent Neural Networks (KBMDTNN) for remote sensing image classification. The architecture of KBMDTNN model provides functional transparency due to application of morphological operators, convolutional and pooling layers, and transparent neural network. In KBMDTNN model, the morphological operator preserve the shape/size information of the objects through efficient image segmentation. Convolution and pooling layers are used to produce minimal number of features from the image. The operational transparency of proposed model is coined based on the mathematical understanding of each layer in the model instead of randomly adding layers to the architecture of model. The transparency of proposed model is also because of assigning the initial weights of NN in output layer of model with computed values instead of random values. The proposed KBMDTNN model outperformed similar type of models as tested with multispectral and hyperspectral remote sensing images. The performance of KBMDTNN model is evaluated with the metrics like overall accuracy (OA), overall accuracy standard deviation (OA STD ), producer's accuracy (PA), user's accuracy (UA), dispersion score (DS), and kappa coefficient (KC).

Study on Remote Sensing Image Classification of Oasis Area Based on ENVI Deep Learning

“…Accurately extracting and identifying information of different feature types are of great significance for digital monitoring and efficient and rational utilization of soil and water resources in the region. Convolutional neural networks (CNNs), the most representative deep learning algorithm, are widely used in research on remote sensing image classification [18][19][20][21]. Although CNN can adaptively extract the most relevant features to the classification [24][25][26][30][31][32], which has also continuously enriched and improved the methods and theories of remote sensing image classification research and accumulated many worthy results and experiences [33][34].…”

“…Deep learning has emerged as a new research direction in the field of machine learning as a result of the continuous advancement of artificial intelligence algorithms, providing novel ideas for the extraction and identification of remote sensing images [16][17]. Convolutional neural network (CNN), as one of the most representative algorithms of deep learning, has been widely used in remote sensing image classification research, and has accumulated many worthwhile results and experiences for remote sensing image classification research [18][19][20][21]. Pan [22] noted that CNNs can extract higher-level spatial features from images in a hierarchical manner, providing more powerful recognition capabilities for target detection and scene classification in high-resolution remote sensing images; Li [23] demonstrated that by using an improved CNN with an overlap pooling method for remote sensing image classification, image details can be effectively improved and obtain high classification accuracy.…”

Adaptive Granulation-Based Convolutional Neural Networks With Single Pass Learning for Remote Sensing Image Classification

“…The model is computationally faster than the conventional graph-based CNN and it can classify unknown samples without retraining the network and improve the classification performance. Kumar [11] proposed a knowledge encoded CNN model for multispectral remote sensing image classification. Morphological operators were used in the convolutional layers of this model to obtain informative features of the objects.…”

Calculating the Interspersion and Juxtaposition Rates for Saint Clair Flats State Wildlife Area

“…In the case of this project, in order to assess the management being done, classification maps are needed. Kumar (2022) recently has done an evaluation of remote sensing classifications, a major conclusion being that image classifications (specifically from remote images) are of great use when evaluating and monitoring resources. Kumar went further to suggest a process that can be done to an image before classification is a convolution process.…”