A Novel Remote Sensing Image Retrieval Method Using Intersected Similarity and Clustering Index (ISCI) Transfer Learning

The volume of remote sensing imagery data has gone up significantly. The capacity to efficiently and rapidly locate the necessary images from vast remote sensing repositories is the valuable parameter toward the structure, administration, and exchange of remote sensor image information. The primary research contributions toward, (1) Analyzing Classification process using Machine Learning Algorithm of satellite images using Histogram of Oriented Gradients (HOG) Features, the model doesn’t lose the high impact features and create the closest definition to the classifiers. (2) High-Resolution Remote Sensing Satellite Images Classification and Retrieval Model based on the Features Gray Level Co-Occurrence Matrix (GLCM). It produces a higher performance level by testing the accuracy greater than the previous model, but not scoring above 85% of results. (3) The novelty of the proposed work is Intersected Similarity and Clustering Index (ISCI). ISCI is a dual optimized technique to retrieve the image from a huge database. The problem statement is unsupervised learning and it produces the results above 95% for all the classes. A unique image retrieval technique has been designed based on the studies that have been thoroughly explored.