Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{203401,
        author = {Prof. Phaltane Anjali Dattatry and Ishwari Zambare and Akshada Rohokale and Eshwari Divate},
        title = {Fusion Based Classification of Irregular Land Parcels Using Explainable AI},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {12},
        pages = {11163-11171},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=203401},
        abstract = {Accurate identification of land types is essential for managing natural resources, planning agricultural activities, monitoring environmental changes, and supporting urban development. Traditional approaches depend heavily on visual inspection of satellite and aerial imagery, making the process slow, inconsistent, and difficult to scale for large or diverse geographic areas. To address these challenges, this work introduces an automated land-type classification framework that combines machine learning and deep learning techniques. The system primarily uses a Convolutional Neural Network (CNN) to learn spatial and textural patterns directly from images, while classical models such as Support Vector Machine (SVM) and Random Forest (RF) are used as comparative baselines. The dataset consists of satellite image tiles representing multiple land categories including vegetation, water bodies, barren soil, agricultural regions, and built-up areas. Images undergo essential preprocessing steps such as resizing, normalization, noise reduction, and illumination correction before training. Experimental analysis shows that the CNN outperforms traditional models by effectively distinguishing visually similar terrain and learning richer spatial features. The proposed framework offers a scalable and reliable solution for automated land-type mapping and can support future real-time geospatial analysis in environmental, agricultural, and urban applications. Index Term- Land classification, irregular land parcels, remote sensing, satellite imagery, Convolutional Neural Network (CNN), machine learning, Support Vector Machine (SVM), Random Forest (RF), image preprocessing, feature extraction, environmental monitoring, geospatial analysis.},
        keywords = {Land Type Classification, Remote Sensing, Satellite Imagery, Image Preprocessing, CNN Feature Extraction, Support Vector Machine (SVM), Random Forest (RF), Machine Learning, Deep Learning, Environmental Monitoring.},
        month = {May},
        }