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{184625,
        author = {Namrata Rajendra Tongale and Payal Adagale and Harshada Deshmukh and Shital A. Karande},
        title = {Automated Classification and Metadata Generation for Space Video Content Using Deep Learning and NLP Techniques},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {4},
        pages = {2889-2896},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=184625},
        abstract = {The management and classification of space video content pose a significant challenge due to the increasing volume of mission-related footage, including satellite launches, technical discussions, and public outreach programs. Manual tagging and categorization of such videos are time-consuming, inconsistent, and prone to human error. To address this challenge, this paper proposes an automated system that integrates deep learning and natural language processing techniques for metadata extraction and content classification.

The system consists of multiple modules: object detection using YOLOv8 for identifying space-related elements in video frames, optical character recognition (OCR) using Tesseract for extracting embedded textual information such as mission names and captions, and named entity recognition (NER) using Stanford NER to identify and classify key entities like scientists, locations, and events. 
These modules feed into a CNN-based genre classifier that categorizes video content into predefined genres such as “Launch Event,” “Interview,” “Educational Program,” and “Public Event.”

The proposed system was evaluated using a space video dataset provided by the Indian Space Research Organization (ISRO) under the Smart India Hackathon 2023 initiative. The CNN-based classifier achieved an accuracy of 90%, with notable improvements in metadata generation efficiency (approximately 70%) compared to manual tagging. The integration of visual and textual metadata enables more effective indexing, search, and retrieval of archival space videos, supporting applications in education, research, and mission documentation.},
        keywords = {Space video analysis, Metadata generation, Video classification, Deep learning, ISRO dataset.},
        month = {September},
        }