Smart Intruder Detection System

Abstract

In recent years, . This creates a dangerous environment in which people live in fear. In today’s environment, the issue of home security is a cause of concern. The conventional systems used for intruder detection are very costly, and there is also a possibility of false alarms. The problem of false alarms is eliminated by using OpenCV and a mobile phone to develop a system that can effectively detect an intruder by eliminating the movements of objects that are also moving. When an intruder is detected, the system sends an alert to the user through an email, and the video is also stored in the local storage.

Introduction

The paper addresses the limitations of traditional security and surveillance systems, which rely on manual CCTV monitoring and basic motion detection, leading to high false alarm rates and delayed responses. Environmental factors such as lighting changes, moving objects, and animals further reduce the reliability of conventional systems. With increasing security threats and unauthorized access, there is a growing need for intelligent, automated surveillance solutions capable of accurately detecting human intrusions in real time.

To overcome these challenges, the study proposes a Smart Intruder Detection System that uses computer vision and human detection techniques to analyze live video streams from surveillance cameras. The system employs image preprocessing and classification models to distinguish human intruders from non-threatening objects, thereby reducing false alarms. It provides real-time monitoring, automated alerts via email, and event storage for evidence, offering a cost-effective and reliable security solution for homes, offices, and restricted areas.

The literature review highlights existing image-processing-based intruder detection systems and identifies their key limitations, including false alarms, reliance on manual supervision, sensitivity to environmental conditions, and high implementation costs. The proposed system addresses these issues through automated detection and intelligent processing.

Methodologically, the system integrates camera input, person detection using OpenCV, Python-based decision logic, alert generation, and data storage. The model is trained and tested using a combination of custom and available surveillance videos, labeled as intruder and non-intruder frames, with a 70:30 training–testing split. Performance is evaluated using accuracy, precision, recall, and F1-score, demonstrating improved detection reliability and reduced false alarms compared to traditional surveillance approaches.

Conclusion

This paper presented a smart intruder detection system that was intended to enhance further the accuracy and dependability of most modern surveillance systems. The proposed system utilizes computer vision and deep learning in detecting human intruders automatically from real-time video streams to reduce reliance on manual monitoring and minimize false alarms generated by traditional motion-based systems. The major outputs of this work are enhanced detection accuracy, reduced false alarm rate, and capability of generating alerts in real time. The proposed system is cost- effective and scalable for smart homes, offices, and restricted areas, hence a practical approach for intelligent security and surveillance applications. We have implemented an automatic text detection technique from an image for Inpainting. Our algorithm successfully detects the text region from the image which consists of mixed text-picture-graphic regions. We have applied our algorithm on many images and found that it successfully detect the text region.

References

[1] G. Bradski, “The OpenCV Library,” Dr. Dobb’s Journal of Software Tools, 2000. [2] G. Bradski, “The OpenCV Library,” Dr. Dobb’s Journal of Software Tools, 2000.N. Dalal and B. Triggs, “Histograms of Oriented Gradients for Detection,” IEEE Computer Society [3] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi,“You Only Look Once: Unified, Real-Time Object Detection,” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 779– 788, 2016. [4] S. Ren, K. He, R. Girshick, and J. Sun, “Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks, ” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 39, no. 6, pp. 1137–1149, 2017. [5] A. M. Patil and R. R. Deshmukh, “Smart Surveillance System Using Image Processing,” Proceedings of the International Conference on Computing, Communication and Automation, IEEE, pp. 110–115, 2017. [6] J. Redmon and A. Farhadi, “YOLOv3: An Incremental Improvement,” arXiv preprint arXiv:1804.02767, 2018. [7] S. Saha, “A Comprehensive Guide to Convolutional Neural Networks,” Towards Data Science, 2018. [8] P. Kumar and S. Sharma, “Human Detection and Tracking for Video Surveillance Applications,” Multimedia Tools and Applications, Springer, vol. 78, no. 14, pp. 19845– 19860, 2019. [9] R. K. Saini and M. Kaur, “An Intelligent Surveillance System for Intruder Detection Using Computer Vision,” International Journal of Information Technology, Springer, vol. 12, no. 3, pp. 789–796, 2020.

Copyright

Copyright © 2026 V. C. Sanap, Tanishka Sonawane, Diya Soni, Aakanksha Nanaware, Antra Prajapati. This 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.