Real-Time Object Detection in Occluded Environment with Background Cluttering Effects Using Deep Learning

Syed Muhammad Aamir,Hongbin Ma,Malak Abid Ali Khan,Muhammad Aaqib

2024-01-02

Abstract:Detection of small, undetermined moving objects or objects in an occluded environment with a cluttered background is the main problem of computer vision. This greatly affects the detection accuracy of deep learning models. To overcome these problems, we concentrate on deep learning models for real-time detection of cars and tanks in an occluded environment with a cluttered background employing SSD and YOLO algorithms and improved precision of detection and reduce problems faced by these models. The developed method makes the custom dataset and employs a preprocessing technique to clean the noisy dataset. For training the developed model we apply the data augmentation technique to balance and diversify the data. We fine-tuned, trained, and evaluated these models on the established dataset by applying these techniques and highlighting the results we got more accurately than without applying these techniques. The accuracy and frame per second of the SSD-Mobilenet v2 model are higher than YOLO V3 and YOLO V4. Furthermore, by employing various techniques like data enhancement, noise reduction, parameter optimization, and model fusion we improve the effectiveness of detection and recognition. We further added a counting algorithm, and target attributes experimental comparison, and made a graphical user interface system for the developed model with features of object counting, alerts, status, resolution, and frame per second. Subsequently, to justify the importance of the developed method analysis of YOLO V3, V4, and SSD were incorporated. Which resulted in the overall completion of the proposed method.

Computer Vision and Pattern Recognition,Artificial Intelligence

What problem does this paper attempt to address?

The paper attempts to address the problem of detecting small or uncertain moving targets in occluded environments and cluttered backgrounds. These issues severely affect the detection accuracy of deep learning models. To overcome these challenges, the authors focus on using SSD and YOLO algorithms for real-time detection and improve detection accuracy through techniques such as data preprocessing, data augmentation, parameter optimization, and model fusion.

Specifically, the main contributions of the paper include:
1. **Classifying different frameworks**: Categorizing existing object detection frameworks and identifying areas for improvement.
2. **Studying detection challenges in occluded environments**: Conducting an in-depth study of the various obstacles faced in accurately detecting moving objects in occluded environments.
3. **Building a graphical user interface**: Developing a graphical user interface to display real-time detection results and adding various features such as object counting, alerts, status display, resolution, and frame rate.

Through these methods, the authors improved the model's detection performance in complex backgrounds, achieving higher accuracy and real-time performance.