Developing an Advanced Deep Learning SSD Algorithm Using Computer Vision Approaches to Enhance Vehicle Detection

Tamara A. Anai

doi:10.71229/w1ta9f93

Authors

Tamara A. Anai Department of Basic Science (Computer Science), Dentistry, Tikrit University, Tikrit, Iraq. https://orcid.org/0000-0002-6010-152X

DOI:

https://doi.org/10.71229/w1ta9f93

Keywords:

Computer Vision,, Deep Learning,, SSD, , Mean Average Precision,, KITTI Vision Benchmark dataset

Abstract

The development of Computer vision and Deep learning frameworks plays a crucial role in Autonomous driving and real-time intelligent traffic surveillance. The Single Shot Detector (SSD) architecture is well known for its ability to perform inference quickly, and generically configured anchors are problematic because of the severe localization errors they cause. This article proposes an optimized architecture for SSD-based vehicle detection that addresses these challenges in two main steps: (1) using an optimized edge-preserving median filter to remove semantic noise in the image caused by the sensors while preserving the sharpness of the image boundaries and (2) applying custom-scaled anchor box aspect ratios specifically designed to capture the real size of the vehicles on the road. Applied to the standard KITTI Vision Benchmark dataset, the proposed framework's Mean Average Precision (mAP) and recall were 95.90% and 96.60%, respectively. The proposed model is compared with the baseline networks and is shown to significantly outperform them, achieving mAP 89.50% for vanilla SSD300 and 94.10% for YOLOv5. The proposed method reached high reliability with a balanced precision-recall rate that is highly suitable for real-time traffic monitoring systems.

Author Biography

Tamara A. Anai, Department of Basic Science (Computer Science), Dentistry, Tikrit University, Tikrit, Iraq.

Tamara Afif Anai received her B.Sc. degree in Computer Science from the University of Technology, Iraq, in 2002. She obtained her M.Sc. degree in Computer Science from Zagazig University, Egypt, in 2016, and her Ph.D. degree in Computer Vision from Suez Canal University, Egypt, in 2023. She is currently a Lecturer at Tikrit University. Her main research interests include Artificial Intelligence, classification problems, optimization techniques, and image processing. She can be contacted at email: tamsamka@tu.edu.iq.

ORCID ID: https://orcid.org/0000-0002-6010-152X

Scopus ID: https://www.scopus.com/authid/detail.uri?authorId=58085154500

Scholar ID: https://scholar.google.com/citations?hl=ar&user=zfxAXnoAAAAJ

ResearchGate: https://www.researchgate.net/profile/Tamara-Anai

Web of Science: https://www.webofscience.com/wos/author/record/GYI-8588-2022

References

1. Wiley V, Thomas L. Computer vision and image processing: a paper review. Artif Intell Res. 2018;2(1):29-36. doi:https://doi.org/10.29099/ijair.v2i1.42

2. Reddy GPO. Digital Image Processing: Principles and Applications, Geospatial Technologies in Land Resources Mapping, Monitoring and Management. Published. Springer International; 2018. doi:http://dx.doi.org/10.1007/978-3-319-78711-4_6

3. Anai, Tamara A, Al-Hashimi M, Anaee, Mays A. Effect of Genetic Algorithm as a Feature Selection for Image Classification. Iraqi J Sci. 2023;64(11):6001–6012. doi:10.24996/ijs.2023.64.11.42

4. Najia, Noora H, Al-Juboori, Ali M. Survey of Palm Print Detection Techniques. J Al-Qadisiyah Comput Sci Math. 2022;14(4):74–81. doi:https://doi.org/10.29304/jqcm.2022.14.4.1088

5. Anai, A T, Mersal S, Mostafa MS. The Effect of Meta-heuristic Methods on the Performance of ImageClassification. Iraqi J Sci. 2024;65(5):2881-2897. doi:10.24996/ijs.2024.65.5.41

6. Shms Aldeen, S. AD. Improving Heart Disease Classification using Multimodal Fusion in Deep Learning. Int J Comput Electron Asp Eng Publ RAME. 2025;6(3):218-227. doi:10.26706/ijceae.6.3.20250812

7. Damien L, Akito H, Geert S, et al. Dicing Lane Quality Quantification & Wafer Assessment Using Image Thresholding Techniques. In: 2024 IEEE 10th Electronics System-Integration Technology Conference (ESTC), Berlin, Germany. ; 2024:1-6. doi:10.1109/ESTC60143.2024.10712068

8. Al-Fatlawia, Talib T, Baawib, Salwa S, Albukhnefis, Adil L. Efficient KeyFrame Extraction based on Adaptive Threshold and HOG for Video Summarization. J Al-Qadisiyah Comput Sci Math. 2025;17(2):308–316. doi:https://doi.org/10.29304/jqcsm.2025.17.22363

9. Hang G, Jinmin L, Tao D, Zhihao O, Xudong R, Shu-Tao X. MambaIR: A Simple Baseline for Image Restoration with State-Space Model. In: European Conference on Computer Vision, Cham: Springer Nature Switzerland. ; 2024:222–241. doi:10.1007/978-3-031-72649-1_13

10. Hamadi C, Abdelhak L, Paolo F. Blind image restoration via fast diffusion inversion. Adv Neural Inf Process Syst. 2025;37:34513-34532. doi:10.48550/arXiv.2405.19572

11. Katamneni , Vinaya S, G. J. An Evolutionary Computing Approach to Solve Object Identification Problem for Fall Detection in Computer Vision-Based Video Surveillance Applications. In: Hemanth, D., Kumar, B., Manavalan, G. (Eds) Recent Advances on Memetic Algorithms and Its Applications in Image Processing. Studies in Computational Intelligence, Vol 873. Springer, Singapore. ; 2020:1-6. doi:10.1007/978-981-15-1362-6_1

12. Minar R, Siddharth C, Asif M, et al. Statistical analysis of design aspects of various YOLO-based deep learning models for object detection. Int J Comput Intell Syst. 2022;13(9):864-874. doi:10.14569/IJACSA.2022.01309100

13. Momina, Liaqat A, Zhou Z. The YOLO framework: A comprehensive review of evolution, applications, and benchmarks in object detection. Computers. Published online 2024:1-36. doi:10.20944/preprints202410.1785.v1

14. Jyotsna J, Prachi R, Prity B. Plant Disease Prediction Using Deep Learning. Int J Comput Electron Asp Eng. 2022;3(2):32-38. doi:10.26706/ijceae.3.2.arset1002

15. Najwan W. Deep Feature Fusion Method for Images Classification. Int J Comput Electron Asp Eng RAME Publ. 2024;5(4):148-153. doi:10.26706/ijceae.5.4.20241103

16. Jin L, Guodong L. An approach on image processing of deep learning based on improved SSD. Symmetry (Basel). 2021;13(3):495. doi:10.3390/sym13030495

17. Jia D, Jialin Z, Chuanwang Z. Detection of cervical cells based on improved SSD network. Multimed Tools Appl. 2022;81(10):13371-13387. doi:10.1007/s11042-021-11015-7

18. Cheng L, Yicai J, Chao L, Xiaojun L, Guangyou F. Improved SSD network for fast concealed object detection and recognition in passive terahertz security images. Sci Rep. 2022;12(1):12082. doi:10.1038/s41598-022-16208-0

19. Xin X, Jing S, Yongqin C, et al. Research on machine vision and deep learning based recognition of cotton seedling aphid infestation level. Front Plant Sci. 2023;14(14):1200901. doi:10.3389/fpls.2023.1200901

20. Juhartini, Dwinita A, Desmiwati. Single Shot Multibox Detector (SSD) in Object Detection: A Review. Int J Adv Comput Informatics. 2025;1(2):118–127. doi:https://doi.org/10.71129/ijaci.v1i2.pp118-127

21. Ramos-Sanchez S, Jinmi L, Ricardo Y, Joyce Z. Object Detection on Road: Vehicles Detection Based on Re-Training Models on NVIDIA-Jetson Platform. J Imaging. 2026;12(1):1-20. doi:10.3390/jimaging12010020

22. Geiger P, Lenz R, Urtasun R. Are we ready for autonomous driving? The KITTI vision benchmark suite. in IEEE Conference on Computer Vision and Pattern Recognition CVPR. Published 2012.

23. Rani S, Kumar R. An effective image preprocessing framework using modified median filtering for disease detection. J Real-Time Image Process. 2023;20(3):45-58. doi:10.1007/s11554-023-01311-x

24. Goodfellow I, Bengio Y, Courville A. Convolutional Networks, in Deep Learning, Ch. 9. 1st ed.; 2016. https://www.deeplearningbook.org/contents/convnets.html

25. Chicco D, Jurman G. The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation. BMC Genomics. 2020;21(1):6. doi:10.1186/s12864-019-6413-7.

26. Zheng Z, Wang P, Liu W, Li J, Ye R, Ren D. Distance-IoU loss: Faster and better learning for bounding box regression. In: In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 34, No. 07. ; 2020:12993-13000. doi:10.1609/aaai.v34i07.6999.

27. Mohammed, Basim O. Detecting Road Depressions Based on Deep Learning Techniques. Int J Comput Electron Asp Eng RAME Publ. 2025;6(3):153-167. doi:10.26706/ijceae.6.3.20250606

28. Mohanad , Dawood S. Performance of fast learning approach to predicting black fungus diseases. Int J Comput Electron Asp Eng RAME Publ. 2022;3(4):70-75. doi:10.26706/ijceae.3.4.2211519

29. Albahli S, Ahmad, G. N. An empirical study of vanilla SSD300 and its variants for efficient object detection. IEEE Access. 2023;11:34567-34578. doi:10.1109/ACCESS

30. Zhang X. Performance evaluation and optimization of YOLOv5 for real-time object detection systems. Sci Rep. 2024;14(1):4321. doi:10.1038/s41598-024-54321-w.

31. Wang Y, Zhang X. Multi-scale feature fusion and anchor box optimization for vehicle detection based on improved Single Shot Detector. J Real-Time Image Process. 2024;21(4):512–525. doi:10.1007/s11554-024-01432-1.

32. Guo B, Hansen JHL. Enhanced Vehicle Detection System with Advanced Deformation Feature Extraction Algorithm. In: 2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC), Edmonton, AB, Canada. IEEE; 2024:707-712. doi:10.1109/ITSC58415.2024.10919643.

33. Bakirci M. Enhancing vehicle detection in intelligent transportation systems via autonomous UAV platform and YOLOv8 integration. 2024;164(Applied Soft Computing):112015. doi:10.1016/j.asoc.2024.112015.

34. Al-Mansoori H, Ahmed M, Al-Ali S. Optimized anchor configurations and preprocessing pipelines for robust urban vehicle detection using SSD networks. IEEE Access. 2025;13:14890–14903. doi:10.1109/ACCESS.2025.3412580.