Characterizing Scale Compatibility in Fixed-Scale Vehicle Detection for Monocular Distance-Based Warning

Sasaki, Seigo

doi:10.51094/jxiv.6245

##article.authors##

Sasaki, Seigo National Defense Academy of Japan https://orcid.org/0000-0002-7596-5291 https://researchmap.jp/s5sskrmp

DOI:

https://doi.org/10.51094/jxiv.6245

キーワード:

Fixed-scale detection、 Scale compatibility、 Observation window、 Distance-dependent behavior、 Monocular warning systems、 Cascade classifiers

抄録

This study analyzes the distance-dependent behavior of fixed-scale object detectors and

examines how this behavior affects the reliability of a simple monocular distance-based

warning system. Four cascade detectors trained with different window sizes were evaluated

across multiple distances. The results show that each window size exhibits a distinct

effective operating region: smaller windows perform well at longer distances, larger windows

at shorter distances, and intermediate windows maintain usable performance over a broader

range. These findings indicate that window size functions as a scale-dependent observation

parameter, determining when a fixed-scale detector can represent the target reliably.

Because the warning logic requires a valid detection to estimate distance, the consistency of warning operation reflects whether detections are available near the threshold.

While the study focuses solely on fixed-scale detectors and does not evaluate additional multi-scale or adaptive methods, the results provide insight

into how fixed observation structures behave in scale-varying environments and may help

inform future work on selecting or adapting observation structures for monocular sensing

under constrained computational resources. Overall, the study provides a structural

characterization of scale compatibility in fixed observation structures, clarifying how

detector design choices influence performance in scale-varying environments.

利益相反に関する開示

The author declares no conflicts of interest.

ダウンロード *前日までの集計結果を表示します

ダウンロード実績データは、公開の翌日以降に作成されます。

引用文献

I. Gat, M. Benady, and A. Shashua, “A monocular vision advance warning system for the automotive aftermarket,” SAE Technical Paper, pp. 1–8, 2005.

H. Isobe and Y. Nakamura, “Driving support system using In-vehicle monocular camera and image processing (in japanese),” Bulletin of Hiroshima Institute of Technology. Research Volume, vol. 43, pp. 317–322, 2009.

L. B. N. De Guzman and A. Raymond See, “Using monocular depth estimation for distance estimation in a moving vehicle,” in 2023 International Conference on Computer Graphics and Image Processing (CGIP), Jan. 2023, pp. 19–22.

P. Viola and M. Jones, “Rapid object detection using a boosted cascade of simple features,” in Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001, vol. 1, Dec. 2001, pp. I511–I518.

R. Lienhart and J. Maydt, “An extended set of Haar-like features for rapid object detection,” in Proceedings. International Conference on Image Processing, vol. 1, Sep. 2002, pp. I900–I903.

N. Dalal and B. Triggs, “Histograms of oriented gradients for human detection,” in 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05), vol. 1, Jun. 2005, pp. 886–893 vol. 1.

S. Sivaraman and M. M. Trivedi, “Active learning for on-road vehicle detection: A comparative study,” Machine Vision and Applications, pp. 1–13, 2011.

A. P. Witkin, “Scale-space filtering,” in Proceedings of the Eighth International Joint Conference on Artificial Intelligence - Volume 2, ser. IJCAI’83. San Francisco, CA, USA: Morgan Kaufmann Publishers Inc., Aug. 1983, pp. 1019–1022.

J. J. Koenderink, “The structure of images,” Biological Cybernetics, vol. 50, no. 5, pp. 363–370, Aug. 1984.

T. Lindeberg, “Scale-space theory: A basic tool for analyzing structures at different scales,” Journal of Applied Statistics, vol. 21, no. 1-2, pp. 225–270, Jan. 1994.

——, “Feature detection with automatic scale selection,” International Journal of Computer Vision, vol. 30, no. 2, pp. 79–116, Nov. 1998.

D. G. Lowe, “Distinctive image features from scale-invariant keypoints,” International Journal of Computer Vision, vol. 60, no. 2, pp. 91–110, Nov. 2004.

Tsung-Yi Lin, P. Doll´ar, R. Girshick, K. He, B. Hariharan, and S. Belongie, “Feature pyramid networks for object detection,” in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Jul. 2017, pp. 936–944.

H. Nguyen, “Multiscale feature learning based on enhanced feature pyramid for vehicle detection,” Complexity, vol. 2021, no. 1, p. 5555121, 2021.

D. Ma, F. Zhao, Y. Li, X. Qu, X. Jiang, H. Wu, X. Chen, and M. Liu, “A scalable monocular 3D detector with superpixel feature pyramid network,” Applied Soft Computing, vol. 180, p.113389, Aug. 2025.

P. Dollar, C. Wojek, B. Schiele, and P. Perona, “Pedestrian detection: An evaluation of the state of the art,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 34,no. 4, pp. 743–761, Apr. 2012.

J. Zhang, Q. Su, C. Wang, and H. Gu, “Monocular 3D vehicle detection with multi-instance depth and geometry reasoning for autonomous driving,” Neurocomputing, vol. 403, pp. 182–192, Aug. 2020.

K. Mikolajczyk and C. Schmid, “A performance evaluation of local descriptors,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 27, no. 10, pp. 1615–1630, Oct. 2005.

S. Sasaki and M. Motoki, “AdaBoost approach to vehicle detection for inter-vehicle distance measurement with monocular camera (in Japanese),” in Proc. IEEJ Annual Conference on Electronics, Information and Systems, Sep. 2013, pp. 1466–1470.