ANALISIS PERFORMA RESNET-50 UNTUK PERSON RE-IDENTIFICATION PADA LINGKUNGAN LINTAS KAMERA

Authors

  • Muhammad Abdul Khafid Universitas Nusantara PGRI Kediri Indonesia
  • Wahyu Cahyo Utomo Universitas Nusantara PGRI Kediri Indonesia
  • Danang Wahyu Widodo Universitas Nusantara PGRI Kediri Indonesia

DOI:

https://doi.org/10.29407/k3rtqp12

Abstract

Masalah utama pemantauan keamanan kampus lintas kamera adalah hilangnya jejak identitas visual pejalan kaki akibat topologi jaringan non-overlapping dan tingginya variasi intra-kelas. Untuk mengatasi masalah ini, penelitian mengevaluasi metode Person Re-Identification (Re-ID) menggunakan arsitektur ResNet-50 standar. Proses pencocokan identitas dilakukan melalui ekstraksi fitur visual dan perhitungan Cosine Similarity antara citra kueri dan galeri. Hasil pengujian pada 20 target di lingkungan kampus menunjukkan performa pencarian awal yang sangat baik, dengan Rank-1 Accuracy mencapai 100% pada 19 target dan metrik Recall 0,9466. Namun, sistem mencatat Precision yang sangat rendah (0,09-0,13) dan mAP tertinggi 0,3979. Kesimpulannya, model ResNet-50 standar andal menempatkan target utama di peringkat pertama, tetapi masih menghasilkan banyak false positive karena gagal membedakan fitur pejalan kaki berpakaian mirip di bawah variasi pencahayaan ekstrem.

Keywords:

Person Re-Identification, Analisis Performa, ResNet-50, Cosine Similarity, Lintas Kamera

##plugins.themes.default.displayStats.downloads##

##plugins.themes.default.displayStats.noStats##

References

[1]Y.-J. Cho and K.-J. Yoon, “Distance-based camera network topology inference for person re-identification,” Pattern Recognit. Lett., vol. 125, pp. 220–227, 2019, doi: https://doi.org/10.1016/j.patrec.2019.04.009.

[2]N. Narayan, N. Sankaran, S. Setlur, and V. Govindaraju, “Learning deep features for online person tracking using non-overlapping cameras: A survey,” Image Vis. Comput., vol. 89, pp. 222–235, 2019, doi: https://doi.org/10.1016/j.imavis.2019.07.007.

[3]A. Zahra, N. Perwaiz, M. Shahzad, and M. M. Fraz, “Person Re-identification: A Retrospective on Domain Specific Open Challenges and Future Trends,” 2022. [Online]. Available: https://arxiv.org/abs/2202.13121

[4]N.-B. Nguyen, V.-H. Nguyen, T. D. Ngo, and K. M. T. T. Nguyen, “Person re-identification with mutual re-ranking,” Vietnam Journal of Computer Science, vol. 4, no. 4, pp. 233–244, 2017, doi: 10.1007/s40595-016-0093-x.

[5]C.-H. Chuang, T.-C. Huang, C.-W. Wang, J.-H. Lo, and C.-L. Lin, “Person Re-Identification Under Non-Overlapping Cameras Based on Advanced Contextual Embeddings,” Algorithms, vol. 18, no. 11, 2025, doi: 10.3390/a18110714.

[6]A. Zahra, N. Perwaiz, M. Shahzad, and M. M. Fraz, “Person Re-identification: A Retrospective on Domain Specific Open Challenges and Future Trends,” 2022. [Online]. Available: https://arxiv.org/abs/2202.13121

[7]A. Zahra, N. Perwaiz, M. Shahzad, and M. M. Fraz, “Person Re-identification: A Retrospective on Domain Specific Open Challenges and Future Trends,” 2022. [Online]. Available: https://arxiv.org/abs/2202.13121

[8]C. H. Chuang, T. C. Huang, C. W. Wang, J. H. Lo, and C. L. Lin, “Person Re-Identification Under Non-Overlapping Cameras Based on Advanced Contextual Embeddings,” Algorithms, vol. 18, no. 11, Nov. 2025, doi: 10.3390/a18110714.

[9]Y. Chen, Z. Fan, Z. Chen, and Y. Zhu, “CA-Jaccard: Camera-aware Jaccard Distance for Person Re-identification.”

[10]A. Zahra, N. Perwaiz, M. Shahzad, and M. M. Fraz, “Person Re-identification: A Retrospective on Domain Specific Open Challenges and Future Trends,” 2022. [Online]. Available: https://arxiv.org/abs/2202.13121

[11]Y.-J. Cho and K.-J. Yoon, “Distance-based camera network topology inference for person re-identification,” Pattern Recognit. Lett., vol. 125, pp. 220–227, 2019, doi: https://doi.org/10.1016/j.patrec.2019.04.009.

Downloads

Published

2026-07-20

Issue

Vol. 10 No. 3 (2026): Prosiding Seminar Nasional Inovasi Teknologi Tahun 2026

Section

Articles

License

Creative Commons License

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Authors who publish with this journal agree to the following terms:

  1. Copyright on any article is retained by the author(s).
  2. The author grants the journal, right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgment of the work’s authorship and initial publication in this journal.
  3. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal’s published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
  4. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work.
  5. The article and any associated published material is distributed under the Creative Commons Attribution-ShareAlike 4.0 International License

How to Cite

ANALISIS PERFORMA RESNET-50 UNTUK PERSON RE-IDENTIFICATION PADA LINGKUNGAN LINTAS KAMERA. (2026). Prosiding SEMNAS INOTEK (Seminar Nasional Inovasi Teknologi), 10(3), 2945-2955. https://doi.org/10.29407/k3rtqp12