Gerçek Zamanlı Derin Öğrenme Yaklaşımı ile Şüpheli Geçişlerin Otomatik Tespiti

Havalimanı, adliye, spor salonu gibi kamusal alanlarda yer alan güvenlik kamera sistemleri ya aranan şahısların bulunması ya da bir operatör tarafından gözlemlenerek şüpheli davranışların tespiti için kullanılmaktadır. Bu tip alanlar içerisinde yer alan koridor gibi bölgelerde sıradan bir kişinin kısa bir süre içinde mükerrer geçişleri şüpheli bir davranış olarak tanımlanabilmektedir. Lakin çok sayıda kameranın tek bir operatör çalışanı tarafından takip edilmesi özellikle kalabalık alanlar içinde bu tip bir şüpheli davranışın tespitini zorlaştırmaktadır. Bu nedenle karar destek sistemlerine ihtiyaç duyulmaktadır. Görüntüler üzerinde kişilerin belirlenmesi, bu kişilerin bir geçici kimlik kazandırılarak bir veri tabanı üzerinde geçişlerinin tutulması gerçekleştirilir ise şüpheli davranışın bir karar destek sistemi tarafından operatöre rapor edilmesi sağlanabilir. Bu nedenle bu çalışma kapsamında halihazırda görüntü sağlayan gerçek zamanlı güvenlik kamera görüntüleri üzerinde MTCNN tabanlı yüz tespiti ve elde edilen şahıs imgeleri üzerinde ResNet ile eğitilmiş bir derin öğrenme algoritması ile gerçekleştirilmiş yüz noktalarından oluşturulmuş kimlik verme yöntemi bir arada kullanılmıştır. Elde edilen geçici kimlik bilgileri yorumlanarak şüpheli davranışların tespit edilebileceği ortaya konulmuştur. Uygulamanın başarısı deneysel olarak sınanmış ve elde edilen sonuçlardaki başarım ve hata nedenleri tartışılmıştır.

Automated Real Time Detection of Suspicious Appearances using Deep Learning

Security camera systems especially in public areas such as airports, courthouses or sports facilities etc. are used to find fugitive persons or detect suspicious behaviors manually under the monitoring of an operator. In hallway-like sections in public facilities, repeated appearances of an unknown ordinary person in a short span of time can be defined as suspicious behavior. However, the fact that multiple cameras are monitored by a single operator makes it harder to detect suspicious behaviors especially in crowded fields. Therefore, support decision systems are required to support operator. If individuals are detected on images automatically and their appearances on the camera are recorded on a database by giving them a temporary identity, suspicious behaviors can be reported to an operator as a support decision system. For this reason, two different methods are used together as a hybrid solution in the study; a MTCNN based facial detection is used on the real time security camera images that currently provide face images, and an identification method, created with facial landmarks produced with a deep learning algorithm that was trained with res-net, was used on the obtained person’s face images. It has been presented that suspicious behaviors can be detected by interpreting the temporary identity information that was obtained. The success of the application was experimentally tested, and the causes of success and failures in the results were discussed.

PDF

___

[1] D. Ciregan, U. Meier and J. Schmidhuber, "Multi- column deep neural networks for image classification," in Proc. of 2012 IEEE Conf. on Computer Vision and Pattern Recognition, June 16-21, 2012, pp. 3642-3649.
[2] A. Memiş, S. Albayrak and F. Bilgili, "A brief overview of medical software tools used in MR image segmentation," in Proc. of 2018 Medical Technologies National Congress, TIPTEKNO 2018, Nov. 8-10, 2018, pp. 1-4.
[3] C. Hark et al., "Doğal dil işleme yaklaşimlari ile yapisal olmayan dökümanlarin benzerliği," in Proc. of 2017 International Artificial Intelligence and Data Processing Symposium, IDAP 2017, Sept. 2017, pp. 1-6.
[4] W. Bao, J. Yue and Y. Rao, “A deep learning framework for financial time series using stacked autoencoders and long-short term memory,” journals.plos.org, Jul. 14, 2017. [Online]. Available: https://journals.plos.org/plosone/article?id=10.1371/jo urnal.pone.0180944. [Accessed: Aug. 10, 2020]
[5] J.C. Russ, “Image Processing,” in Computer- Assisted Microscopy, 1nd ed. Reading, Eds. Boston, MA:Springer, 1990, pp. 33-69.
[6] Y. LeCun, Y. Bengio and G. Hinton, “Deep Learning,” Nature, vol. 521, pp. 436-444, May 2015.
[7] A. Makandar and A. Patrot, "Malware class recognition using image processing techniques," in Proc. of 2017 International Conference on Data Management, Analytics and Innovation, ICDMAI 2017, February 24-26, 2017, pp. 76-80.
[8] J. Kour, M. Hanmandlu, A.Q. Ansari, “Biometrics in Cyber Security,” Defence Science Journal, vol. 66, pp. 600-604, October 2016.
[9] E. Saykol, M. Bastan, U. Gudukbay, O. Ulusoy, “Keyframe Labeling Technique for Surveillance Event Classification” Optical Engineering, vol. 49, pp. 1-12, November 2010.
[10] A. Sukumar, V. Subramaniyaswamy, L. Ravi, V. Vijayakumar and V. Indragandhi, “Robust image steganography approach based on RIWT-Laplacian pyramid and histogram shifting using deep learning,” link.springer.com, Jul. 01, 2020[Online]. Available: https://link.springer.com/article/10.1007/s00530-020- 00665-6. [Accessed: Aug. 10, 2020]
[11] S. Setiowati, Zulfanahri, E. L. Franita and I. Ardiyanto, "A review of optimization method in face recognition: Comparison deep learning and non-deep learning methods," 9th International Conference on Information Technology and Electrical Engineering, ICITEE 2017, Phuket, Thailand, October 12-13, 2017, pp. 1-6.
[12] Y. Taigman, M. Yang, M. Ranzato and L. Wolf, "DeepFace: Closing the Gap to Human-Level Performance in Face Verification," Conference on Computer Vision and Pattern Recognition, IEEE 2014, Columbus, OH, USA, June 23-28, 2014, pp. 1701- 1708.
[13] Y. Sun, L. Ding, X. Wang, and X. Tang. “Deepid3: Face recognition with very deep neural networks,” arxiv.org, Feb. 3, 2015. [Online]. Available: https://arxiv.org/abs/1502.00873. [Accessed: Jul. 2, 2019].
[14] F. Schroff, D. Kalenichenko and J. Philbin, "FaceNet: A unified embedding for face recognition and clustering," IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015, Boston, MA, USA, June 7-12, 2015, pp. 815-823.
[15] M. Liao and X. Gu, “Face recognition based on dictionary learning and subspace learning,” Digital Signal Processing, vol. 90, pp. 110-124, July 2019.
[16] V. N. Vapnik, The Nature of Statistical Learning Theory, 2nd ed. Reading, Eds. New York, MA: Springer, New York, 2000. [Online] Available: SpringerLink.
[17] K. Zhang, Z. Zhang, Z. Li and Y. Qiao, "Joint Face Detection and Alignment Using Multitask Cascaded Convolutional Networks," IEEE Signal Processing Letters, vol. 23, pp. 1499-1503, October 2016.
[18] “Face Detection: Haar Cascade vs. MTCNN,” datawow.io, May. 20, 2020. [Online]. Available: http://datawow.io/blogs/face-detection-haar-cascade- vs-mtcnn. [Accessed: Aug. 10, 2020].
[19] D. E. King, “Dlib-ml: A Machine Learning Toolkit,” Journal of Machine Learning Research, vol. 10, pp. 1755–1758, December 2009.
[20] In the Mind Films. “High School Hallways”, YouTube, Mar. 1, 2017 [Video file]. Available: http://www.youtube.com/watch?v=1ojGb6QU7qY&t= 33s&ab_channel=IntheMindFilms. [Accessed: Jul. 21, 2019].
[21] P. Viola and M. Jones, "Rapid object detection using a boosted cascade of simple features," Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001, Kauai, HI, USA, December 8-14, 2001, pp. 511-518.