KNN-Based Automatic Cropping for Improved Threat Object Recognition in X-Ray Security Images
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
One of the most important applications of computer vision algorithms is the detection of threat objects inx-ray security images. However, in the practical setting, this task is complicated by two properties inherent tothe dataset, namely, the problem of class imbalance and visual complexity. In our previous work, we resolvedthe class imbalance problem by using a GAN-based anomaly detection to balance out the bias induced bytraining a classification model on a non-practical dataset. In this paper, we propose a new method to alleviatethe visual complexity problem by using a KNN-based automatic cropping algorithm to remove distracting andirrelevant information from the x-ray images. We use the cropped images as inputs to our current model. Empirical results show substantial improvement to our model, e.g. about 3% in the practical dataset, thusfurther outperforming previous approaches, which is very critical for security-based applications
Abstract
One of the most important applications of computer vision algorithms is the detection of threat objects inx-ray security images. However, in the practical setting, this task is complicated by two properties inherent tothe dataset, namely, the problem of class imbalance and visual complexity. In our previous work, we resolvedthe class imbalance problem by using a GAN-based anomaly detection to balance out the bias induced bytraining a classification model on a non-practical dataset. In this paper, we propose a new method to alleviatethe visual complexity problem by using a KNN-based automatic cropping algorithm to remove distracting andirrelevant information from the x-ray images. We use the cropped images as inputs to our current model. Empirical results show substantial improvement to our model, e.g. about 3% in the practical dataset, thusfurther outperforming previous approaches, which is very critical for security-based applications
참고문헌 (16)
- Vladimir Riffo, "Automated Detection of Threat Objects Using Adapted Implicit Shape Model", IEEE Transactions on Systems, Man, and Cybernetics: Systems, 46(4), pp.472~482 (2016), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Domingo Mery, "Object Recognition in X-ray Testing Using Adaptive Sparse Representations", Journal of Nondestructive Evaluation, 35(3), pp.9~ (2016), Springer Science and Business Media LLC[DOI]
- Domingo Mery, "Modern Computer Vision Techniques for X-Ray Testing in Baggage Inspection", IEEE Transactions on Systems, Man, and Cybernetics: Systems, 47(4), pp.682~692 (2017), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Samet Akcay, "Using Deep Convolutional Neural Network Architectures for Object Classification and Detection Within X-Ray Baggage Security Imagery", IEEE Transactions on Information Forensics and Security, 13(9), pp.2203~2215 (2018), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Domingo Mery, "GDXray: The Database of X-ray Images for Nondestructive Testing", Journal of Nondestructive Evaluation, 34(4) (2015), Springer Science and Business Media LLC[DOI]
- C. Miao, "SIXray: A Large-scale Security Inspection X-Ray Benchmark for Prohibited Item Discovery in Overlapping Images", Proc. IEEE Conf. Comput. Vis. Pattern Recognit (CVPR) (2019)
- H. He, "Learning from Imbalanced Data", IEEE Trans. Knowl. Data Eng, 21(9), pp.1263~1284 (2009)
- J. K. Dumagpi, "A New GAN-based Anomaly Detection(GBAD)Approach for Mult7i-Threat Object Classification on Large-Scale X-Ray Security Images", IEICE Trans. Inf. & Syst., E103-D(2) (2020)
- J. Chen, "Automatic Image Cropping: A Computational Complexity Study", IEEE Conf. Comput. Vis. Pattern Recognit (CVPR) (2016)
- M. Nishiyama, "Sensation-based photo cropping", Proc. 17th ACM Int. Conf. Multimedia (2009)
- A. Santella, "Gaze-based interaction for semi-automatic photo cropping", Proc. SIGCHI Conf. Human Factors Computing Systems (2006)
- M. Imura, "Automatic Cropping Method for Chest Radiographs Based on Adaptive Binarization", Conf. Proc. IEEE Eng. Med. Biol. Soc. (2013)
- K. He, "Deep Residual Learning for Image Recognition", Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR) (2016)
- D. Mery, "Computer Vision for X-Ray Testing" (2015), Springer International Publishing
- Mark Everingham, "The Pascal Visual Object Classes (VOC) Challenge", International Journal of Computer Vision, 88(2), pp.303~338 (2010), Springer Science and Business Media LLC[DOI]
- A. Paszke, "Automatic Differentiation in Pytorch", NIPS Autodiff Workshop (2017)