63 Two-Dimensional Spectral Detector for Baggage Inspection X-Ray System Mykola Shutko1, Volodymyr Shutko2, Lidiya Tereshchenko3, Maksym Zaliskyi4, Iuliia Silantieva5 1. Information Security Devices Department, National Aviation University, UKRAINE, Kyiv, 1 Komarova av., email: Lenochka597@ukr.net 2. Electronics Department, National Aviation University, UKRAINE, Kyiv, 1 Komarova av., email: vnshutko@ukr.net 3. Aviation Radioelectronic Complexes Department, National Aviation University, UKRAINE, Kyiv, 1 Komarova av., email: 10118@ukr.net 4. Aviation Radioelectronic Complexes Department, National Aviation University, UKRAINE, Kyiv, 1 Komarova av., email: maximus2812@ukr.net 5. International Transportation and Customs Control Department, National Transport University, UKRAINE, Kyiv, 1, M. Omelianovycha- Pavlenka Str., email: gmelanine@gmail.com Abstract: The paper deals with two-dimensional II. LITERATURE REVIEW AND PROBLEM STATEMENT spectral detector for baggage inspection X-ray devices. This detector is based on construction of analytical The paper addresses applied research challenges models for internal structure of object under control and concerning development and application of a new method of their spectrum calculation. The methods of projective determination (visualization) of the internal structure of the geometry and Bouguer-Lambert law are applied to obtain objects under control (OC), that enables dangerous OC to be the analytical models for shadows of the three- identified with high probability in real time, increases the dimensional objects. Spectral detector are designed speed of dangerous substances identification in luggage, and according to Neyman-Pearson criterion. Analysis shows provides automation of these processes. In addition, that proposed spectral detector has good operating automatic generation of images of hazardous OC allows for characteristics even at low signal-to-noise ratios. periodic inspections of aviation security service operators. Keywords: aviation security service, X-ray, optical Detection systems based on X-ray, computer tomography imaging, shadow of the three-dimensional objects, and spectroscopy of mobile ions have certain shortcomings spectral detector. [1 – 7]. Some of these systems can detect well-hidden explosives, but their implementation requires considerable I. INTRODUCTION funds. In addition, they have a high level of false alarms (approximately 0.2 ... 0.4). Ensuring effective protection against terrorism is the most difficult issue, especially for countries with a developed air Thus, the development of analytical models for the receipt transport network, a large number of airlines and airports. of multidimensional shadows of translucent objects for further processing will allow the classification of OC, which The problem is complicated by unpredictability of terrorists’ will greatly facilitate the work of operators serving actions. In addition, vulnerabilities in aviation security supervision devices in Aviation Security Service (AvSS), systems (such as procedures for screening airline passengers reducing the value of false alarms. and their baggage, freight shipments, mail, etc.) that can be exploited by law violators should be taken into consideration. Literature analysis showed that modernization of The main way to improve aviation safety is to prevent equipment for AvSS is carried out in two directions: in the part of the improvement of hardware and software. In [8] hazardous objects and substances, explosive devices and authors proposed new X-ray backscatter technique using an weapons on aircraft board. This requires a comprehensive un-collimated powerful (high kW) X-ray beam and an development and introduction of new methods of screening, efficient pinhole camera encompassed with a high resolution detection and identification of dangerous objects under control. matrix detector for imaging of an object. Moreover, a high- Insights of the direct visualization methods indicate that energy X-ray inspection technique for the reliable inspection of air freight container was presented in [9]. they are inherent in the same type of operations: primary Analysis of various strategies for object detection in X-ray radiation exposure of the objects under control in security imagery is given in [10]. Moreover, paper [11] also configuration space (in the case of active method), reradiation deals with a technique for the classification of X-ray baggage reception (scattered or passed through the object), its conversion into an electrical signal, signal processing and images using convolutional neural networks. Application of deep convolutional neural network as classification method in electrical-to-optical signal conversion. medicine X-ray image analysis was considered in [12]. ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic 64 In [13] authors investigated the feasibility of applying Internal visualization of the OC with a complex form, in straight-line-trajectory-based tomographic imaging this case a sphere in the sphere, designed with point source is configurations to security inspections. The method of shown in the Fig. 3. automated target recognition with usage of reference database, which contains X-ray images of OC, for cargo scanning systems was proposed in [14]. The papers [15, 16] deal with procedures of handguns, shuriken and razor blades recognition for baggage inspection. The simulation of the internal structure for OC with simple and complex forms using the point source of irradiation in the center, as well as with the bias relative to the center, is considered in [17]. The method developed for optical imaging of the inner structure of the three-dimensional objects allows obtaining a shadow of these objects, exposed to electromagnetic radiation. It has useful applications in different life spheres, as in medicine, manufacturing industry, in a process of customs supervision of goods and means of Fig. 3. Inner structure imaging for OC transport for commercial use, etc. It allows the AvSS to The simulation shows that the simplest objects have increase the probability of correct detection of hazardous shadows with transient characteristics, half-dooms, materials and reduce false alarms of its security system. For distortions of the type of the crater, where there are generally medicine the method may help to increase the probability of flat irradiating planes. Changing the irradiation angle changes health hazard anomaly detection. the shadow to unrecognizability. To accurately identify the So aim of this paper is synthesis of two-dimensional intended OC, it is necessary to automate the process of spectral detector for baggage inspection X-ray devices. recognizing shadows, taking into account possible distances III. TWO-DIMENSIONAL SPECTRAL DETECTOR between the source, the OC and the screen-receiver, the irradiation angles, etc. The construction of an analytical model reduces to the Methods of analytical modeling of the OC with different calculation of a projective image of isotropic object in the shapes, geometrical dimension, foreshortening, substance and case of homogeneous irradiation by a point source located on appropriate extinction coefficients, used to develop the axis of object symmetry perpendicular to the plane of the procedures for identifying dangerous objects under security image (screen). supervision of passengers and baggage, allow to image OC To determine a position of the radiation source, the OC and inner structure. the screen with a point source it is appropriate to use In order to verify the developed models multidimensional cylindrical coordinate system applied to the Fig. 1. The OC spectra of visualization images are obtained. model with complex form is presented in Fig. 2. Procedure for image processing consists of using a shadow of the object of given shape to construct a two-dimensional spectrum and its subsequent use in developing the standard spectral detector proposed in the research. This detector is invariant with respect to the location of the OC in the working area. The invariance of the calculated spectrum to the location of the OC on the plane of the screen provides the possibility of applying algorithms for the calculation of two-dimensional spatial spectra of the visualization image in relation to the wanted images of some image anomalies in the introscopic imaging systems of the AvSS. Fig. 1. OC scanning: (а) is the setting a cylindrical coordinate That is, the desired density distribution of the object of system; (b) is the setting a scanning beam position control μ(x, y) must be matched to fit its two-dimensional ( ) spatial spectrum – Fourier-image M k x , k y . In the further processing of visualization data, we find solutions in the ( ) frequency space M k x , k y , and then, through the inverse Fourier transform, the desired distribution is calculated µ * ( x, y ) . The resulting distribution is selected according to those images, which are in the memory of the supervision system. A decision is made to detect a particular object after Fig. 2. OC with complex form matching the resulting image µ * ( x, y ) and mask µ * ( x, y ) . ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic 65 Figures 4, 5 shows the spectra of images of different When using X-ray systems in order to provide automation shades of opaque OC of a simple shape on size a 100x100 of care and increase the reliability of decision-making on the screen plane located almost above the center of the screen. presence of prohibited articles and substances in the OC, there are problems of identifying different forms and locations of the OC. For this purpose, on the example of spectral detector model was constructed in the Matlab environment. In this case, the detection occurs regardless of the OC location and regardless of its shape and size. The considered models are the shadows of two objects in a field with specified boundaries. One object is a regular square а) b) (this kind of can have a dynamite), and the other is a model of the machine gun (Fig. 8). Also, white Gaussian noise and a Fig. 4. Shadow (a) of a parallelepiped and its spectrum (b) mixture of image and noise are modeled (Fig. 9). The developed program allows us to detect an OC with a given probability of false alarms for the corresponding threshold decision depending on the size of the OC. The program calculates the probability of correct detection of a signal from an OC. а) b) Fig. 5. Shadow (a) of the cylinder and its spectrum (b) On one plane, the shadows of two parallelepipeds are located, and their spectral images are obtained (Fig. 6). a) b) Fig. 8. Model of shadow OC A mixture of useful signal and noise is shown in Fig. 9. a) b) c) Fig. 6. Shadows of two parallelepipeds and their spectrum: a) shadows of two parallelepipeds; b) three-dimensional image of the spectrum those shadows; c) a two-dimensional projection of the a) b) spectrum of shadows of parallelepipeds Fig. 9. A mixture of signal with noise in cases of signal-to-noise The following figures show the spectral images of the ratios equaled to 2 (a) and 0.5 (b) shadows of the parallelepiped and the spheres that were located in space (Fig. 7). The Neyman-Pearson criterion is applied for optimal detection of an OC. According to the Neyman-Pearson criterion, the threshold level V is determined from the condition that the probability of a correct detection D with the given probability of false alarm F was maximal. Hence, the optimal character of the Neyman-Pearson criterion is that it maximizes the probability of correct detection at a fixed probability of false alarms. a) b) c) In addition, it should be noted that the program calculates Fig. 7. Shadows of parallelepiped and spheres and the spectrum of the characteristics of the detection. An example of these their compatible shadows: a) shadows; b) three-dimensional image characteristics is shown in Fig. 10. of the spectrum those shadows; c) a two-dimensional projection of On these graphs it is seen that when the decision threshold the spectrum of shadows is reduced, the detection characteristic is more efficient, Analysis of the spectra of hazardous and forbidden OC however, the probability of false detection is increased. allows us to create an appropriate database for the further The analysis shows that the developed spectral detector has detection of OC of various shapes and complexity. good detection characteristics even at low signal-to-noise ratios. ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic 66 [5] V. N. Epifanov, et al. Nondestructive inspection, in 5 books. Book 4. Radiation control: Practicum, Ed. by V.V. Sukhorukov, Moscow, Vysshaya shkola, 1992, 321 p. (in Russian). [6] V. V. Sukhorukov, et al. Nondestructive inspection in 5 books. Book 5 Introscopy and automation of control: Practicum, Ed. by V.V. Sukhorukov, Moscow, Vysshaya shkola, 1993, 329 p. (in Russian). [7] X-ray technology: Directory in 2 books, Ed. by V. V. Klyuyev, Moscow, 1980, B.1, 431 p., B.2, 383 p. (in Russian). [8] S. Kolkoori, N. Wrobel, and U. Ewert, "A new X-ray backscatter technology for aviation security applications," Fig. 10. Characteristics of signal detection for sample size 1000 and in 2015 IEEE International Symposium on Technologies probabilities of false alarms F = 0.05 and F = 0.03 for Homeland Security (HST), Waltham, MA, USA, 14-16 IV. CONCLUSION April 2015, pp. 1-5. [9] S. Kolkoori, N. Wrobel, S. Hohendorf, and U. Ewert, The analysis of scientific publications has shown that the "High Energy X-ray Imaging Technology for the most effective methods for the detection and identification of Detection of Dangerous Materials in Air Freight hazardous OCs are transient multi-energy direct X-ray ones. Containers," in 2015 IEEE International Symposium on They provide reliable detection of hazardous OCs. However, Technologies for Homeland Security (HST), Waltham, these methods are complicated, their implementation in the MA, USA, 14-16 April 2015, pp. 1-6. supervisory systems has a significant expenditure of material [10] S. Akcay, and T. P. Breckon, "An evaluation of region resources, and they do not work efficiently with dynamic based object detection strategies within X-ray baggage OCs. At a high probability of correct detection to 0.99, there security imagery," in 2017 IEEE International Conference is a high probability of false alarms from 0.3 to 0.4. on Image Processing (ICIP), Beijing, China, 17-20 Sept. The simulation shows that the simplest OC have shadows 2017, pp. 1337-1341. with transient characteristics, half-dooms, distortions of the [11] S. Akcay, M. E. Kundegorski, M. Devereux, and T. P. type of the crater, where there are generally flat irradiating Breckon, "Transfer learning using convolutional neural planes. Changing the irradiation angle changes the shadow to networks for object classification within X-ray baggage unrecognizability. To accurately identify the intended OC, it security imagery," in 2016 IEEE International Conference is necessary to automate the process of recognizing shadows, on Image Processing (ICIP), Phoenix, AZ, USA, 25-28 taking into account possible distances between the source, the Sept. 2016, pp. 1057-1061. OC and the screen-receiver, the irradiation angles, etc. [12] K. S. Kurachka, and I. M. Tsalka, "Vertebrae detection in Procedure for image processing consists of using given shape X-ray images based on deep convolutional neural OC shadow to construct a two-dimensional spectrum and its networks," in 2017 IEEE 14th International Scientific subsequent use in developing the standard spectral detector. Conference on Informatics, Poprad, Slovakia, 14-16 Nov. This detector is invariant with respect to the location of the 2017, pp 194-196. OC in the working area. In order to solve the problem, a [13] Hewei Gao, Li Zhang, Zhiqiang Chen, Yuxiang Xing, spectral detector model is developed using MatLab software Hui Xue, and Jianping Cheng, "Straight-Line-Trajectory- environment. In this case, the detection occurs regardless of Based X-Ray Tomographic Imaging for Security the OC location or its shape and size. It allows detecting Inspections: System Design, Image Reconstruction and dangerous objects with a high probability of correct detection Preliminary Results," IEEE Transactions on Nuclear and low probabilities of false positives (from 0.03 to 0.05). Science, 2013, Volume 60, Issue 5, pp. 3955-3968. [14] W. Visser, et al. "Automated comparison of X-ray REFERENCES images for cargo scanning," in 2016 IEEE International [1] O. O. Semenov, Theoretical foundations and principles of Carnahan Conference on Security Technology (ICCST), construction of technical devices for aviation security: Orlando, FL, USA, 24-27 Oct. 2016, pp. 1-8. training guide, Kyiv, NAU, 2001, 214 p. (in Ukrainian). [15] D. Mery, and A. K. Katsaggelos, "A Logarithmic X-ray [2] V. N. Epifanov, et al. Nondestructive inspection, in 5 Imaging Model for Baggage Inspection: Simulation and books. Book 1 General questions. Control of Propagating Object Detection," in 2017 IEEE Conference on Substances: Practicum, Ed. by V.V. Sukhorukov, Computer Vision and Pattern Recognition Workshops, Moscow, Vysshaya shkola, 1993, 350 p. (in Russian). Honolulu, HI, USA, 21-26 July 2017, pp. 251-259. [3] Nondestructive inspection, in 5 books. Book 2. Acoustic [16] C. Meghare, and C. S. Gode, "Automated Detection Of methods of control, Ed. by V.V. Sukhorukov. Moscow, Threat Object in X-ray Images of Baggage," International Vysshaya shkola, 1993, 380 p. (in Russian). Journal of Electrical, Electronics and Data [4] V. N. Epifanov, et al. Nondestructive inspection, in 5 Communication, 2017, Volume 5, Issue 6, pp. 97-101. books. Book 3. Electrical, magnetic and eddy current [17] A. A. Semenov, and L. Y. Tereshchenko, "Modeling of testing methods and instruments: Practicum, Ed. by V.V. the visualization of the internal structure of objects of Sukhorukov, Moscow, Vysshaya shkola, 1993, 420 p. (in control," Electronics and control systems, 2008, № 1, pp. Russian). 144-148. (in Ukrainian). ACIT 2018, June 1-3, 2018, Ceske Budejovice, Czech Republic