Mobile objects and sensors within a video surveillance system: Spatio-temporal model and queries Dana Codreanu, Ana-Maria Manzat, Florence Sedes Université de Toulouse – IRIT – UMR 5505 118 Route de Narbonne, 31062 Toulouse Cedex 9, France {codreanu, manzat, sedes}@irit.fr ABSTRACT available for all these videos are the id of the camera (eventually The videos recorded by video surveillance systems represent a key GPS coordinates) and a local date/timestamp that are not element in a police inquiry. Based on a spatio-temporal query homogenous throughout the different systems. specified by a victim, (e.g., the trajectory of the victim before and A great majority of the existing video surveillance systems after the aggression) the human operators select the cameras that are manual or semi-automatic (they employ some form of video could contain relevant information and analyse the corresponding processing but with significant human intervention) [11]. Taking video contents. This task becomes cumbersome because of the into account the huge amount of video contents that need to be huge volume of video contents and the cameras’ mobility. This handled, the purely manual approach (agents watching the videos paper presents an approach, which assists the operator in his task and detecting events) becomes insufficient. The main objective in and reduces the research space. We propose to model the the video surveillance domain is to provide users with tools that cameras’ network (fixed and mobile cameras) on top of the city’s could assist them in their research by reducing the research space transportation network. We consider the video surveillance system and therefore the response time. These tools depend on the as a multilayer geographic information system, where the cameras research context and complexity (e.g., real time surveillance of big are situated into a distinct layer, which is added on top of the events, police inquiry) [22]. other layers (e.g., roads, transport) and is related to them by the location. The model is implemented in a spatio-temporal database. Our work is situated in the context of the police inquiry Our final goal is that based on a spatio-temporal query to which involves an a posteriori processing of the data in order to automatically extract the list of cameras (fixed and mobile) help the investigator to highlight (isolate) the relevant elements concerned by the query. We propose to include this automatically (e.g., persons, events). To do that, the investigators dispose of the computed relative position of the cameras as an extension of the set of recorded videos from different video surveillance systems standard ISO 22311. (e.g., public, private, RATP). In order to assist the investigators in their tasks, it is important that the different outputs of the systems are interoperable, which is not currently the case. The interoperability between any video surveillance systems from the 1. INTRODUCTION simple ones with only few cameras to the large scale systems is The number of video surveillance cameras increases in the main goal of the standard ISO 223111. It specifies a format for public and private areas (e.g., in train and metro stations, on- the data which can be exchanged between the video surveillance board of buses and trains, inside commercial areas, inside systems in the inquiry context. enterprises buildings). For example, some estimations show that This standard does not consider the video surveillance there are more than 400000 cameras in London and that only the cameras’ mobility or their fields’ of view modification. In fact, at RATP also known as Régie Autonome des Transports Parisiens the beginnings of video surveillance systems the cameras were (English: Autonomous Operator of Parisian Transports) placed in fixed locations in order to monitor indoor and outdoor surveillance system comprises around 9000 cameras in Paris. In places. With the improvements in the hardware and software these conditions, any person that lives and walks in those two big technologies, on-board cameras are more and more employed in European capitals is likely to be captured many times during a day mobile vehicles (e.g., buses, police cars). This cameras’ mobility (up to 300 times in London) by several video surveillance systems makes the task of security agents even more difficult in the (e.g., the traffic surveillance cameras, the cameras in the subway, context of an inquiry, when they have to analyse a huge amount of and the cameras of a commercial centre). The only markers video contents and to have supplementary knowledge on the system’s characteristics (e.g., the bus timetables, the city transport plan) in order to select the most appropriate video contents. In this context, our goal is to provide users with tools that could assist them in their research and reduce the research space. In order to achieve this objective, in this article, we propose an extension of the ISO 22311 standard in order to take into account 1 http://www.iso.org/iso/fr/catalogue_detail.htm?csnumber=5346 1 Proceedings IMMoA’13 52 http://www.dbis.rwth-aachen.de/IMMoA2013/ the cameras’ mobility. We consider the video surveillance system  Camera type: optical, thermal, infrared as a multilayer geographic information system, where the cameras  Sensor type and dimension: CMOS, CCD are situated on a distinct layer, which is added on top of the other  Transmission type: analogous/ IP layers (e.g., roads, transport) through the location. We  Angle of view (horizontal and vertical), focal implemented our solution using a spatial database in order to distance, pan-tilt-zoom, field of view orientation, select the cameras that might have acquired video contents visible distance etc. corresponding to a user’s spatio-temporal query. The remainder of this paper is organized as follows. After a review of related work concerning the three aspects addressed in this paper, video surveillance systems, standard ISO 22311 and mobile objects modelling in the Section 2, Section 3 presents our multilayer modelling approach. This model is implemented using a spatio-temporal database. Some queries that can be answered based on this database are presented in Section 4. Finally, Section 5 concludes and discusses possible future research. Figure 2: Examples of video surveillance cameras having the same position but different fields of view 2. STATE OF THE ART We started by analysing the way a query is processed in a 2.1 Video Surveillance Systems video surveillance system today. When a person (victim of an The generic schema of a video surveillance system is aggression for example) files a complaint, he is asked to fill a illustrated in Figure 1. The content is captured and stored in a form describing the elements that could help the investigators to distributed manner and analysed in a control centre by human find the relevant video segment (the Figure 3 illustrates an operators that watch a certain number of screens displayed in a example of such form). Based on the spatial and the temporal matrix (the Video Wall in Figure 1). aspects of the query, the surveillance operator uses his own knowledge concerning the spatial disposal of the cameras’ network in order to select the most relevant video contents. Then he analyses these contents by playing them on the different screens that he has in front of him. The monitors themselves show no spatial relationship of any kind, only the numbering of the cameras is in a somewhat logical order. Figure 3: Example of a form filled by a victim Figure 1: Video surveillance system’s schema Therefore, the operators’ tasks become cumbersome taking into consideration the huge volume of video contents to be analysed, the mobility and the different characteristics of cameras. There is a big diversity of cameras and sensors that constitute Moreover, in the current systems, most of the stored contents is the acquisition part of surveillance systems and a heterogeneity of not exploitable because of the recording’s low quality. This lack their installation contexts (e.g., on the halls or platforms of of quality is often caused by inappropriate installation of cameras, railway or metro stations, on-board of trains and buses, on the bad shooting, bad illumination conditions etc. The operator has no streets, in commercial centres or office buildings). Therefore, we a priori knowledge on the quality of the video contents and thus have fixed and mobile cameras having different technical he loses time by visualizing the low quality contents also. characteristics (most of the time dynamic) (see Figure 2 for an example of such cameras) [14]: 2 Proceedings IMMoA’13 53 http://www.dbis.rwth-aachen.de/IMMoA2013/ Figure 4: ISO 22311 sensor description The video surveillance domain has seen a big number of for the data issued from video surveillance systems and the commercial systems developed [8]. In the research area, many metadata needed to exploit that data. projects were developed as well: CROMATICA [5], In the following, we are going to present the ISO 22311 CARETAKER2 [3], VANAHEIM3 for the indoor static video standard, especially the part concerning the description of the surveillance, and SURTRAIN [20], BOSS4 [13], 5 cameras characteristics and mobility. We are going to highlight PROTECTRAIL projects for the on-board mobile surveillance. the interesting elements which relate to our research. All these heterogeneous projects concentrate on the development of the system’s physical architecture and of better detection algorithms in order to obtain a fully automatic system [12], [24]. We can summarize by saying that there is a growing concern 2.2 Standard ISO 22311 The Standard ISO 22311 defines an interoperability format in the research and industrial environments for developing for the data generated by video surveillance systems and for the algorithms for video content analysis (VCA) in order to metadata needed to exploit these huge volumes of data. automatically index content and detect objects (e.g., abandoned packets or luggage) and events (e.g., intrusions, people or vehicles The audio visual packages (containing audio, video or going the wrong way) [16] or to draw operators’ attention to metadata files) have to be structured hierarchically (in files, events of interest (e.g., alarms). However, solutions for assistance folders and groups of folders) according to time intervals in to a posteriori investigation are at a lesser stage of maturity, and to Coordinated Universal Time (UTC). For each group of folders it date most of the data remain unexploited. is mandatory for the system to provide a XML description of the source(s) (e.g., cameras, GPS, video analysis tools), codec(s), file In this article, we are going to address also the lack of formats and a temporal index enabling an easy access to the interoperability between different surveillance systems. In the content. context of an inquiry, the police might need to analyse data from different sources (systems), so it is important that the different The current technologies and processing power enable the outputs of the systems to be interoperable. As a consequence, the analysis of video content and the extraction of metadata big actors of the domain started to unify efforts in order to describing objects, events, scenes etc. This analysis depends on standardize the structure of folders and of metadata files generated the acquisition context (e.g., the position of the camera, the image by video surveillance systems. A result of these efforts is quality, the type of sensors). Therefore, the standard distinguishes represented by the ISO 22311 standard that proposes a structure between the systems, those that can generate such metadata (i.e., level 2 systems) and provides a general structure and dictionary for describing sensors and events (i.e., metadata). 2 http://cordis.europa.eu/ist/kct/caretaker_synopsis.htm As in this paper we are going to address the problem of 3 http://www.vanaheim-project.eu/ cameras’ geo-localization we present the schema for the sensors 4 http://celtic-boss.mik.bme.hu/ description in Figure 4. 5 http://www.protectrail.eu/ 3 Proceedings IMMoA’13 54 http://www.dbis.rwth-aachen.de/IMMoA2013/ Each camera has an absolute location (GPS coordinates) as defined by [9] (e.g., moving points, moving lines, moving more and more of the installed cameras have an embedded GPS regions), or using the dynamic attributes [23] (e.g., motion vector) transmitter. But, there are many cases when the GPS is not which enables to limit the size of the data that has to be stored and enough because: (1) we need to model the position of the camera queried. with regards to the video surveillance system and not to the world; As far as we know, the video content’s mobility is not taken (2) in some situations, for example in indoor environments, the into account in the video surveillance domain. In this article, we GPS positions do not provide a good precision. want to exploit the advances in the field of mobile objects and In the context of a video surveillance system: apply them in the video surveillance domain in order to consider the mobile aspect of surveillance cameras.  The mobile cameras are embedded in buses, train, police cars;  The movement of these vehicles is constrained by a road network and a transportation network. 3. Extension of the Standard 22311 for the By analysing the standard, we can notice that it defines a management of cameras mobility relative position for a camera that is today a simple link to an As you could see in Section 2.2, the Standard 22311, image (the plan of the network of cameras or of a building). This defines a fix position of video surveillance camera, through the kind of location is not easily exploitable. Furthermore, the GPS coordinates and a link to an image containing the plan of the standard does not consider the video surveillance cameras’ network. In order to overcome this issue, we propose to compute a mobility. In order to overcome these issues, we propose to extend relative position with regards to a map which will enable us to: this standard through a multilayer modelling approach, where the  Model the distances between the cameras and select the network of cameras is put on top of a transportation network. relevant cameras for a certain trajectory; In the following, we present a state of the art of the mobile  Model the connections between the cameras ( e.g., possible objects modelling as the cameras’ mobility management path between camera1 and camera2 but not between represents the main focus of this paper. camera2 and camera3 );  Model trajectories for mobile cameras;  Model the fields of view and the maximum detection distances of fixed and mobile cameras. 2.3 Mobile Objects Modelling With the technology’s evolution, the mobility became very In order to achieve this goal we took our inspiration from the important in the context of video surveillance systems. Not only domain of GIS (Geographical Information Systems) [4] and the objects (e.g., persons, cars) are moving in the monitored mobile objects modelling. scene, but also the surveillance cameras are moving. The great majority of the research papers concerning the mobile objects in By considering the video surveillance system as a GIS we the video surveillance domain concentrate on the video content benefit from the separation between the conceptual layers. Thus at analysis in order to detect and track the objects, to interpret their any time, a new layer can be added without modifying the existing behaviour and to understand the visual events of the monitored layers. scene [10]. Thus, the mobility of the cameras is not exploited. In our approach, we propose a four layer model: (1) Road In the field of moving objects, a mobile object means the network, (2) Transportation network, (3) Objects and (4) Cameras continuous evolution of any object over the time, in terms of network. The Figure 5 illustrates the UML model for the first position and dimension [21]. This movement of the mobile three layers. objects can be effectuated in an unconstrained environment [18] The “Road network” layer, presented in blue in Figure 5, is (e.g., for hurricanes, fires) or in a constrained environment [17] based on the graph modelling approach well-known in the (e.g., cars move on road and transportation networks). literature. The road network is considered as an undirected graph In the video surveillance domain, the objects are moving in a G= (V, E), with V a set of vertices and E a set of edges defined constrained environment, mainly by the road network. This according to the granularity level that we want to consider (for a environment is represented as a graph-based model [6], [15], [25], big boulevard of a European capital for example we can consider where the vertices are junctions and the edges are the roads each segment of the road, each segment between two intersections between the two junctions. [9] considers also the connectivity at or the entire boulevard). Each vertex has an identifier and a 2D each junction in order to represent the road network. [19] extends position. Each edge is determined by two vertices. the model proposed by [9] in order to consider the predefined The “Transportation network”, presented in yellow in trajectories that some objects could have (e.g., buses). [7] Figure 5, is also based on a graph model. At this level, the vertices proposes a mobile object data model where they consider the road of the transportation network are intersections between roads, and and rail networks. [2] takes into account the transport network in bus stations. Each transportation vertex has a position with a city as a graph and they add to each graph vertex the transport regards to a road segment. Ordered sequences of transportation modes available (i.e., pedestrian, auto, urban rail, metro, bus). vertices constitute sections, which form lines (e.g., bus lines). The In the management of mobile objects, a major issue is the advantage of our approach with regards to the ones proposed in storage of the objects’ spatio-temporal positions. Several the state of the art [9] is that we have two independent graphs that strategies can be considered: using the spatio-temporal data types are connected to each other through the positions of transportation vertices. That way if the buses stations are modified or new buses 4 Proceedings IMMoA’13 55 http://www.dbis.rwth-aachen.de/IMMoA2013/ lines are introduced we do not have to recompute the underlying On top of all these layers, we model a video surveillance road graph. cameras’ network. A simplified schema of this model is illustrated in Figure 6. The “Objects” layer, presented in red in Figure 5, models the positions of fixed and mobile objects with regards to the The cameras’ network is composed of fixed and mobile underlying layers. cameras. The fixed cameras have a 2D position that is given at installation time. The mobile cameras are associated with mobile The Fixed Object has a position on a road segment. Its objects (e.g., buses) and their trajectory is the same as the object’s position is defined as a distance from each end of the segment. one. For this kind of objects, we adopt the same localisation as the one proposed by [9]. The new generation of digital surveillance cameras has embedded GPS transmitters and even compasses. The In the case of Mobile Objects (e.g., buses, police cars, technologies developed around these cameras make it possible to persons), the position changes in time. Each object will automatically extract information from the camera related to its periodically transmit its position using different strategies (e.g., orientation, pan, tilt, zoom, focal distance, compression each Δt seconds, each time the object is changing the segment, parameters etc. when the object's position predicted by the motion vector deviates from the real position by more than a threshold [23]) that are out Based on all these elements it is possible to model the field of the scope of this article. We suppose that we periodically of view for each camera and track its modifications in time. The receive updates containing time-stamped GPS points that we field of view is computed based on four parameters [1]: the 2D transform into a relative position with regards to the road network position, the viewable angle, the orientation and the visible (i.e., the segments). We use this information to reconstitute distance. A schema of a 2D field of view proposed by [1] is object’s trajectory. shown in Figure 7. We distinguish two types of mobile objects: objects that move freely within the road and transportation networks (e.g., car, person) and objects of which trajectories are constrained by a “line” (e.g., buses). Figure 6: "Cameras network" layer P : camera location () θ : viewable angle d : camera direction vector R: visible distance Figure 7: Illustration of the field of view model in 2D [1] Figure 5: "Road network", "Transportation network" and "Objects" layers 5 Proceedings IMMoA’13 56 http://www.dbis.rwth-aachen.de/IMMoA2013/ Figure 8: General architecture of the system In order to select the most appropriate attributes to describe More precisely, the idea is to compare a spatio-temporal a video surveillance camera, we studied the sensor description query of the user (e.g., Rivoli Street from Louvre to Metro proposed by the ISO 22311 standard, SensorGML6, KML7. We Chatelet the 14th of July between 10h and 14h) with the separated the identified camera’s properties in two categories: trajectories stored in our database and, for a better precision, with properties that could be modified over the time, and fixed the cameras fields of view. The Figure 8 illustrates the generic characteristics. architecture of a system based on our spatio-temporal database for assisting the video surveillance in their research. Thus, the extension of the standard ISO 22311 is realised at three levels: From the Figure 8 it is easy to observe that there are two main questions when developing such system: How to query the  Taking into account the road and transportation networks as system? and How to update the system?. As explained in the a graph and not as an image; previous section our work addresses only the querying aspect that  Taking into account the camera’s relative position and its we are going to describe in the following. mobility on the networks; First, a Query Interpreter module will transform the user  Taking into account the camera’s characteristics change query (e.g, Rivoli Street from Louvre to Metro Chatelet the 14 th of over the time. July between 10h and 14h) in a spatio-temporal query. By spatio- Our model is implemented in a spatio-temporal database temporal query we understand a sequence of road segments and a that can be queried by users in order to retrieve the relevant time interval that will be further transformed in a SQL query, by cameras for a given trajectory. The originality of our research the SQL Query Generator module. The SQL query is executed on work is given by: the database having as a result a list of cameras. Based on some image quality parameters a score per camera can be computed and  the fact that it combines different spatio-temporal the initial list can then be ranked according to this relevance information (e.g., road network, transportation network, score. objects’ positions) and computation (e.g., trajectories, field of view) within the same database; In the following we present two examples of spatio- temporal queries executed on our database implemented in Oracle  the twofold mobility, of the target objects and of the Spatial 8: cameras. In the next section we present the general architecture of the  The first selects the fixed cameras of which geometry tool that could assist the video surveillance operators in their (field of view) intersects the geometry of the Rivoli street; research based on our spatio-temporal database and some SELECT IdCamera examples of queries. FROM FixedCamera WHERE SDO_RELATE( camera_geom, (SELECT street_geom 4. Spatiotemporal database and queries FROM Road Based on the presented model, our goal is to automatically WHERE Name ='Rivoli‘ ), select the cameras (fixed and mobile) that could contain relevant 'mask=OVERLAPBDYDISJOINT querytype=WINDOW' video content with regards to the user query (their field of view )='TRUE'; intersected the query trajectory). 6 http://www.opengeospatial.org/standards/sensorml 8 http://www.oracle.com/fr/products/database/options/spatial/index 7 http://www.opengeospatial.org/standards/kml .html 6 Proceedings IMMoA’13 57 http://www.dbis.rwth-aachen.de/IMMoA2013/  The second selects the mobile cameras that are associated Another perspective of our work is the improvement of the with the buses that crossed the street within the given time resulted cameras list by re-ranking it based on cameras’ interval. characteristics (e.g., image quality, visible distance). LET TimePeriod = Timestamp(hour(2013,1,14,10), hour(2013,1,14, 12)); SELECT ObjetID 6. ACKNOWLEDGMENTS FROM ConstrainedObject This work has been supported by the ANR CSOSG-National WHERE Type.MobileObject= “Bus” AND Security (French National Research Agency) project TimePeriod.ConstrainedObject (atperiods (Timestamp, METHODEO. TimePeriod)); SELECT DISTINCT IdMobileCamera FROM ConstrainedObject, FreeObject, MobileCamera 7. REFERENCES [1] Ay S. A., Zimmermann R., and Kim S.O.. Relevance WHERE Intersect (MobileCamera.geom, Ranking in Georeferenced Video Search. In Multimedia ConstrainedObject.geom) AND Intersect Systems Journal, 16, 2 (March 2010), Springer, 105-125. (MobileCamera.geom, FreeObject.geom); [2] Booth J., Sistla P., Wolfson O., and Cruz I. F. A data model for trip planning in multimodal transportation systems. In Proceedings of the 12th International Conference on 5. CONCLUSION Extending Database Technology: Advances in Database In this paper, we presented a spatio-temporal modelling Technology, ACM, 2009, 994-1005. approach of fixed and mobile cameras within a common transportation network. Taking our inspiration from the multilayer [3] CARETAKER Consortium. Caretaker puts knowledge to representation of the geographical information systems, we model good use. In European Public Transport Magazine. 2008. spatial information about the road and transportation [4] Chang K.-T. Introduction to Geographic Information infrastructures and mobile objects’ trajectories in four Systems, McGraw-Hill Higher Education, 2006, 450 pages independent layers: (1) Road network, (2) Transportation [5] Deparis J.P., Velastin S.A., and Davies A.C. Cromatica network, (3) Objects and (4) Cameras network. project. In Advanced Video-Based Surveillance Systems, The Based on this modelling approach we also proposed a Springer International Series in Engineering and Computer generic architecture for a system that could assist the video Science. volume 488, 1999, 203-212. surveillance operators in their research. Starting from a sequence [6] Dingin Z., and Deng K. Collecting and managing network- of trajectory segments and a temporal interval, such system matched trajectories of moving objects in databases. In generates the list of cameras that could contain relevant Proceedings of the 22nd international conference on information concerning the query (that “saw” the query’s Database and expert systems applications, LNCS Volume trajectory). 6860, 2011, pp 270-279 2011, 270-279. The need of such assisting tools was identified within the [7] El bouziri A., Boulmakoul A., Laurini R. Mobile Object and French National Project METHODEO. Among the project’s Real Time Information System Modeling for Urban partners, we mention the French National Police, Thales and the Environment, In Proceedings of the 26th Urban and RATP also known as Régie Autonome des Transports Parisiens Regional Data Management Symposium, 2007, 403-413. (English: Autonomous Operator of Parisian Transports). Our [8] GDANSK, KU. Deliverable 2.1 – Review of existing smart approach has been validated and will be evaluated within the video surveillance systems capable of being integrated with project. ADDPRIV. ADDPRIV consortium, 2011, www.addpriv.eu Obviously, many questions are still left with no answer [9] Güting R.H., Almeida, V.T., and Ding Z. Modeling and giving way to a large number of perspectives. We will present Querying Moving Objects in Networks. In VLDB Journal, several of them in the following. 15, 2 (2006), 165-190. For now, our model considers only outdoor transportation [10] Joshi K. A., and Thakore D. G. A Survey on Moving Object and surveillance networks. We plan to extend our model to indoor Detection and Tracking in Video Surveillance System, in spaces also in order to model cameras inside train or subway International Journal of Soft Computing and Engineering, 2, stations for example. 3 (July 2012), 44-48 Our work is situated in the context of the a posteriori [11] Kim I.S., Choi H. S., Yi K. M., Choi J. Y., and Kong S. G. research in the case of a police inquiry. We would like to extend Intelligent Visual Surveillance - A Survey. In International this context in the future in order to be able to process real time Journal of Control Automation and Systems, 8, 5 (april queries or to predict trajectories based on some statistics realized 2010), 926-939 based on the stored data (e.g., average speed on some road [12] Lakshmi Devasena C., Revathí R., and Hemalatha M., Video segments). Surveillance Systems-A Survey, in International Journal of Computer Science Issues, 8, 4(July 2011), 635-642 7 Proceedings IMMoA’13 58 http://www.dbis.rwth-aachen.de/IMMoA2013/ [13] Lamy-Bergot C., Ambellouis S., Khoudour L., Sanz D., Conference on Advanced Geographic Information Systems, Malouch N.,. Hocquard A, Bruyelle J-L., Petit L., Cappa A., Applications, and Services, IARIA, 2012, 222-231 Barro A., Villalta E., Jeney G., and Egedy K. Transport [20] Priam Q.-C., Lapeyronnie A., Baudry C., Lucat L., Sayd P., system architecture for on board wireless secured a/v Ambellouis S., Sodoyer D., Flancquart A., Barcelo A.-C., surveillance and sensing. In Proceedings of the 9th Heer F., Ganansia F., and Delcourt V. Audio-video international conference on Intelligent Transport Systems surveillance system for public transportation. In Proceeding Telecommunications, IEEE, 2009, 564-568. of the 2nd international conference on Image Processing [14] Le Barz, C. and Lamarque, T. Video Surveillance Cameras. Theory Tools and Applications, IEEE, 2010, 47-53. In Jean-Yves Dufour, editor, Intelligent Video Surveillance [21] Schneider M. Moving Objects in Databases and GIS: State- Systems, chapter 4, Wiley, novembre 2012. 33-46 of-the-Art and Open Problems, in Research Trends in [15] Liu K., Li Y., He F., Xu J., and Ding Z. Effective Map- Geographic Information Science. Springer-Verlag, 2009, matching on the Most Simplified Road Network. In 169-188. Proceedings of the 20th International Conference on [22] Sedes F., Sulzer J.F., Marraud D., Mulat Ch., and Cepas B.. Advances in Geographic Information Systems (SIGSPATIAL A Posteriori Analysis for Investigative Purposes. In Jean- '12). ACM, 2012, 609-612. Yves Dufour, editor, Intelligent Video Surveillance Systems, [16] Marraud, D. Cepas, B. and Reithler, L., Semantic browsing chapter 3, Wiley, novembre 2012, 33-46. of video surveillance databases through Online Generic [23] Sistla A.P., Wolfson O., Chamberlain S., and Dao S. Indexing, in Proceeding of the 3rd ACM/IEEE International Querying the Uncertain Posi-tion of Moving Objects. In: Conference on Distributed Smart Cameras, 2009, .1-8. Temporal Databases: Research and Practice, LNCS 1399, [17] McKenney M, and Schneider M, Spatial Partition Graphs: A Springer, 1998, 310–337 Graph Theoretic Model of Maps. In Proceedingsof the 10th [24] Soro S. and Heinzelman W. A Survey of Visual Sensor Int. Symp. on Spatial and Temporal Databases, LNCS 4605, Networks, in Advances in Multimedia, vol. 2009, 2009, 1-22 Springer, 2007, 167–184. [25] Xu J., Guo L., Ding Z., Sun X., and Liu C. Traffic aware [18] Parent Ch., Spaccapietra S., and Zimányi E. Conceptual route planning in dynamic road networks. In Proceedings of Modeling for Traditional and Spatio-Temporal Applications: the 17th international conference on Database Systems for The MADS Approach. Springer-Verlag New York, 2006 Advanced Applications, LNCS 7238, 2012, 576-591 [19] Popa I.S., and Zeitouni K. Modeling and Querying Mobile Location Sensor Data, In Proceedings of the 4th Int. 8 Proceedings IMMoA’13 59 http://www.dbis.rwth-aachen.de/IMMoA2013/