KR101502505B1 - Method for evaluating safety performance of target space and apparatus thereof - Google Patents

Abstract

The present invention relates to a method and a device thereof to evaluate a security safety grade for a target space. According to the present invention, provided are a method to evaluate a security safety grade for a target space, which includes s a step of modeling a 3D virtual space in which multiple virtual cameras are placed in response to the target space; a step of setting a moving path for a virtual object having a set size in the 3D virtual space; a step of separately obtaining photographing images from the virtual cameras while moving the virtual object along the moving path at set intervals; a step of calculating scores for each of the virtual cameras corresponding to a recognition degree of the virtual object in the photographing images, and adding up the scores; and a step of evaluating a security safety grade for the moving path or monitoring utilization degree for each of the virtual cameras by using the added scores calculated for each of the virtual cameras, and a method to evaluate a monitoring safety grade for sub areas made by dividing the target space at regular intervals. According to the method and the device thereof to evaluate a security safety grade, the monitoring safety grade for the target space in which security cameras are installed is evaluated in a 3D virtual space through simulation and thus, the utilization of a CCTV is improved and a more reliable and safer security environment is built.

Description

FIELD OF THE INVENTION [0001] The present invention relates to a method,

The present invention relates to a method and an apparatus for evaluating security safety of a target space, and more particularly, to a method and system for evaluating security safety of a target space, .

Since the introduction of CCTV in Korea, CCTV has been used as an excellent means of detecting movements of animals and plants including human beings due to the development of information and communication technology. Recently, the use of CCTV cameras has been increasing due to the increase of the environment exposed to crime, the social atmosphere in which women and socially weak people such as sexual crimes and special crimes are anxious. In particular, the CCTV in the dense area has brought the psychological stability of the residents due to the crime prevention effect as the scope of application for crime prevention and investigation is expanded.

Construction companies are also installing CCTV in and out of public housing complexes to help the residents to live their daily lives safely, but the feeling of safety felt by residents is insufficient. In case of crime, the general countermeasures are to increase the number of CCTV installations or to increase the number of pixels of the camera. In case of CCTV, The focus direction and the angle of view are displayed. Therefore, it is time to pay attention not only to the hardware investment but also to the monitoring efficiency and safety through software approach.

Generally, the owner of the building or place installs the ground and underground CCTV in the apartment complex according to their own standards, but relies on the designer's discretion. The verification of efficiency by CCTV cameras and the consideration of areas that CCTV can not detect This is not done at all and a quantitative approach is needed. That is, up to now, there has been no evaluation on the monitoring efficiency depending on the installation location or angle of the CCTVs deployed. From the viewpoint of spatial information, as the utilization of underground space has recently increased, the scale of the underground parking lot and the direction of the passageway have become complicated, so that a more quantitative approach is required. If we can increase the utilization of CCTV through a software approach, we can contribute to the safety improvement of the people with the same budget.

The technology to be a background of the present invention is disclosed in Korean Registered Patent No. 1310238 (registered on Sep. 13, 2013).

An object of the present invention is to provide a security security evaluation method and apparatus for a target space in which security surveillance about a target space in which security cameras are installed can be evaluated through simulation.

The present invention relates to a method for designing a virtual space, the method comprising the steps of: modeling a three-dimensional virtual space in which a plurality of virtual cameras are arranged corresponding to a target space; setting a movement path for a virtual object having a predetermined size in the three- A step of acquiring a captured image from each of the plurality of virtual cameras while moving the virtual object at a set interval along the movement path; and a step of acquiring a score corresponding to the degree of recognition of the virtual object in the captured image Calculating a surveillance degree for each virtual camera using the summed score calculated for each virtual camera or a security security degree for the moving route; To provide a method for evaluating the security safety of the system.

Here, the security degree of security is proportional to the summed score, and the degree of recognition of the virtual object may correspond to the degree of recognition of the region of interest predetermined for the virtual object.

Also, if the virtual object is an avatar, the region of interest is a head region, and if the virtual object is a vehicle, the region of interest may be a license plate region.

In addition, the degree of recognition of the virtual object may correspond to the imaging size of the pixel of interest pertaining to the region of interest recognized in the captured image.

In addition, the score corresponding to the recognition degree may be a score corresponding to the ratio of the imaging size to the reference size in units of pixels known to the interested region.

Here, the plurality of virtual cameras are respectively modeled on the three-dimensional virtual space in correspondence with the positions and angles of the plurality of security cameras installed in the target space, and the score corresponding to the recognition degree is the Is a mean or a product of a score corresponding to a horizontal size recognized per pixel unit recognized for the region of interest and a score corresponding to the vertical size, and the recognized horizontal size is calculated as a direction of movement of the virtual camera Obtained by projecting the image of the virtual object on the surface of the virtual object, and a value obtained by projecting the recognized vertical size on a plane perpendicular to the ground surface, The corrected value can be used to convert the score.

The present invention also provides a method of designing a virtual space, comprising the steps of: modeling a three-dimensional virtual space in which a plurality of virtual cameras are arranged corresponding to a target space; dividing the three-dimensional virtual space into a plurality of sub- A step of acquiring a captured image from each of the plurality of virtual cameras by arranging a virtual object having a size corresponding to the degree of recognition of the virtual object in the captured image with respect to each of the plurality of sub- And summing the scores for each of the plurality of virtual cameras; and evaluating security security for each subarea using the summed scores calculated for each subarea. And provides a method for evaluating the degree of the road.

Here, the security security degree is proportional to the summed score, and the security security degree evaluation method for the target space may further include displaying each of the sub zones with a color corresponding to the evaluated security degree of safety have.

The virtual object is an avatar or a vehicle, and the degree of recognition of the virtual object corresponds to a degree of recognition of a head part of the avatar or a license plate part, which is a predetermined interested area with respect to the virtual object, May be a score corresponding to the number of pixels constituting the region of interest recognized in the captured image.

Also, the security security evaluation method for the object space may include setting a virtual surface on the bottom surface or the virtual surface perpendicular to the bottom surface constituting the virtual space to the virtual object in each subarea, It is possible to further evaluate the security security degree of the bottom surface.

The three-dimensional modeling unit includes a three-dimensional modeling unit for modeling a three-dimensional virtual space in which a plurality of virtual cameras are arranged in correspondence with the object space, a moving path setting unit for setting a moving path for a virtual object having a predetermined size in the three- An image acquiring unit for acquiring a captured image from each of the plurality of virtual cameras while moving the virtual object along a moving path at a set interval; A point conversion unit for calculating and adding a score corresponding to the degree of recognition of an object for each of the setting intervals; and a monitoring utilization for each virtual camera or a security for the moving route using the summed scores calculated for each virtual camera A security safety evaluation device for a target space including a safety evaluation part for evaluating safety Provided.

The three-dimensional modeling unit includes: a three-dimensional modeling unit for modeling a three-dimensional virtual space in which a plurality of virtual cameras are arranged in correspondence with the object space; a zone dividing unit for dividing the three-dimensional virtual space into a plurality of sub- An image acquiring unit for arranging a virtual object having a predetermined size in each of the subareas to acquire a captured image from each of the plurality of virtual cameras; and an image acquiring unit for acquiring, for each of the subareas, A score conversion unit for calculating a score corresponding to the degree of recognition of a virtual object and summing the score corresponding to the plurality of virtual cameras; and a safety evaluation unit for evaluating a security security degree for each subarea using the summed scores calculated for each subarea And an evaluation unit for evaluating the security of the object space.

Here, the security degree of safety is proportional to the sum of the scores, and the safety degree evaluation unit may display each of the sub-zones in a color corresponding to the evaluated security degree of security.

According to the method and apparatus for evaluating security security of a target space according to the present invention, monitoring safety of a target space in which security cameras are installed can be evaluated through simulation in a 3D virtual space. That is, the present invention can effectively evaluate the monitoring efficiency according to the installation location and angle of CCTV through a software approach, and can increase the utilization of CCTV. Furthermore, it is possible to induce the most effective CCTV placement or relocation to the target space, and there is an advantage that a more secure and reliable security environment can be established.

FIG. 1 is a block diagram of an apparatus for evaluating security security of a target space according to a first embodiment of the present invention.
FIG. 2 is a flowchart of a security security evaluation method using FIG. 1. FIG.
FIG. 3 is an example of a three-dimensional modeling of an underground parking lot according to the method of FIG.
4 is a partially enlarged view of Fig.
Fig. 5 is an example of setting a main movement route for Fig. 3. Fig.
FIG. 6 is an exemplary view illustrating an avatar used in the method of FIG. 2; FIG.
Fig. 7 shows an example of photographing of an avatar by a virtual camera in the method of Fig.
FIG. 8 is a block diagram of an apparatus for evaluating security security of a target space according to a second embodiment of the present invention.
FIG. 9 is a flowchart of a security security evaluation method using FIG. 8. FIG.
FIG. 10 is a diagram illustrating the monitoring safety degree of each area (floor) analyzed through the method of FIG. 8 by color legend.
FIG. 11 shows a result of evaluating security security of the floor by using the method of FIG. 8. FIG.
Fig. 12 shows a conceptual diagram of photographing license plates in the method of Fig. 8. Fig.
FIG. 13 shows the result of evaluating the security degree of security for each license plate by using the method of FIG.
Fig. 14 shows a conceptual diagram for photographing an avatar in the method of Fig. 8. Fig.
FIG. 15 shows a result of evaluating the security security degree of each zone for the avatar using the method of FIG.
Fig. 16 shows the results of Figs. 11, 13, and 15 in an integrated fashion.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention.

The present invention is for evaluating the safety and security of a target space. It is possible to determine the monitoring efficiency of a security camera (ex. CCTV) disposed in a target space through software simulation using 3D modeling, A reliable security environment can be established.

Generally, a plurality of CCTVs (hereinafter referred to as security cameras) are installed in the target space (ex, underground parking lot) at a predetermined angle. In the conventional case, the position and angle of the security camera are arbitrarily determined based on the design plan of the target space. However, this does not directly reflect the surveillance efficiency of the security camera, which may cause an area that the security camera can not detect, and there is a problem in that the security security is lowered.

In the case of this embodiment, it is possible to effectively evaluate the security security degree according to the installation position and angle of the security cameras to be installed or installed in the target space through a software approach.

In the following embodiment, the target space is an indoor space such as an underground parking lot, but the present invention is not necessarily limited thereto. That is, the object space is indoors or outdoors where the security camera can be placed. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an apparatus and method for evaluating security security of a target space according to an embodiment of the present invention will be described in detail.

FIG. 1 is a block diagram of an apparatus for evaluating security and safety of a target space according to a first embodiment of the present invention, and FIG. 2 is a flowchart of a method for evaluating security and security using FIG. 1 and 2, the apparatus 100 for evaluating security of safety according to the first embodiment of the present invention includes a three-dimensional modeling unit 110, a path setting unit 120, an image obtaining unit 130, A conversion unit 140, and a safety evaluation unit 150.

First, the three-dimensional modeling unit 110 models a three-dimensional virtual space in which a plurality of virtual cameras are arranged corresponding to a target space (S210). The step S210 is a process of modeling a plurality of virtual cameras on the three-dimensional virtual space, respectively, corresponding to the installation positions and angles of the plurality of security cameras installed in the target space or expected to be installed.

In general, indoor space is composed of many elements. In the present embodiment, since the surveillance performance of the security camera disposed in the indoor space is spatially analyzed, it is not necessary to consider all indoor space factors. Also, in general, a security camera tends to be disposed at the bottom of a structure (various piping, VMS, etc.) attached to a ceiling for securing a viewing angle. Accordingly, in the present embodiment, the indoor space information defines the floor, the pillar, the wall, the parking surface, and the entry entrance as constituent elements. Floor, wall, entry entrance, and parking surface are entered as surface information, and columns are entered as section and height information. In case of ceiling, it is not included in the modeling component of indoor spatial information because it may interfere with inquiry of other components. The information necessary for 3D modeling can be obtained through the floor plan of underground parking lot, the sectional view of the underground parking lot (vertical and horizontal sectional view), and the details of the parking lot entrance and exit.

FIG. 3 is an example of a three-dimensional modeling of an underground parking lot according to the method of FIG. FIG. 3 is a 3D model of indoor space information for an arbitrary underground parking lot. The components of the indoor spatial information are wall, pillar, floor, parking surface, entrance entrance, and each component in the apartment complex is defined as individual color. The floor is gray, the pillars are blue, the walls are green, the entrance is black, and the parking side is brown. Of course, the color of each component corresponds to one embodiment and can be changed at any time.

4 is a partially enlarged view of Fig. Referring to FIG. 3 and FIG. 4, it can be seen that a plurality of virtual cameras are arranged at individual positions on the three-dimensional space. In addition, the line of sight (direction vector) and the angle of view information of each virtual camera are shown by corresponding red and green solid lines

In step S210, the position, the direction vector, the angle of view, and the like of each virtual camera are arranged according to the field or the design scheme. When exploiting the design diagrams, the virtual camera is 3D modeled on the basis of the documented specifications of the security camera and the resulting 2D layout. In order to reflect the results of the field survey, a method of arranging the markers in the real space and mapping the virtual space and the real space, and a method of calculating the number of pixels of the real space markings on the image pickup unit, A method of measuring the camera specifications can be used.

For reference, if a person with a height of 1.7 m is photographed with a 1/3 inch CCD camera at a distance of 10 m, the focal length of the lens 'f = (distance from the subject x effective vertical height of the CCD) / subject size' F = (10000 mm x 3.6 mm) / 1700 mm = 21.17 mm. It can be seen from this equation that the size of the lens is preferably 25 mm or more.

Table 1 shows the horizontal and vertical dimension conversion tables for each CCD type.

Size nominal = inch / mm Calculation Dimensions (mm) Effective screen size (mm) Diagonal (diameter D) Horizontal W Vertical H Horizontal w Height h 1 "/ 25.4 20.3 15.2 12.8 9.6 2/3 "/ 17.0 13.6 10.3 8.8 6.6 1/2 "/ 12.7 10.2 7.6 6.4 4.8 1/3 "/ 8.45 6.7 5.0 4.8 3.6 1/4 "/ 6.35 5.1 3.8 3.6 2.7

After the 3D virtual space is modeled as described above, the path setting unit 120 sets a movement path for a virtual object having a predetermined size in the 3D virtual space (S220).

In the present embodiment, a moving line (moving path) of a virtual object (ex, avatar) is defined after 3D modeling of the virtual camera. As the avatar is moved through this line, it is evaluated how much the virtual cameras can monitor the avatar and the number of points is quantified. Of course, at the time of setting the moving route, the position of at least one I / O port dependent on the target space is taken into consideration. A description of the size of the virtual object will be given later.

In step S220, the main moving path of the pedestrian to the indoor space is estimated. In this case, it is necessary to select an appropriate number of routes, and set the main route considering the entrance. Here, the movement route can be set to be slightly offset toward the parking space from the center of the road between the parking spaces so as not to overlap with the vehicle motion line. In the case of the underground parking lot of the apartment, since the parking area of the residents by the east is generally set near the same entrance, three to four standard travel routes can be set for each entrance.

Fig. 5 is an example in which a main movement route is set with respect to Fig. 3, and the movement route is indicated by a dotted line. There are three stairwells in the site (A, B, C). Residents using each staircase usually park close to it. Three to four representative routes were estimated for each staircase. The names of the routes are A-1, A-2, A-3, A-4, B-1, B-2, B-3, B-4, C-1, C-2 and C-3 do. In this way, simulations are performed by estimating several travel routes for each apartment. Of course, the path to be used for the simulation is not necessarily limited to the example shown in FIG. 5, and various paths (not shown) may be set as a simulation target. This path setting can be performed automatically by the system in consideration of the attributes and position of each component of the 3D modeling, or can be directly input from the user.

Thereafter, the image obtaining unit 130 obtains the captured images from the plurality of virtual cameras while moving the virtual object along the moving path at a set interval (ex, interval of 1 m) (S230). This image acquisition process is also a process of simulation performed in virtual space.

The virtual object may correspond to an avatar or a vehicle. Specifically, it is possible to reflect the recognition degree of the head part in the case of the avatar and the degree of the photographing recognition in the case of the license plate in the safety degree evaluation in the case of the vehicle. Each of these areas is a part that is easy to use for crime arrest, so it needs to be watched more carefully than other parts.

In the case of this embodiment, an avatar simulating a criminal is used in relation to security and safety of the target space. FIG. 6 is an exemplary view illustrating an avatar used in the method of FIG. 2; FIG. The avatar is simply defined as the head and torso. Using the body dimension data of the Ministry of Commerce, Industry and Energy, the criminals were assumed to be men aged 30 to 34, with an average height of 1724.43 mm, an average head vertical length of 213.09 mm and an average head width of 153.06 mm. The size of the avatar of FIG. 6 reflects this.

Next, the score conversion unit 140 calculates a score corresponding to the degree of recognition of the virtual object in the captured image for each virtual camera for each setting interval, and adds the calculated score (S240). Thereafter, the safety evaluation unit 150 evaluates security utilization degree for each virtual camera or security security degree for the movement route using the summed scores calculated for each virtual camera (S250).

Here, the degree of recognition of the virtual object may correspond to the degree of recognition of the region of interest predetermined for the virtual object. That is, if the virtual object is an avatar, the region of interest is the head region, and if the virtual object is a vehicle, the region of interest may correspond to the license plate region.

In this embodiment, the object of the virtual object can be changed. According to this, the efficiency of the camera can be determined individually for each object of the virtual object. The monitoring efficiency of the camera may vary depending on the position (height, size, etc.) of the recognition target site. The human head is generally located at a relatively higher position than the license plate area. That is, depending on how the line of sight and the angle of view of the camera are set in the space, the head portion and the license plate portion of the person can be recognized effectively, or only one of them can be recognized effectively, .

Hereinafter, for convenience of explanation, a case in which a virtual object is an avatar will be exemplified. In this embodiment, the avatar is moved to a specific interval (ex, 1m) according to the movement route, and the degree of recognition of the avatar by the virtual camera is scored.

1m 2m 3m 4m 5m Sum Virtual Camera 1 30 points 40 points 50 points 30 points 20 points 170 points Virtual camera 2 - 10 points 20 points 10 points - 40 points Virtual Camera 3 10 points 30 points 10 points - - 50 points

Table 2 shows the results obtained by moving the avatar at intervals of 1 m for one arbitrary moving route while obtaining scores corresponding to the recognition degree for each moving position of the avatar from the surrounding virtual cameras 1 to 3. The result of summation according to the movement interval shows that the virtual camera 1 has the highest score. Here, the surveillance utilization of each camera is proportional to the summed score. That is, the virtual camera 1 has the highest monitoring utilization for the random movement path, and the virtual cameras 2 and 3 have relatively low monitoring utilization. From these results, in the case of two security cameras corresponding to the virtual cameras 2 and 3, it is possible to induce resetting of the installation position, direction or angle.

In the example of Table 2, for the sake of convenience of explanation, the monitoring utilization of the virtual camera in the vicinity of one moving route is evaluated. Of course, the monitoring utilization of the virtual camera can be evaluated in the same manner for all the set travel paths. Further, since the total sum of the recognition scores of the virtual cameras 1, 2, and 3 on the movement path indicates the security degree of security for the movement route, the doctor for relocation, supplementation, and reduction A determination can be made.

Hereinafter, the process of recognizing or scoring a virtual object in step S240 of FIG. 2 will be described in detail. In the embodiment of the present invention, the degree of recognition of a virtual object corresponds to the imaging size in units of pixels for the region of interest recognized in the captured image. That is, the recognition degree of the avatar corresponds to the imaging size per pixel for the head part recognized by the virtual camera. At this time, at least one of the horizontal size and the vertical size of the head part can be utilized.

Fig. 7 shows an example of photographing of an avatar by a virtual camera in the method of Fig. The CCD size, CCD pixel, and focal length of the security camera corresponding to this virtual camera are already known. Also, in the three-dimensional modeling environment, the angle between the surveillance direction line vector a of the virtual camera and the avatar face front surface vector b can be calculated, and the distance c from the center of the lens to the center of the face (head) can be calculated. And, since it is generally known to require 80 pixels per foot to identify a stranger, 2.625 pixels per centimeter is required. In the case of underground car parks, a lot of cameras with a focus lens of 3.6 ~ 8mm are often used.

In the example of FIG. 7, when the head of the criminal actor's avatar 10 m away is observed through a 1/3 "CCD camera having a focal length of 8 mm and 1,080 vertical pixels, the height of the face on the CCD = (focal distance x actual face height ) / Distance c '= (8 mm x 213.09 mm) / 10000 mm = 0.17 mm.

Here, since the vertical height of the CCD is 5.0 mm in calculation, when the number of vertical pixels within 5 mm is 1,080, the number of vertical pixels within 0.17 mm is (1,080 pixels x 0.17 mm) / 5 mm = about 36 pixels. That is, the vertical size in pixel units is converted to 36. Of course, the width of the face (width) can also be converted in pixels in a similar way.

The horizontal size of the face is influenced by the angle between the gaze of the virtual camera (direction vector a) and the direction of movement of the avatar (vector b) on the ground surface, and the vertical size of the face is the gaze direction of the virtual camera The vector a) and the direction of movement of the avatar (vector b), respectively, on a plane perpendicular to the above-mentioned ground surface.

Accordingly, the vertical size of the pixel unit is converted into 36 pixels. If the tilt angle is 45 degrees, cos45 = 1/2 is reflected in 36 pixels, and the vertical size is corrected to 18 pixels. If the horizontal size of the pixel is converted to 30 pixels and the angle between the vectors a and b is observed at 45 degrees, the horizontal size is corrected to 15 pixels using the same method.

Here, in the embodiment of the present invention, a score corresponding to the degree of recognition of the head portion of the avatar can be calculated, which corresponds to the ratio of the imaging size to the reference size of the pixel unit, . In addition, the pixel size itself (the corrected horizontal size and the corrected vertical size product) may be used for monitoring safety evaluation without additional conversion of the score.

The score conversion for horizontal size and vertical size utilizes the fact that 2.625 pixels per 1cm (number of pixels per 1cm) is required for human identification.

The vertical size of the avatar head is 21.309 cm, and it requires 21.309 × 2.625 = 56 pixels in total to recognize it. However, since the vertical size actually recognized in Fig. 7 is 18 pixels, if 56 pixels are 100 points, 18 pixels become 32 points (= (18/52) x 100). With this same principle, the horizontal size of the avatar head is also converted to the corresponding score using the actual horizontal size of 15.306 cm.

In the embodiment of the present invention, the average value (or the multiplied value) of the score of the horizontal size and the score of the vertical size obtained by photographing the avatar can be used for the monitoring safety evaluation. In the case of Table 2, as a result of acquiring the score corresponding to the degree of recognition of the head part from the surrounding virtual cameras 1 to 3 while moving the avatar at intervals of 1 m, the score of each cell is calculated for each avatar moving position It corresponds to the average of the horizontal size and the vertical size of the recognized head part. For example, in Table 2, 30 points of the avatar obtained from the virtual camera 1 when the avatar moves 1m may be an average value of 28 points of the horizontal size and 32 points of the vertical size.

Of course, in the embodiment of the present invention, the multiplication of the number of pixels of the horizontal size and the number of the pixels of the vertical size can directly be used for the monitoring safety evaluation without converting the scores.

Hereinafter, a method for evaluating the monitoring safety degree of the floor forming the object space will be described, unlike the first embodiment. This is a method of evaluating the monitoring safety by the floor unit area of the indoor space.

FIG. 8 is a block diagram of an apparatus for evaluating security security of a target space according to a second embodiment of the present invention, and FIG. 9 is a flowchart of a method of evaluating security security using FIG. 8 and 9, an apparatus 200 for assessing security of safety according to a second embodiment of the present invention includes a three-dimensional modeling unit 210, a region dividing unit 220, an image obtaining unit 230, A conversion unit 240, and a safety evaluation unit 250.

First, the three-dimensional modeling unit 210 models a three-dimensional virtual space in which a plurality of virtual cameras are arranged corresponding to the object space (S910). The step S210 refers to the three-dimensional modeling process of the first embodiment.

Next, the area dividing unit 220 divides the 3D virtual space into a plurality of sub-areas (S920). This is a process of dividing a three-dimensional virtual space corresponding to a target space into a plurality of sub-zones, and more particularly, dividing the entire bottoms constituting the indoor space into grid shapes by unit area. For example, the entire floor can be divided into a plurality of forward grids of 1.5 m by 1.5 m in size.

The image acquiring unit 230 arranges virtual objects in the subareas and acquires the captured images from the plurality of virtual cameras, respectively (S930). Here, the virtual object disposed in the sub-zone has a predetermined size, and its type may correspond to an avatar, a vehicle (or license plate), or a virtual surface (virtual surface that is perpendicular to the bottom surface or the bottom surface) have. You can analyze the monitoring safety of people, license plates, and floors for each sub-zone depending on which object the virtual object is used for.

Thereafter, the score conversion unit 240 calculates a score corresponding to the degree of recognition of the virtual object in the sensed image with respect to each of the plurality of cameras, for each of the sub-zones, and sums each of the plurality of virtual cameras S940).

 Here, the degree of recognition of the virtual object corresponds to the degree of recognition of the head portion of the avatar, which is the predetermined region of interest, with respect to the virtual object, or the license plate portion of the vehicle, respectively. Of course, when a virtual surface is used as a virtual object, it means the degree of recognition of the camera with respect to the virtual surface itself.

In step S940, for example, the avatars are arranged in the first sub-zone, and then the degree of recognition of the avatars in all surrounding virtual cameras is determined and the results are summed. After arranging the avatars in the second sub-zone, the degree of awareness of the avatars is determined from all surrounding virtual cameras in the same manner, and the results are summed.

Here, the score corresponding to the recognition degree may correspond to a score corresponding to the number of pixels constituting the region of interest recognized in the captured image. That is, similar to the first embodiment, the number of reference pixels known to the region of interest is set to 100 points, and the score is converted using the ratio of the number of recognized pixels of each virtual camera It can be added by camera. Of course, you can also use the number of pixels itself as a score in addition to converting the score.

In the second embodiment, the head shape of the avatar may be a spherical shape having a spherical shape, unlike the first embodiment. If the avatar is not a spherical body but has the same shape as that of the first embodiment, the avatar may be rotated in eight directions to acquire a camera-specific shot image to evaluate the surveillance performance.

Thereafter, the safety evaluation unit 250 evaluates the security security degree for each subarea using the summed scores calculated for each subarea (S950). SECOND EMBODIMENT The security degree of security is also proportional to the sum of the above scores.

In some sub-zones, recognition may be very low or close to zero, and in other sub-zones the camera may be much more aware than other areas. If it is judged that the security safety degree is low, it is possible to change the angle of the camera or induce additional placement of the camera.

The safety evaluation unit 250 may display each of the sub-zones in a color corresponding to the evaluated security degree of security. That is, spatial correlation between the security camera and the floor area in the indoor space is analyzed in 3D to score the safety degree of safety, and then the color can be expressed differently according to the score according to the floor section.

FIG. 10 is a diagram illustrating the monitoring safety degree of each area (floor) analyzed through the method of FIG. 8 by color legend. The monitoring safety score according to the summed score can be quantified to a value within the range of 0 to 100, and the color gradually changes according to the quantified safety range. In FIG. 10, the first column shows the case where a virtual plane is placed in each sub-zone, not the object, the second column indicates a case where a license plate is arranged in a sub-zone, and the third column indicates a security safety evaluation Represents a color legend corresponding to the result. In each of these three cases, an example of the color indication for the evaluation result of security safety for each will be described later.

The second embodiment will be described in more detail as follows. First, the analysis result of security security of the floor itself is explained. The monitoring safety degree of the floor itself is determined by setting the floor surface itself constituting the virtual space or a virtual surface perpendicular to the floor surface as a virtual object for each subarea so that the security degree of security of the floor for each subarea . In other words, the monitoring safety of the floor itself does not place any special objects on the floor, but rather the monitoring safety of the floor, assuming that the floor itself is facing vertically.

FIG. 11 shows a result of evaluating security security of the floor by using the method of FIG. 8. FIG. The color of the safety of the floor itself refers to the color legend of the first column of Fig. For areas where the color is not displayed, it corresponds to a shaded area where camera monitoring can not be performed. As a result of evaluating the safety of the floor itself, it was confirmed that there is a difference in the monitoring safety of the floor according to the vertical angle of the camera (tilting angle of the camera to the ground). Especially, And the monitoring safety was relatively low.

Fig. 12 shows a conceptual diagram of photographing license plates in the method of Fig. 8. Fig. The size of the license plate is 335 mm in width and 155 mm in length, and the length from the bottom to the center of the license plate is 650 mm. The license plate object is placed in each subarea, and the sum of the points that the surrounding camera observes is calculated as the security security score of that subarea.

FIG. 13 shows the result of evaluating the security degree of security for each license plate by using the method of FIG. Refer to the color legend in the second column of FIG. 10 for the color of the zone safety for the license plate. The result of FIG. 13 shows a result comparable to the result of analyzing the bottom surface itself of FIG. 12 as an object.

Fig. 14 shows a conceptual diagram for photographing an avatar in the method of Fig. 8. Fig. The avatar has a height of 1724.43mm, a head vertical length of 213.09mm, and a head width of 153.06mm. Place the avatar in the sub-zone and calculate the sum of the observation points of the surrounding cameras and make it the floor surveillance safety of the corresponding sub-zone.

FIG. 15 shows a result of evaluating the security security degree of each zone for the avatar using the method of FIG. As a result of the security safety analysis by these person surveillance, it can be seen that the subareas for recognizing persons are very limited. This indicates that human face recognition is very difficult if it is avoided in front of the camera.

Fig. 16 shows the results of Figs. 11, 13, and 15 in an integrated fashion. In this embodiment, as shown in FIG. 16, security security degrees by floor, license plate, and avatar can be integrated. Since the floor security itself and the license plate security are similar to each other, only the license plate and avatar security security can be integrated.

According to the security security evaluation method and apparatus for the target space according to the present invention, the monitoring security degree of the target space in which the security cameras are installed can be evaluated through simulation in the 3D virtual space. The present invention minimizes the rectangular area of the detection area through the simulation of the arrangement of the CCTV camera fused with the spatial information.

In addition, the present invention can effectively evaluate the monitoring efficiency according to the installation position and angle of the CCTV camera through a software approach, and can increase the utilization of the CCTV camera. Furthermore, there is an advantage that it is possible to induce the most effective CCTV camera placement or relocation to the target space, and to establish a more secure and reliable security environment.

While the present invention has been described with reference to exemplary embodiments, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

100,200: security safety evaluation device for the target space
110, 210: Three-dimensional modeling unit 120: Path setting unit
130, 230: Image acquisition unit 140, 240: Score conversion unit
150, 250: Safety evaluation part 220:

Claims (16)

Modeling a three-dimensional virtual space in which a plurality of virtual cameras are arranged corresponding to a target space;
Setting a movement path for a virtual object having a predetermined size in the three-dimensional virtual space;
Acquiring a captured image from the plurality of virtual cameras while moving the virtual object along a moving path at a set interval;
Calculating and adding a score corresponding to a degree of recognition of a predetermined region of interest for the virtual object in the captured image for each virtual camera for each virtual camera for each virtual camera; And
And evaluating the security utilization degree for each virtual camera or the security security degree for the movement route by using the summed scores calculated for each virtual camera,
The degree of recognition of the region of interest,
Corresponding to an image pickup size in units of pixels for the region of interest recognized in the captured image. The method according to claim 1,
Wherein the security security degree is proportional to the summed score. The method of claim 2,
Wherein if the virtual object is an avatar, the region of interest is a head region, and if the virtual object is a vehicle, the region of interest is a license plate region. delete The method according to claim 1,
The score corresponding to the recognition degree may be,
Wherein the score is a score corresponding to a ratio of the imaging size to a reference size of a pixel unit known to the region of interest. The method of claim 5,
Wherein the plurality of virtual cameras include:
Dimensional virtual space corresponding to positions and angles of a plurality of security cameras installed in the object space,
The score corresponding to the recognition degree may be,
Wherein the virtual camera is an average or product of a score corresponding to a horizontal size and a vertical size corresponding to a perceived pixel size of the region of interest recognized in the captured image, A value obtained by correcting the recognized vertical size on the plane of the virtual camera and the moving direction of the virtual object on a plane perpendicular to the surface of the ground, A method for evaluating security safety of a space in which the value corrected through the angle is used for conversion of points. Modeling a three-dimensional virtual space in which a plurality of virtual cameras are arranged corresponding to a target space;
Dividing the three-dimensional virtual space into a plurality of sub-zones;
Arranging a virtual object having a predetermined size in each of the subareas to acquire a captured image from each of the plurality of virtual cameras;
Calculating a score corresponding to a degree of recognition of a predetermined region of interest with respect to the virtual object in the sensed image for each of the plurality of cameras and summing the scores for each of the plurality of virtual cameras; And
And evaluating security security for each subarea using the summed scores calculated for each subarea,
The score corresponding to the recognition degree may be,
Wherein the score is a score corresponding to the number of pixels constituting the region of interest recognized in the captured image. The method of claim 7,
The security security degree is proportional to the summed score,
And displaying each of the sub-zones in a color corresponding to the evaluated security security degree. The method according to claim 7 or 8,
The virtual object is an avatar or a vehicle,
The degree of recognition of the virtual object may be,
Wherein the avatar's head part or the license plate part of the vehicle, which is a pre-set interest area for the virtual object, is recognized. The method of claim 9,
A floor surface constituting the virtual space or a virtual surface perpendicular to the floor surface is set as the virtual object in each subarea and a virtual space is formed in a target space for further evaluating security security degree of the floor surface for each subarea Security Assessment Methods for. A three-dimensional modeling unit for modeling a three-dimensional virtual space in which a plurality of virtual cameras are arranged corresponding to a target space;
A path setting unit for setting a movement path for a virtual object having a predetermined size in the three-dimensional virtual space;
An image acquiring unit for acquiring a captured image from each of the plurality of virtual cameras while moving the virtual object along a moving path at a set interval;
A point conversion unit for calculating and adding a score corresponding to a degree of recognition of a predetermined region of interest for the virtual object in the captured image for each virtual camera for each virtual camera for each virtual camera; And
And a safety evaluation unit for evaluating surveillance utilization for each virtual camera or security security for the movement route using the summed scores calculated for each virtual camera,
The degree of recognition of the region of interest,
Corresponding to an imaging size of a pixel unit for the region of interest recognized in the captured image. The method of claim 11,
The security security degree is proportional to the summed score,
Wherein if the virtual object is an avatar, the region of interest is a head region, and if the virtual object is a vehicle, the region of interest is a license plate region. The method of claim 12,
The score corresponding to the recognition degree may be,
Wherein the score is a score corresponding to a ratio of the imaging size to a reference size of a pixel unit known to the region of interest. 14. The method of claim 13,
Wherein the plurality of virtual cameras include:
Dimensional virtual space corresponding to positions and angles of a plurality of security cameras installed in the object space,
The score corresponding to the recognition degree may be,
Wherein the virtual camera is an average or product of a score corresponding to a horizontal size and a vertical size corresponding to a perceived pixel size of the region of interest recognized in the captured image, A value obtained by correcting the recognized vertical size on the plane of the virtual camera and the moving direction of the virtual object on a plane perpendicular to the surface of the ground, A security safety evaluation system for a space in which a value corrected through an angle is used for score conversion. A three-dimensional modeling unit for modeling a three-dimensional virtual space in which a plurality of virtual cameras are arranged corresponding to a target space;
A region dividing unit dividing the three-dimensional virtual space into a plurality of sub-regions;
An image acquiring unit for arranging a virtual object having a predetermined size in each of the subareas to acquire a captured image from the plurality of virtual cameras;
For each of the sub-zones, a score corresponding to a degree of recognition of a predetermined region of interest of the virtual object in the captured image with respect to the plurality of cameras, and sums each of the plurality of cameras with respect to each of the plurality of virtual cameras; And
And a safety evaluation unit for evaluating a security security degree for each subarea using the summed scores calculated for each subarea,
The score corresponding to the recognition degree may be,
Wherein the score is a score corresponding to the number of pixels constituting the region of interest recognized in the captured image. 16. The method of claim 15,
The security security degree is proportional to the summed score,
The safety evaluation unit includes:
And displays each of the sub-zones in a color corresponding to the evaluated security degree of safety.

Images