KR20190116154A - Calibration system and method for camera using object identification - Google Patents

Abstract

The present invention relates to a camera calibration system and method using object identification, and more particularly, to a camera calibration system and method using object identification which can perform calibration by sending a sensed image to a control center for a predetermined period after installing a camera and a DVR, analyzing the object of the image at the control center and analyzing a state where the same object is displayed on various regions of a screen. The present invention can support performing automatically calibration during an initial set period without the complicated work of a skilled person to support the user′s convenience for calibration and to use a variety of low-cost cameras in a system configuration. In addition, it is possible to reduce the operation cost of the skilled person and to greatly reduce the cost of constructing a system composed of an image recording device. The camera calibration system includes a video recording device and a control server.

Description

Camera calibration system and method for object identification {Calibration system and method for camera using object identification}

The present invention relates to a system and method for calibrating a camera through object identification. More particularly, the present invention relates to a camera calibration system and a method, and more particularly, sends a sensed image to a control center for a predetermined period after installing a camera and a DVR, and analyzes the object of the image in the control center to display the same object. The present invention relates to a camera calibration system and method through object identification for performing a calibration by analyzing a state displayed in various areas of the apparatus.

With the development of various object tracking algorithms for tracking objects appearing in images captured by cameras and the performance of the cameras, the development of object monitoring systems using them has become more prominent.

Such an object monitoring system is applied to various fields including a surveillance system, and is used to detect an object desired by a user in an image and to locate a corresponding object.

The existing object monitoring system acquires an image of an object located in a real three-dimensional space by using a camera. In this case, the objects appearing in the image may not only be different in size depending on the distance from the camera, but may be the same object. Since it can have different sizes, in order to accurately identify the desired object in the image and locate the object, spatial mapping through calibration between the virtual space where the camera is photographed and the real space must be preceded.

Therefore, to operate the existing object monitoring system, the initial parameters (height, angle, FOV, camera position, camera placement altitude, etc.) of the camera are input and the virtual space displayed in the image captured by the camera and the actual target object are taken. It is necessary to calibrate the 3D space so that the objects located in the 3D space correspond exactly to the objects appearing in the image.

However, in the conventional calibration method, in consideration of various parameters based on the specification of the camera, the user directly sets the size value of the object and the depth value of the space corresponding to the area where the known object is located in the 3D space. This calibration process is performed by a skilled person by exposing an object whose size is known in the actual three-dimensional space to the camera and then matching the image information of the camera with the size information of each position in the object to obtain parameters. As a result, the mapping process between the virtual space and the physical space is not only very complicated and inefficient, but also greatly increases the system complexity.

In addition, the existing calibration method has different sizes of objects at positions corresponding to a plurality of different positions in the real space, respectively, so that the size information and the image of the object are mutually matched to all regions in the real space to be monitored. Although the calibration should be performed in a manner that makes it possible to do this, it is not only difficult to perform such a task but also needs to be performed again when the area to be monitored changes.

Moreover, such a calibration method is difficult to set up by a user who is not an experienced expert due to a complicated working method, and there is a problem that an initial installation cost burden for a system construction is excessive due to excessive work cost of an expert.

In addition, although various low-cost cameras are currently being provided for monitoring purposes, and the system construction cost can be lowered, users cannot perform such calibration, so the system must be limited to cameras of specific companies during the construction of the monitoring system. Not only does this cause inefficiency, but the calibration required to rely on specialists creates labor costs in the process of operating the system, increasing the cost of operating the system.

In addition, when using the low-cost camera when building a security system, there is a problem that it is difficult to cope with the security threats caused by hacking, there is a problem that the system reliability is not guaranteed and must be relied on by a skilled person.

Korea Patent Registration No. 10-0969576

The present invention enables to perform the camera calibration required for intelligent video security after the initial DVR and camera installation through the collection of own sample screens without the need for expensive technicians, and to lower the cost, and also without the automatic setup of these skilled personnel. The user can check the progress of the initial setting by explicitly presenting the situation in which the calibration process is performed, and if necessary, actively participate in the initial setting to improve the convenience of calibration by improving the quality and reducing the setting time. Its purpose is.

In addition, the present invention, after setting the access authority for the image transmitted to the server of the control center for image analysis during the setup process for the calibration of the camera, and transfers to the user, and then access information on the user image to blockchain technology Its purpose is to support safety and security through access management separate from the control center.

According to an embodiment of the present invention, a camera calibration system through object identification in which camera calibration is automatically performed during an initial setting period according to the installation of an image recording apparatus and a camera unit for monitoring a surveillance area of a user may include: an object received from a connected camera unit; An image recording device for transmitting the existing image and the camera information related to the specification of the camera unit through a communication network and an image received from the image recording device, and an object existing in the received image is identified according to a preset deep learning algorithm. Collects and stores object information including a type, a location, and a size of the camera; and stores state information about a calculation-related progress state of the calibration information obtained through the object information and the camera information according to a preset information collection criterion for calibration of the camera unit. Generate and award And a control server for transmitting the status information to the video recording apparatus whenever the status information is updated according to the collection of object information, wherein the video recording apparatus corresponds to the camera unit when receiving the status information. The status information may be output through the connected monitor so that the user may check the real-time progress of the calibration.

As an example related to the present invention, the image recording apparatus may detect an object in the image according to a preset image analysis algorithm, and transmit the object presence image in which the object is detected to the control server.

As an example related to the present disclosure, the control server may include an image of a monitored object corresponding to the specific object type based on a standard object preset in correspondence with the specific object type of the object identified in the image through the deep learning algorithm. The object information is generated by calculating the size of each location, and a plurality of object information related to the specific object type obtained from the plurality of images received from the camera unit are averaged for each location, and the space according to the change in the average size of each object location. And generating the information, and updating the spatial information by collecting the object information until the deviation between the spatial information of each of the different object types is equal to or less than a preset reference value according to the information collection criteria.

As an example related to the present invention, the control server compares the size of the monitored object at a specific position in the image obtained based on the standard object with an average size of the same object type at the same position as the specific position. It may be characterized by excluding the object information that the difference occurs more than a predetermined reference value.

As an example related to the present invention, when the control server satisfies the information collection criterion, the control server collects and averages the spatial information for each object type and generates the averaged spatial information, and then relates the viewpoint of the camera unit based on the camera information and the reference spatial information. The calibration information including the plurality of attribute-specific parameters may be calculated.

As an example related to the present invention, the control server calculates the state information on the basis of the deviation between the spatial information of each of the different object types, and the analysis by the deep learning algorithm according to the collection of the object information The status information may be updated whenever the deviation is changed.

As an example related to the present invention, the control server transmits the identification information and the state information of the camera unit to a user terminal connected to the video recording apparatus through a wired or wireless connection, and thus the calibration state of the camera unit is changed through the user terminal. It may be characterized in that the display.

As an example related to the present invention, the control server identifies a user terminal preset in correspondence with the video recording apparatus, and generates a message for inducing user participation according to the progress of calibration according to the status information. It may be characterized in that the transmission to the terminal.

As an example related to the present invention, the control server generates the state information for each of the one or more camera units connected to the image recording apparatus, and the image information is displayed in the state information and the identification information of the camera unit whenever the state information is updated. And the image recording apparatus identifies any one of the one or more camera units based on the state information received from the control server and identification information of the camera unit, and monitors the camera unit corresponding to the identified camera unit. It may be characterized by outputting the status information.

As an example related to the present invention, the control server generates and transmits transfer information for transferring an access right to an image of the camera unit to a user corresponding to the image recording apparatus when the calibration according to the state information is completed. The video recording apparatus generates and stores the authority information for setting the access authority to a user who is preset based on the transfer information, and is connected to one or more other video recording apparatuses in a P2P manner to provide transaction information related to the access history of the authority information. At the time of generation, the transaction information may be shared with the other video recording apparatus and stored in a block chain manner.

According to an embodiment of the present invention, a camera calibration method through object identification in which camera calibration is automatically performed during an initial setting period according to the installation of an image recording apparatus and a camera unit for monitoring a surveillance area of a user, wherein the image recording apparatus includes a camera unit. Transmitting an object existence image received from the camera and spec-related camera information of the camera unit through a communication network, and a control server communicating with the image recording apparatus through a communication network stores the image received from the image recording apparatus, Collecting and storing object information including the type, position, and size of the object by identifying an object existing in the image according to a preset deep learning algorithm; and controlling the camera to a preset information collection criterion for calibration of the camera unit. According to the object information and camera information Generating state information on a progress state related to the calculation of the obtained calibration information, and transmitting the state information to the video recording apparatus every time the state information is updated according to the collection of the object information; And receiving the status information so that the user can check the real-time progress of the calibration through a monitor connected to the image recording apparatus corresponding to the camera unit upon reception.

According to the present invention, after installing a video recording apparatus such as a camera unit and a DVR, a video of a camera unit in which a monitoring target object is detected by a control server for a predetermined period is received and collected from the video recording apparatus, and the object of the image is analyzed for each object type. As the object is displayed in various areas of the screen, the three-dimensional space in the image mapped to the real space related to the surveillance area of the camera unit is calculated for each object type, and then compared with each other for each object type. It is possible to calculate the space, and based on this, it is possible to automatically calculate the calibration information of the camera unit according to the reference space, and to automatically calculate the calibration information of the camera unit. Number of calibrations during In addition to the user's convenience of calibration, various low-cost cameras can be used to configure the system, and the operation cost of the skilled person can be greatly reduced, thereby greatly reducing the cost of constructing the system consisting of the video recording device. It works.

In addition, the present invention can be displayed on the monitor connected to the video recording device for the information on the progress of the calibration for each camera connected to the video recording device so that the user can easily check the calibration progress of the user In addition to greatly improving the ease of building the system, it is possible to increase the user's satisfaction through quick calibration by supporting the user's participation to shorten the initial setting period in order to speed up the calibration progress of the slow camera. There is.

In addition, the present invention transfers the access authority for the image transmitted to the control server for image analysis during the initial setting process for the calibration of the camera unit to the user after the completion of the calibration, and then the access information for the user image blockchain technology By using it to manage, there is an effect of supporting the improvement of safety through accessibility management (blockchain management using P2P) separated from the control center.

1 is a block diagram of a camera calibration system through object identification according to an embodiment of the present invention.
2 to 5 are diagrams illustrating an operation of a camera calibration system through object identification according to an embodiment of the present invention.
6 is an exemplary blockchain-based operation of a camera calibration system through object identification according to an embodiment of the present invention.
7 is a flowchart illustrating a camera calibration method through object identification according to an embodiment of the present invention.

Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings.

The present invention specializes in camera calibration required for intelligent video security at the time of initial installation of a video recording device such as a DVR (Digital Video Recorder) and a camera for the construction of a video-based security system for a user's surveillance area. In order to improve the problem that the system construction cost is increased due to the setting of a skilled worker, the calibration is automatically performed by collecting a sample screen without the need for a costly expert, and the cost is lowered. Explicitly present the user with an autocalibration process without the involvement of these skilled personnel so that the user can check the progress of the initial setup and actively participate in the initial setup if necessary to increase quality and reduce setup time. Support.

The camera calibration for image analysis refers to setting of information for calculating a transformation matrix between a 3D virtual space and a real space in an image in order to identify the size and position of an object in the image.

To this end, basically, calibration parameters, such as a camera's viewpoint height, a camera's viewing angle, and camera lens distortion information, must be set. The present invention can automatically distinguish the shape and the type of an object from a still image. AI image analysis (e.g., machine learning, deep learning, etc.) is a still image, such as deep learning, that contains objects in the image that are collected as camera images instead of a manual and complicated calibration procedure. The camera may be automatically calibrated by extracting the type and shape of the object by analyzing the analysis method and repeating the process of obtaining the type and position of the object from the sample image of the camera for setting the calibration of the camera.

With reference to the above-described configuration, a detailed embodiment of the present invention will be described below with reference to the drawings.

First, FIG. 1 is a configuration diagram of a camera calibration system through object identification according to an exemplary embodiment of the present invention. As shown in FIG. 1, the image recording apparatus 10 connected to at least one camera unit 11 is connected through a communication network. It may include a control server 100 in communication with the video recording device 10.

In this case, the camera unit 11 may be configured as an IP (Internet Protocol) camera, and the video recording apparatus 10 may be configured as a digital video recorder (DVR) or a network video recorder (NVR).

In addition, a variety of wired and wireless communication schemes that are widely known may be applied to a communication network for communication between the image recording apparatus 10 and the control server 100.

Referring to the configuration of FIG. 1, a detailed operation configuration of the camera calibration system through object identification will be described. As shown in FIG. 1, an image connected to one or more camera units 11 installed for the security of a surveillance region designated by a user is illustrated. Initial installation of the recording device 10 requires the calibration of each of the one or more camera units 11 connected to the image recording device 10.

To this end, the image recording apparatus 10 detects an object by analyzing an image received from the camera unit 11 so that the calibration is automatically performed during an initial setting period for calibration of the camera unit 11. In addition, only the object existence image from which the object is detected may be transmitted to the control server 100.

In addition, the control server 100 stores the image received from the image recording device 10, and identifies the type, location, size, etc. of the object by identifying the object present in the received image according to a preset deep learning algorithm It can create object information to include and collect and store the object information.

In addition, the control server 100 generates the state information on the progress related to the calculation of the calibration information obtained through the object information and the camera information according to a preset information collection criteria for the calibration of the camera unit 11 and The status information may be transmitted to the video recording apparatus 10 whenever the status information is updated according to the collection of object information.

2 to 5, first, as shown in FIG. 2, the control server 100 receives the one or more camera units received from the image recording apparatus 10 through the deep learning algorithm. The object information may be generated by calculating a size for each position in the image of the monitored object corresponding to the specific object type identified in the image of the specific camera unit 11 in FIG. 11.

In the description of FIGS. 2 to 5, a calibration process for one specific camera unit 11 will be described.

In this case, the control server 100 may be set in advance a standard object by object type (for example, a person, an animal, a car, etc.), and the size of the standard object may be preset.

Accordingly, the control server 100 may calculate (or convert) the size according to the position in the image of the monitored object identified in the image based on the size of the standard object.

In this case, the control server 100 continuously receives sample images from which the object is detected in the specific camera unit 11 from the image recording apparatus 10 for automatic calibration of the specific camera unit 11, and the sample. After generating the object information for each image, the object information may be stored in the DB 101 by matching the identification information of the specific camera unit 11.

Here, the image recording apparatus 10 may transmit identification information of the specific camera unit 11 associated with the corresponding image in the image and transmit the image, and the object information also includes identification information of the specific camera unit 11. It may include.

In addition, the image recording apparatus 10 includes camera information related to specifications of the camera including identification information of the camera unit 11 for each of the one or more camera units 11 connected to the image recording apparatus 10. The image recording apparatus 10 may transmit the camera information to the control server 100 when the initial calibration is set.

Accordingly, the control server 100 may match the object information generated based on the image of the specific camera unit 11 with the camera information of the specific camera unit 11 and store it in the DB 101.

In addition, the control server 100 averages a plurality of object information related to a specific object type obtained from a plurality of images received from the specific camera unit 11 through the image recording apparatus 10 for each location, and averages each location. It is possible to generate spatial information according to the change of the average size.

For example, the control server 100 collects and averages object information having the same position among a plurality of object information having the same object type collected and stored corresponding to the specific camera unit 11 as shown, and averages the specific location. An average size corresponding to a specific object type in E may be calculated.

Accordingly, the control server 100 can know the average size of the object at each of the different positions in the image of the specific camera unit 11 corresponding to the specific object type, and thus the object corresponding to the specific object type Spatial information can be calculated according to the change in average size for each location.

That is, the control server 100 may calculate the distance between different positions according to the size change of each position of the object, and calculate the spatial arrangement state of the surveillance region appearing in the image based on the distance between the different positions. To generate spatial information.

In addition, as shown in FIG. 3, the control server 100 generates the spatial information in correspondence with another object type different from the specific object type, similarly to a method of generating spatial information generated corresponding to the specific object type. can do.

In this case, in order to accurately calibrate the specific camera unit 11, spatial information must be uniform regardless of the object type. For this purpose, the control server 100 corresponds to a plurality of different object types, respectively. It is possible to compare different spatial information of each other, and to calculate the deviation between a plurality of different spatial information.

In addition, the control server 100 collects object information from the video recording device 10 until the deviation between the spatial information of each different object type is less than or equal to a preset reference value according to the preset information collection criteria. The spatial information may be updated, and when the predetermined information collection criteria are satisfied, the reference spatial information may be calculated to minimize the deviation from the spatial information of each of a plurality of different object types.

In this case, the control server 100 may generate reference space information obtained by averaging the space information for each object type.

In addition, in the above-described configuration, the control server 100 sets the size of the monitored object at a specific position obtained based on the standard object corresponding to a specific object type to the average size of the same object type at the same position as the specific position. Compared with, the error object information having a difference exceeding a predetermined reference value may be excluded from the calculation of the spatial information or the error object information may be deleted.

In this case, the control server 100 may be different in height when the object type is a person due to individual deviation, so that if the measurement is relatively large or small among the various sample images collected from the image recording apparatus 10, By calculating the remaining cases by converting them into standard keys, a relatively accurate size of each object can be measured, and the accuracy of the three-dimensional space in the image corresponding to the real space can be increased based on the continuously received sample image.

In addition, in the above-described configuration, the control server 100 may check the information on the object by providing the user with reference image or object information to be used for calibration if necessary.

For example, the control server 100 may inquire the user of the type or key of a specific object appearing in the image through communication with the video recording apparatus 10, and whether the object appearing in each of the different sample images is the same object. You can also query whether or not.

In addition, the control server 100 may receive the response information generated by the user in response to the query from the video recording device 10 and generate the object information by reflecting the response information.

On the other hand, as shown in Figure 4, the control server 100 is generated corresponding to the camera information and the specific camera unit 11 of the specific camera unit 11 corresponding to the specific camera unit 11 Based on the reference space information, calibration information including a plurality of property-related parameters of a viewpoint of the specific camera unit 11 may be calculated.

That is, the control server 100 calculates a preset calibration-related attribute parameter for mapping the camera's viewpoint to the reference spatial information based on the spec-related camera information of the specific camera unit 11, Calibration information including parameters may be calculated (generated) in correspondence with the camera unit 11.

In this case, the attributes set in advance for the calibration may include a camera's viewpoint height (h), a camera's orientation angle, a camera's field of view (FOV), lens characteristics, a focal length, and the like.

In addition, the control server 100 may identify the camera unit 11 corresponding to the received image when receiving the image from the image recording apparatus 10 when there are a plurality of camera units 11 connected to the image recording apparatus 10. Information may be received together with the image and stored in the DB 101. The calibration information may be calculated in correspondence with each of the plurality of different camera units 11 by individually managing the camera units 11 based on the identification information. Can be.

Through the above-described configuration, the control server 100 installs the camera unit 11 and the image recording apparatus 10 such as a DVR, and then the camera unit 11 in which the monitored object is detected for a predetermined period (initial setting period). ) Is received from the video recording device 10 and collected, the object of the image is analyzed and the object is displayed in various areas of the screen for each object type. After calculating the three-dimensional space in the image mapped to the real space and comparing it with different object types, it is possible to calculate the reference space for the calibration of the camera unit 11, based on the real space and the image as the surveillance area The real object located in the real space by automatically calculating the calibration-related parameters for the viewpoint of the camera unit 11 for the mapping of the three-dimensional space in accordance with the reference space The calibration information for cross-mapping an object on the three-dimensional space displayed on the image of the camera unit 11 can be automatically calculated.

In addition, the control server 100 may transmit the calibration information to the image recording apparatus 10, and the image recording apparatus 10 controls the camera unit 11 corresponding to the calibration information to control the camera unit. The parameter per view-related attribute of (11) may be automatically set according to the calibration information.

Accordingly, the present invention can automatically calculate the calibration information corresponding to the specific camera unit 11 in the control server 100 so that the calibration is automatically performed during the initial setting period without having to go through the complicated work of the skilled person as before. In addition to the user's convenience for calibration, various low-cost cameras can be used to configure the system, as well as to reduce the operation cost of the skilled person, thereby greatly reducing the system construction cost of the video recording device 10. Can support

On the other hand, in the above-described process, when the monitoring area with a low frequency of object appearance on the installation position of the camera unit 11 is the monitoring target area of the camera unit 11 is not enough sample image collected by the control server 100 There may be variable factors such as the time required for the calibration process may be long, and in the case of many exceptional circumstances, an error may occur in the automatic calibration.

Accordingly, a non-expert user has a problem that it is difficult to clearly know the calibration process or quality of the installed image recording apparatus 10 and the camera.

Detailed configuration of the present invention to compensate for this will be described in detail with reference to FIG.

As shown in FIG. 5, the control server 100 generates state information on a progress state related to the calculation of the calibration information obtained through the object information and the camera information, corresponding to a specific camera unit 11, and the object information. Whenever the state information is updated according to the collection of the state information, the state information corresponding to the specific camera unit 11 may be transmitted to the image recording apparatus 10.

In this case, the state information may include identification information of the specific camera unit 11.

For example, the control server 100 calculates spatial information for each object type based on the object information stored in the DB 101 corresponding to the specific camera unit 11, and corresponds to a plurality of different object types, respectively. Based on the degree of deviation between different spatial information of may be calculated the degree of progress for the calculation related progress state of the calibration information.

That is, the control server 100 may calculate the state information having a low progress level when the deviation degree is greatly increased, and continuously update the state information by increasing the progress level as object information is continuously collected and the deviation decreases.

In addition, the control server 100 may transmit the state information every time the state information is updated, so that the user may check the progress of calculating the calibration information in real time.

In addition, the image recording apparatus 10 outputs the state information through the monitor 12 connected to the image recording apparatus 10 corresponding to the specific camera unit 11 corresponding to the state information when the state information is received. Through this, the user may check the calibration progress state of the camera unit 11 corresponding to the corresponding monitor 12 in real time through the monitor 12.

In addition, the control server 100 identifies a user terminal 20 that is set in advance in correspondence with the video recording apparatus 10, and generates a message for inducing user participation according to the progress of calibration according to the status information. Can be transmitted to the user terminal 20.

Accordingly, the user can check the calibration progress for each camera unit 11 in real time through the monitor 12 or the user terminal 20, identify the specific camera unit 11 that is slow in calibration, and identify the specific camera. In order to speed up the calibration process of the unit 11, a sample image that appears directly in the surveillance area of the specific camera unit 11 and is detected by the user as an object is continuously collected from the control server 100 so that the calibration process can proceed quickly. Participate in the calibration process.

That is, the control server 100 provides the user with the state information on the calibration progress state through the monitor 12 connected to the image recording apparatus 10 to correspond to the camera 11 having a slow progression of calibration. In order to speed up the calibration process, user participation can be encouraged. This can shorten the initial setup period and increase user satisfaction.

In this case, the control server 100 may transmit and provide the state information for each of the camera units 11 to the user terminal 20 preset in correspondence with the user through a communication network, or through the user terminal 20. Status information for each of the camera units 11 may be displayed.

Through the above-described configuration, the present invention allows the user to display information on the progress of the calibration for each camera connected to the image recording apparatus 10 through the monitor 12 connected to the image recording apparatus 10. By making it easy to check the calibration progress, it greatly improves the user's convenience of system construction and encourages the user's participation to shorten the calibration progress of the camera, which slows the calibration progress. Fast calibration can greatly increase user's satisfaction.

Meanwhile, in the above-described configuration, in order to identify and calibrate an object through a sample image, the control server 100 needs to access (access) the user's video recording apparatus 10 to collect images, thereby protecting personal information or access. Complaints of the user may appear due to the burden that the authority exists in the control center operating the control server 100.

In particular, there is a risk of hacking in the case of a low-cost camera or DVR, and there is also anxiety that the administrator of the control center can randomly access the user's video.

Accordingly, only the calibration process is allowed to access the image collection of the control server 100, and when the calibration is completed, the access authority is transferred to the user so that only the user can set the access rights of the camera and the image recording apparatus 10. Security is required to cope with hacking threats by managing access rights afterwards.

To this end, the configuration of the present invention will be described in detail with reference to FIG.

As shown in the drawing, the control server 100 automatically releases the access right of the control server 100 to the image of the user when the calibration of the video recording apparatus 10 is completed, and the access right is removed. The transfer information for transferring may be generated and transmitted to the image recording apparatus 10 associated with the user through a communication network.

To this end, setting information for setting the access right of the user's video recording apparatus 10 to the control server 100 only during the initial setting period is stored in the control server 100, and the setting information is Identification information for each of the one or more camera units 11 connected to the image recording apparatus 10 that requires calibration may be included, and access rights may be set in the setting information by matching the identification information.

Accordingly, the control server 100 determines whether the calibration is completed according to the progress state according to the state information or the generation of the calibration information for each camera unit 11 connected to the image recording apparatus 10, and then the calibration is completed. In response to the camera unit 11, the access authority may be deleted from the setting information, and the transfer information for transferring the access authority to the user may be transmitted to the image recording apparatus 10.

In this case, the control server 100 may transmit the transfer information to the user terminal 20 to inform the user that the access authority for the calibrated camera unit 11 is transferred and the access authority exists only to the user. .

In addition, the image recording device 10, when receiving the transfer information, the authority information to limit the access right to the image of the camera unit 11 corresponding to the identification information of the camera unit 11 included in the transfer information to the user Can be set by itself.

At this time, the image recording apparatus 10 is completed calibration with respect to the other camera unit 11 connected to the image recording apparatus 10 other than the specific camera unit 11 has been completed calibration from the control server 100 When receiving transfer information for the other camera unit 11, the authority information may be updated by setting an access right to the image of the other camera unit 11 in the previously generated authority information.

In addition, the video recording device 10 is to maintain the security from hacking threats that unauthorized users access the user's video by changing the authority information for the user of the calibration is complete image of the camera unit 11 At least one other recording device connected to at least one other recording device in a peer-to-peer manner to generate transaction information related to the access history (or access history) of the authority information. Can be shared and stored in a block chain manner.

That is, the image recording apparatus 10 may generate the transaction information when generating new transaction information on the access history according to the change of the authority information in order to access the image of the camera unit 11 connected to the image recording apparatus 10. Based on the transaction information generated by the video recording device 10 and the transaction information received from the other video recording device 30 through the P2P method with the other video recording device 30, the blockchain method is used. After generating a block hash (Hash of Block) value for the storage of the new transaction information according to can generate and store a block corresponding to the new transaction information including the block hash value.

In addition, the video recording device 10 may transmit the new transaction information to the other video recording device 30 so that the block corresponding to the new transaction information is generated and stored in the other video recording device 30. .

Through this, the video recording apparatus 10 supports the user to check all the access histories of the authority information by storing them in the blockchain, and through this, assist the user to know the case of access or export for an unfair purpose. Can be.

Through this, when another unauthorized user accesses the transaction information to remove his / her unauthorized access history, the transaction information is recorded in a block stored in another video recording device 30 in a blockchain manner. It helps to track hacking easily.

Especially in the case of export, the validity of the exported image can be checked on the basis of block chain through the hash storage information (or block hash value) of the block for exporting the image so that the user can confirm the export and secure the reliability of the exported image. Can be.

In addition, when the change of the authority information is requested, the video recording apparatus 10 may check the integrity through a block hash value within a single block to check whether the change is a legitimate change, and in the case of hacking, automatically change the authority information. have.

Through the above-described configuration, the present invention transfers the access right for the image transmitted to the control server 100 for image analysis during the initial setting process for calibration of the camera unit 11 to the user after the completion of the setting of the calibration, Afterwards, access information for user images can be managed using blockchain technology, which can support security enhancement through accessibility management (blockchain management using P2P) separated from the control center.

On the other hand, in the above-described configuration, the video recording device 10 is configured with a device controller for performing the overall control function of the video recording device 10, the control server 100 is the overall control of the control server 100 A server controller for performing a control function may be configured.

In this case, the device control unit and the server control unit may operate as a control unit, wherein the control unit uses the program and data stored in the storage unit of the video recording device 10 or the control server 100 or the video recording device 10 or The overall control function of the control server 100 is executed. The controller may include a RAM, a ROM, a CPU, a GPU, a bus, and the RAM, a ROM, a CPU, a GPU, and the like may be connected to each other through a bus. The CPU may access the storage unit to perform booting using an operating system (O / S) stored in the storage unit, and various operations described in the present disclosure using various programs, contents, data, etc. stored in the storage unit. Can be performed.

The device controller may perform object detection through an image analysis algorithm, and the server controller may perform various operations described in the present invention through object identification based on the deep learning algorithm.

In addition, each of the image recording apparatus 10 and the control server 100 may be configured with various components such as a user input unit for receiving user input, a communication unit for communicating with an external device through a communication network, a storage unit for storing various data, and the like. Of course.

7 is a camera calibration through object identification in which camera calibration is automatically performed during an initial setting period according to the installation of the image recording apparatus 10 and the camera unit 11 for monitoring a surveillance area of a user according to an exemplary embodiment of the present invention. A flowchart of the method.

As shown, the image recording apparatus 10 transmits the object existence image received from the camera unit 11 connected to the image recording apparatus 10 and camera information related to the specification of the camera unit 11 through a communication network. It may be (S1).

In addition, the control server 100 which communicates with the video recording device 10 through a communication network stores the image received from the video recording device 10, and stores an object present in the received image in a preset deep learning algorithm. According to the identification, the object information including the type, position, and size of the object may be collected and stored (S2).

In addition, the control server 100 generates the state information on the progress related to the calculation of the calibration information obtained through the object information and the camera information according to a preset information collection criteria for the calibration of the camera unit 11 and Each time the state information is updated according to the collection of object information, the state information may be transmitted to the video recording device 10 (S3).

Thereafter, the image recording apparatus 10 corresponds to the camera unit 11 when receiving the state information so that the user can check the real-time progress of the calibration through the monitor 12 connected to the image recording apparatus 10. The state information may be output (S4).

Various devices and components described herein may be implemented by hardware circuitry (eg, CMOS based logic circuitry), firmware, software, or a combination thereof. For example, it may be implemented using transistors, logic gates, and electronic circuits in the form of various electrical structures.

The above description may be modified and modified by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are not intended to limit the technical idea of the present invention but to describe the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. The protection scope of the present invention should be interpreted by the following claims, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of the present invention.

10: video recording device 11: camera unit
12: monitor 100: control server

Claims (1)

In the camera calibration system through object identification that the camera calibration is automatically performed during the initial setting period according to the installation of the image recording device and the camera unit for monitoring the user's surveillance area,
An image recording apparatus for transmitting the object existence image received from the connected camera unit and camera information related to the specification of the camera unit through a communication network; And
Stores the image received from the image recording device, identifies an object existing in the received image according to a preset deep learning algorithm, collects and stores object information including the type, position, and size of the object, and calibrates the camera unit. Generates state information on a progress state related to the calculation of the calibration information obtained through the object information and the camera information according to a preset information collection criterion, and updates the state information according to the collection of the object information to the image recording apparatus. Control server for transmitting the status information
Including;
When the image recording apparatus receives the state information, the image recording apparatus outputs the state information so that the user can check the real-time progress of the calibration through a monitor connected to the image recording apparatus. Camera calibration system.