KR20190086166A - Intelligent video recording apparatus and method - Google Patents

Abstract

Embodiments of the present invention relate to an intelligent video recording device and an intelligent video recording method performed by the same, comprising: an image photographing device for photographing an image in real-time; a sensor for detecting an impact applied to a target to detect an impact amount and an impact direction; a control part for generating an impact pattern on the basis of the impact amount and the impact direction, and determining whether an impact applied on the basis of the impact pattern corresponds to an event for triggering video recording; and a storage part including a first storage part for storing the photographed image and a second storage part for storing the image for a predetermined time on the basis of the event occurrence time when the applied impact is determined as an event. Therefore, the video recording device determines whether an event for triggering an image recording is generated when an impact is being applied to the target equipped with an image recording device, and initiates video recording when it is determined that an event has occurred.

Description

[0001] INTELLIGENT VIDEO RECORDING APPARATUS AND METHOD [0002]

The present invention relates generally to an image recording apparatus and method, and more particularly, to an image recording apparatus and method for determining whether an event triggering an image recording during an impact applied to an object equipped with the image recording apparatus has occurred, The present invention relates to a video recording apparatus and method for starting video recording.

Recently, an image recording apparatus called a black box, which is intended to store images in order to detect an accident scene or an accident occurring in the absence of a driver, is becoming popular. Such a video recording apparatus is configured to judge whether or not a shock has occurred through an impact detection sensor, to store an image of a corresponding time zone when an impact is detected, and to provide a notice of an impact occurrence and an image storage to the user. Accordingly, it is possible to analyze the accident site based on the recorded image when an accident occurs, and also to recognize an accident occurred in the vehicle in the future even when the driver is absent.

However, in the conventional video recording apparatus, regardless of whether or not the impact is necessary for video recording (for example, even when video recording and notification are unnecessary, such as door closing), if the detected impact is equal to or greater than a predetermined threshold value, Started and saved. Since the impact position is not taken into account, the ordinary impact that is not worthwhile as a notification object that occurs repeatedly at a specific position can not be distinguished, which causes waste of storage space.

In addition, another conventional video recording apparatus can additionally provide a notification of the start of video recording. Such a notification does not include information on the impact position, There is an inconvenience to navigate. Furthermore, the frequency of notifications for everyday impact (which does not require video recording) is more frequent than the frequency of specific shocks (which require video recording), which causes the user to experience extreme fatigue for notifications, There is another problem.

Patent Laid-Open Publication No. 10-2016-0099783

There is provided a video recording apparatus and method for determining whether or not an event triggering a video recording has occurred during an impact applied to an object equipped with the video recording apparatus and for starting video recording when it is determined that an event has occurred.

An intelligent video recording method performed by an intelligent video recording apparatus according to an aspect of the present invention includes: detecting an impact applied to an object to detect an impact amount and an impact direction; Generating an impact pattern based on the impact amount and the impact direction; Determining whether the impact is an event triggering a video recording based on the impact pattern; And storing the image for a predetermined time based on the event occurrence time when the impact is determined to be an event.

In one embodiment, the step of generating the impact pattern based on the impact amount and the impact direction includes the steps of: estimating a position of the applied change by applying the amount of change and the direction of the change to the position model; And generating a change pattern based further on the estimated change position.

In one embodiment, the position model is modeled based on a plurality of training samples, and each training sample may include an impact amount, impact direction and impact position on the training impact.

In one embodiment, the step of determining whether the applied impact corresponds to an event may include applying the impact pattern to an event pattern model to determine whether the impact associated with the impact pattern corresponds to an event.

In one embodiment, the event pattern model has been generated based on a plurality of training samples, and each training sample may include an impulse amount, impact direction and impact position of the training impact, and a first label. Here, the first label indicates whether each training impact corresponds to an event.

In one embodiment, the intelligent video recording method may further include inputting the intrinsic behavior information for each user for the target and customizing the event pattern model for each user using the intrinsic behavior information.

In one embodiment, an intelligent video recording method includes: requesting a first input for generating a second label; Generating the second label in response to the first input; Generating first feedback information including an impact amount and an impact direction used for the notification, and a second label; And updating the position model using the first feedback information. Here, the second label indicates whether the estimated impact position is the actual impact position.

In one embodiment, the intelligent video recording method further comprises: requesting a second input to generate actual impulse position information if the second label indicates that the estimated impulse position is not an actual impulse position; And generating actual impulse position information in response to the second input. In this case, the first feedback information is generated so as to further include the actual impact position information.

In one embodiment, the intelligent video recording method further comprises: requesting a third input for generating a third label; Generating the third label in response to the third input; Generating impulse amount and impact direction used in the notification, and second feedback information including the third label; And updating the event pattern model using the second feedback information. Here, the third label indicates whether the previously determined impact is an actual event.

In one embodiment, the intelligent video recording method includes: determining the impact position as an event occurrence position when the impact is determined as an event; And providing a notification of the event occurrence and the location.

In one embodiment, the event notification may include a graphical representation of the location of the occurrence of the event impact in at least some of the features of the object.

In one embodiment, the event notification may further include at least one of a predicted degree of damage due to the event impact and a literal representation of the estimated loss amount.

A computer-readable recording medium according to another aspect of the present invention is a computer-readable storage medium capable of storing program instructions operable by the computer. Here, the program command may be executed by a processor of the computer to perform the intelligent video recording method according to the above-described embodiments.

 According to another aspect of the present invention, there is provided an intelligent video recording apparatus for recording an image captured when an event occurs, the video recording apparatus comprising: A sensor for detecting an impact applied to an object and detecting an impact amount and an impact direction; A control unit for generating an impact pattern based on the impact amount and the impact direction and determining whether an impact applied based on the impact pattern corresponds to an event triggering the image recording; And a storage unit including a first storage unit for storing the photographed image and a second storage unit for storing an image for a predetermined time based on the event occurrence time when the impact is determined as an event.

According to embodiments of the present invention, it is possible to estimate an impact position at an impact applied to an object, and to determine whether an impact applied to the object corresponds to an event requiring image recording, in consideration of the impact position. This enables more accurate event determination. In addition, it is possible to reduce the unnecessary image recording by the impact repeatedly applied at a specific position (for example, door closing of the vehicle), and to use the storage space more efficiently.

Furthermore, the user will be provided with impact notifications only for impacts that need to be notified, thereby freeing them from fatigue for indiscriminate notification.

In addition, since the notification provided to the user includes information on the impact position, the user who receives the notification can reduce the trouble of searching for the damage location due to the impact.

In addition, it is possible to customize the user's behavior by reflecting the unique behavior pattern of each user such as a habit of closing the vehicle door. In addition, it is possible to continuously enhance the determination function by updating the event determination function by generating feedback information related to a previously applied impact.

In addition to being installed in cars such as cars, commercial vehicles such as trucks, motorcycles, motorcycles, high-speed railways, and trains, it is also used in a variety of fields that need to distinguish between daily impacts and special impacts .

The effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned can be clearly understood by those skilled in the art from the description of the claims.

1 is a simplified block diagram of an intelligent video recording apparatus according to an embodiment of the present invention.
2 is a flowchart of an intelligent video recording method according to an embodiment of the present invention.
3 is a conceptual diagram of an intelligent video recording method according to an embodiment of the present invention.
Figure 4 is an exemplary diagram of an impulse location indication, in accordance with an embodiment of the present invention.
Figure 5 is a flow diagram of a process for updating a location model, in accordance with an embodiment of the present invention.
6 is a flowchart of a process of updating an event pattern model according to an embodiment of the present invention.
The figures depict various embodiments of the invention only for purposes of illustration. Those skilled in the art will readily appreciate from the following description that alternative embodiments of the structures and methods described herein may be used without departing from the principles of the invention described herein.

Embodiments will now be described with reference to the accompanying drawings, however, the principles disclosed herein may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. For example, although the vehicle is described as the main embodiment in the object on which the image recording apparatus is mounted, it will be apparent to those skilled in the art that this is merely exemplary and can be mounted on various objects. In the description of the invention, a detailed description of well-known features and techniques may be omitted to avoid unnecessarily obscuring the features of the embodiments.

In this specification, an event is a trigger that starts image storage when a special situation (for example, an impact of an object, a sudden acceleration / deceleration of an object, a rapid acceleration or sudden acceleration, ). Here, situations in which video recording is necessary include conflicts between the perpetrator / victim of the event (for example, legal, mental, temporal, and economic problems), conflicts between insurers and insurers This means that the video recording at the time of event occurrence is likely to be used as an important means for conflict resolution.

On the other hand, the shock caused by the user's general actions such as simply closing the door or closing the trunk does not cause a series of conflicts as described above, so that image recording is unnecessary and is not judged as an event.

In the present invention, when an event or a plurality of different events are detected, image recording and storage are started. In other words, image recording and storage can be started at the time of impact detection or image recording and storage can be started in the event of a special situation (for example, air bag deployment due to defects in the vehicle, vehicle sudden acceleration, etc.) In the description of the event of the present invention, an object to be mounted will be described as a vehicle, and an event will be described as an example of an impact.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

1 is a simplified block diagram of an intelligent video recording apparatus according to an embodiment of the present invention. The intelligent video recording apparatus 100 includes a camera 110, a sensor 130, a controller 150, a first storage unit (not shown) And a storage unit 170 including a first storage unit 171 and a second storage unit 172.

The intelligent video recording apparatus 100 according to embodiments may be entirely hardware, or partially hardware, and partially software. For example, the control unit 150 included in the intelligent video recording apparatus 100 may collectively refer to hardware and related software for processing data of a specific format and contents and / or for exchanging data by an electronic communication method. That is, the term "unit," "module," "device," or "system" as used herein refers to a computer-related entity such as hardware, a combination of hardware and software, or software. A processor, an object, an executable, a thread of execution, a program, and / or a computer, for example, a computer, but is not limited to, a computer. For example, both an application running on a computer and a computer may correspond to a part, module, device or system of the present specification.

The image capturing apparatus 110 is capable of capturing a continuous image projected on the field of view of the image capturing apparatus 110 as one means of capturing an image. In one embodiment, the camera 110 converts the photographed analog video signal into a digital video signal. Accordingly, the image recording apparatus 100 can capture the outside of the vehicle and / or the inside of the vehicle to acquire image data such as color, monochrome, and infrared rays. In some embodiments, the camera 110 detects only the analog video signal, and the conversion process described above can be performed in the control unit 150 below.

In some embodiments, the number of cameras 110 may be plural. In this case, the video recording apparatus 100 may include a plurality of channels connected to the plurality of cameras 110. For example, a front camera for acquiring images of the front of the vehicle may be connected to the first channel, and a rear camera for acquiring images of the rear of the vehicle may be connected to the second channel.

The sensor 130 is a means for sensing an impact. In one embodiment, when the image recording apparatus 100 is installed in a vehicle, the sensor 130 senses an impact occurring during a vehicle running or during a stop (or parking). The sensor 130 detects the impact and detects impact data including impact amount data related to the impact amount and impact direction data related to the impact direction.

The sensor 130 may be one of an IMU (Inertial Measurement Unit) sensor, a gravity sensor, a motion sensor, an acceleration sensor, a gyro sensor, or a combination thereof. However, And may be implemented using various types of sensors capable of detecting the impact direction.

The control unit 150 determines whether or not an event triggering the image recording is generated based on the impact sensed by the sensor 130, and performs an image recording when the event is determined to have occurred And can be implemented as a kind of application processor (AP) such as an MCU (Micro Control Unit) and a CPU (Central Processing Unit). The operation of the control unit 150 will be described in more detail below with reference to FIG. 2 and FIG.

The storage unit 170 stores program instructions for performing the intelligent video recording method, temporarily stores images acquired by the camera 100, and stores images for a predetermined period of time from the event occurrence time.

In one embodiment, the storage unit 170 includes a first storage unit 171 and a second storage unit 172.

The first storage unit 171 is provided with a buffer function for temporarily storing images acquired by the image capturing apparatus 110. The first storage unit 171 may be implemented as a RAM or the like as a volatile storage medium.

The second storage unit 172 stores the image data acquired by the image capturing apparatus 110 by the data writing of the controller 150, preferably the image data temporarily stored and / or stored in the first storage unit 171 A flash memory such as an SD card or a USB memory, an EEPROM, a hard disk, or the like.

In another embodiment, the first storage unit 171 and the second storage unit 172 are functionally divided on the storage unit 170, which are the same components, and may not necessarily mean separate elements separated from each other have. For example, the first storage unit 171 and the second storage unit 172 may be divided into partitions in the storage unit 170.

It will be apparent to those skilled in the art that the intelligent video recording apparatus 100 may include other components. May include other hardware elements required for the operations described herein, including, for example, input devices for data entries, and output devices for printing or other data display. In some embodiments, it may further include a network, a network interface, and a protocol connecting the intelligent video recording apparatus 100 and an external device (e.g., a smart phone, etc.) Notifications can also be provided via smartphones.

FIG. 2 is a flowchart of an intelligent video recording method according to an embodiment of the present invention, and FIG. 3 is a conceptual diagram of an intelligent video recording method according to an embodiment of the present invention.

The image capturing apparatus 110 captures an image in real time (S110), and the image captured in real time is temporarily stored in the first storage unit (S120). If no shock is detected by the sensor 130, the image data continuously photographed through the camera 110 may be temporarily stored in the first storage unit 171 by buffering or the like. At this time, the image data photographed in real time before the impact is detected by the sensor 130 is not stored in the second storage unit 172. [

When the impact is applied to the object mounted with the intelligent video recording apparatus 100 during the execution of steps S110 to S120, the sensor 130 senses an impact applied to the object, (S140).

When the amount of impact and the direction of impact to the applied impact are detected, the intelligent video recording apparatus 100 generates an impact pattern for the impact applied based on the detected impact amount and the impact direction. In one embodiment, the intelligent video recording apparatus 100 estimates the position of the impact applied by applying the impact amount and impact position to the position model 150 (S150), and calculates the impact amount and impact direction of the impact applied, An impact pattern for the applied impact is generated (S160).

The position model 150 estimates the impact position based on the impact amount and the impact direction. In one embodiment, the position model 150 has been modeled based on a plurality of training samples, where each training sample includes the amount of impact, impact direction and impact position on the training impact to produce the sample.

Where the training samples can be represented by a number of features associated with estimating the impact location. In one embodiment, the impact applied to the vehicle can be represented by the amount of impact, impact direction and impact position. The impact amount is data on the magnitude of the impact sensed by the sensor 130, the impact direction is data on the direction of the impact sensed by the sensor 130, and the impact position is data on the position where the impact has occurred in the vehicle.

In some embodiments, the training samples may further include attribute information of the vehicle. Here, the attribute information of the vehicle may include vehicle-related attributes such as the position, thickness, material and size of the vehicle door, position, thickness, material and size of the vehicle trunk, various dimensions of the vehicle body, and the like. Each property information A floating point number, or a binary value, or it may be categorical.

In this way, when modeling the position model 150, it is possible to estimate a different position result for each vehicle regardless of the impact amount and the impact direction by considering the vehicle property in addition to the impact amount, the impact direction and the impact position, .

In one embodiment, the location model 150 may be modeled using a variety of instructional learning algorithms such as a Decision Tree, Bayesian Network, CART, artificial neural network (ANN) algorithms, etc., It is not.

Using the position model, the intelligent video recording apparatus 100 generates a real-time impact pattern for the impact applied based on the impact amount and impact direction of the impact applied to the vehicle, and the estimated impact position (S160).

In some embodiments, the impact pattern is calculated by combining a first pattern associated with the impact amount and a second pattern associated with the impact location. For example, the impact pattern may be a function of (x, y, z), where x is a value associated with the amount of impact, and y, z may be a value for the impact location.

The intelligent video recording apparatus 100 determines whether the impact applied is based on the impact pattern generated in step S160, that is, whether the impact is worthy of video recording, that is, an event (S170). In one embodiment, the intelligent video recording apparatus 100 applies the impact pattern to the event pattern model 170 to determine whether the impact associated with the impact pattern corresponds to the event (S170).

The event pattern model 170 is generated based on a plurality of training samples, as a classification model for determining whether an impact pattern for an impact applied to a vehicle is associated with an event. Each training sample contains many features and binary labels associated with determining event impact. Here, the binary label indicates whether each training sample corresponds to an event impact. In one embodiment, each training sample includes an impulse amount, impact direction, and binary label, and may include a positive label if each training sample corresponds to an event and a negative label if not corresponding to the event. In some embodiments, the training samples may further include attribute information of the vehicle.

The event pattern model 170 is generated by support vector machine SVM, support vector data description SVD, support vector regression SVR, artificial neural network ANN, and K-means clustering algorithm. But the present invention is not limited to this, and it is possible to classify the currently stored data into two classes (event class and non-event class) and to classify the newly entered data into a variety of classes Algorithm. ≪ / RTI >

In one embodiment, the location model 150 and / or the event pattern model 170 may be modeled by the manufacturer of the intelligent video recording apparatus 100. Alternatively, in another embodiment, the location model 150 and / or the event pattern model 170 may be modeled in the intelligent video recording apparatus 100. [

Meanwhile, in the above description, it is apparent to those skilled in the art that the event pattern model 170 is modeled by a plurality of impacts as training samples and features such as physical quantities associated with such impacts. As described above, in this specification, an event refers to various special situations that require image recording such as sudden braking, sudden shooting, etc. in addition to the impact. Therefore, in this case, a physical quantity other than the physical quantity associated with the impact (speed, The gas change amount to be injected, and the like) can be modeled.

If the impact is determined as an event, the intelligent video recording apparatus 100 stores an image for a predetermined time based on the event occurrence time (S180). In this case, the intelligent video recording apparatus 100 can display an image from a time before a specific time to a time after a specific time, that is, for a predetermined time, based on the time of occurrence of an event among the images temporarily stored in the first storage unit 171 And can be stored in the second storage unit 172.

In this way, the intelligent video recording apparatus 100 can more accurately determine the specific usage style and habit of the user, which occurs repeatedly at a specific position, such as a door close of a vehicle, have. As a result, the image data stored in the second storage unit 172 can be image data associated with an event requiring image recording with higher accuracy, so that the storage space can be more efficiently used.

Figure 4 is an exemplary diagram of an impulse location indication, in accordance with an embodiment of the present invention. In this embodiment, when the controller 150 determines that an event has occurred, the controller 150 determines the estimated impact position as the event occurrence position, and provides the notification about the event occurrence position (S190). Here, the notification includes pictures, characters, and combinations thereof associated with the event occurrence location. For example, the event notification may include a character representation of the location of the event impact and / or a graphic representation can do.

In one example, referring to FIG. 4, an event notification may include a graphical representation of the location of the event impact in a perspective view of the entire vehicle.

In another example, the event notification may include a graphical representation of the location of the event impact in a perspective view magnifying a portion of the vehicle.

In yet another example, the event notification may include a graphical representation of the location of the event impact on the floor plan of the entire vehicle.

In addition, the event notification may include a graphic representation and a character representation thereof, as shown in FIG. Here, the literal expression may include the date and time at which the event impact occurred, and the phrase describing the location of the event impact to the vehicle.

In addition, the literal expression of the event notification may further include a statement describing the expected degree of damage and / or the estimated repair cost of the vehicle. In one embodiment, the notification system 1000 may further provide a phrase describing the expected damage level and / or the estimated repair cost using the pre-stored correlation, wherein the correlation is based on the impression pattern, And / or to calculate an estimated repair cost. The operation principle in which such a correlation is generated is similar to the operation principle of generating the event pattern model, and thus a detailed description thereof will be omitted.

In one embodiment, the event notification as shown in FIG. 4 may be provided through the output unit 190 (S190).

In another embodiment, the intelligent video recording apparatus 100 may be configured to communicate with a software application operable to execute on a user ' s mobile input / output device. The intelligent video recording apparatus 100 can transmit and receive data electronically with a user's mobile input / output device through a wired or wireless communication network. In this case, the event notification as shown in FIG. 4 may be provided through the user's mobile input / output device (S190).

In this way, the intelligent video recording apparatus 100 determines whether or not the impact position is notified in consideration of the impact position, so that the estimated impact position is detected at a position at which a repeated impact is routinely applied (for example, around a door such as a door outer panel, Trunk lid close to the trunk lid), it is determined that the event is not an impact and the user is not notified. This can provide a more accurate and efficient notification of the impact. In addition, the user can intuitively inform the user of the impact position through a graphical representation and provide related information, thereby maximizing user convenience.

In addition, prior to performing the steps S120 to S180 for detecting a real-time impact applied to the vehicle and starting video recording, the intelligent video recording apparatus 100 may generate an event pattern model 170 And can be customized for each user (S110).

In one embodiment, the intelligent video recording apparatus 100 receives the intrinsic behavior information for each user and can customize the event pattern model 170 for each user using the intrinsic behavior information. Herein, the intrinsic behavior information for each user includes characteristics related to the impact caused by behaviors unique to each user (for example, a vehicle door closing habit), i.e., an impact amount and an impact direction. The input of the inherent behavior information can be performed by the user performing a specific action on a user-specific basis in a specific setting mode, and the impact applied thereby is detected by the sensor.

The customization process described above may correspond to one or more locations, respectively, for one or more users. For example, in the case of a family of two, the unique input to the borrower is a feature associated with the habit of closing the door to the door of the driver's seat, while the unique input to the other person is the opposite of the driver's seat Quot; door " position). ≪ / RTI >

Through this customization process, it is possible to obtain a difference in the process steps that can be finely generated according to the manufacturing process even in the same vehicle type (for example, material, size of the door, and degree of rotation in the rotary joint connecting the door and the main body) The use of the event pattern model 170 generated collectively by the manufacturer can be improved by reflecting the individual habits and / or the usage style in the event pattern model 170 every time.

Additionally, the intelligent video recording apparatus 100 may update the location model 150 (S200). In addition, the event pattern model 170 may be updated (S300).

5 is a flow chart of a process of updating the position model 150 according to an embodiment of the present invention. The intelligent video recording apparatus 100, at the same time as or after the event notification, And requests the user to input user information to generate feedback information. The first feedback information includes a second label indicating that the estimated impact position is the actual impact position. If the second label value included in the first feedback information is 1, the estimated impact position corresponds to the actual impact position, and if it is the opposite, it has a value of zero. If the second label value is 0, the first feedback information is generated to further include the actual impact position. For example, when the second label value is 0 by the first user input, a user input for the actual impact position in the vehicle can be additionally performed.

The intelligent video recording apparatus 100 may request a first input for generating a second label (S210). In one embodiment, the request may be performed via the user ' s mobile input / output device.

The intelligent video recording apparatus 100 generates the second label in response to the first input in response to the user input inputted through the input device (S230). Thereafter, the first feedback information including the amount of impact and the impact direction used in the notification and the second label is generated (S270), and the position model 150 may be updated using the first feedback information (S290) .

In one embodiment, when the second label indicates that the estimated impact position is not the actual impact position, the intelligent video recording apparatus 100 requests a second input for generating actual impact position information (S250) After generating the actual impact position information in response to the second input (S260), the first feedback information may be generated so as to further include the actual impact position information (S270).

6 is a flowchart of a process of updating the event pattern model 150 according to an embodiment of the present invention. The intelligent video recording apparatus 100 requests the user to input user input to generate the second feedback information related to the previously determined event impact at the same time as or after the event notification S310. The second feedback information includes a third label indicating whether the impact determined as the event impact is an actual event impact. That is, it indicates whether the event impact provided by the notification actually needs to be notified. If the third label value included in the second feedback information is 1, the impact determined as the event impact corresponds to the actual event impact, and if it is the opposite, it has the value of 0. The process of updating the event pattern model 170 is similar to the process of updating the location model 150, so a detailed description will be omitted.

The operations of the intelligent video recording apparatus and method according to the embodiments described above can be implemented at least partly by a computer program. Here, the program for implementing the present embodiment is configured in the form of functional program instructions and codes that can be executed in the computing device. The computing device is a device having one or more alternative and special purpose processors, memory, storage space, and networking components (either wireless or wireline). The computing device may be, for example, an operating system compatible with Microsoft's Windows, an operating system such as Apple OS X or iOS, a Linux distribution, or Google's Android OS.

The above-described operations and their associated units, modules, etc., may be implemented within software, firmware, hardware, or any combination thereof. For example, a software module may be implemented with a program product comprised of a computer-readable medium including program code, which may be executed by a processor for performing any or all of the steps, operations, or procedures described herein .

Functional program instructions and codes for implementing the present embodiment may be readily understood by those of ordinary skill in the art to which this embodiment belongs.

Also, the operation by the intelligent video recording apparatus and method can be recorded on a computer-readable recording medium. An intelligent video recording apparatus and method according to embodiments of the present invention can be implemented by a computer-readable recording medium. The computer-readable recording medium includes all kinds of recording apparatuses in which data that can be read by a computer is stored. Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like. The computer readable recording medium may also be distributed over a networked computer system so that computer readable code in a distributed manner may be stored and executed.

While the invention has been shown and described with reference to certain embodiments thereof, it will be understood by those skilled in the art that various changes and modifications may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. However, it should be understood that such modifications are within the technical scope of the present invention. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

Claims (14)

An intelligent video recording method performed by an intelligent video recording apparatus,
Detecting an impact applied to an object and detecting an impact amount and an impact direction;
Generating an impact pattern based on the impact amount and the impact direction;
Determining whether the impact is an event triggering a video recording based on the impact pattern; And
And storing an image for a predetermined time based on the event occurrence time when the impact is determined as an event. The method according to claim 1,
Wherein the step of generating the impact pattern based on the impact amount and the impact direction comprises:
Applying the change amount and the change direction to a position model to estimate a position of the applied change; And
And generating a change pattern based further on the estimated change position. 3. The method of claim 2,
Wherein the position model is modeled based on a plurality of training samples, each training sample including an impact amount, an impact direction and an impact position on a training impact. The method according to claim 1,
Wherein the step of determining whether the impact is an event includes:
Applying the impact pattern to an event pattern model to determine whether an impact associated with the impact pattern corresponds to an event. 5. The method of claim 4,
Wherein the event pattern model is generated based on a plurality of training samples, each training sample comprising an impact amount, an impact direction and an impact position of a training impact, and a first label, / RTI > The method according to claim 1,
Further comprising the step of receiving user specific behavior information for the target and customizing the event pattern model for each user using the unique behavior information. 3. The method of claim 2,
Requesting a first input for generating a second label;
Generating the second label in response to the first input;
Generating first feedback information including an impact amount and an impact direction used for the notification, and a second label; And
Further comprising updating the location model using the first feedback information,
And the second label indicates that the estimated impact position is an actual impact position. 8. The method of claim 7,
Requesting a second input to generate actual impact location information if the second label indicates that the estimated impact location is not an actual impact location; And
Further comprising generating actual impulse position information in response to the second input, wherein the first feedback information is generated to further include the actual impulse position information. 5. The method of claim 4,
Requesting a third input for generating a third label;
Generating the third label in response to the third input;
Generating impulse amount and impact direction used in the notification, and second feedback information including the third label; And
The method of claim 1, further comprising updating the event pattern model using the second feedback information, wherein the third label indicates whether the previously determined impact is an actual event. The method according to claim 1,
Determining the impact position as an event occurrence position when the impact is determined to be an event; And
And providing a notification of the occurrence and location of the event. 10. The method of claim 9,
Wherein the event notification includes a graphical representation of the occurrence location of the event impact on at least a portion of the object. 11. The method of claim 10,
Wherein the event notification further includes at least one of a predicted degree of damage due to the event impact and a character expression indicating an estimated loss amount. A computer-readable medium having stored thereon program instructions readable by a computer and operable by the computer, wherein the program instructions are executable by a processor of the computer, wherein the instructions of any one of claims 1 to 12 The method comprising the steps of: (a) An intelligent video recording apparatus for recording an image photographed when an event occurs,
A video camera for capturing an image in real time;
A sensor for detecting an impact applied to an object and detecting an impact amount and an impact direction;
A control unit for generating an impact pattern based on the impact amount and the impact direction and determining whether the impact applied based on the impact pattern corresponds to an event triggering the image recording; And
And a storage unit including a first storage unit for storing the photographed image and a second storage unit for storing an image for a predetermined time based on the event occurrence time when the impact is determined as an event.

Images