KR102554572B1 - Optical zoom acquisition image sensing device with crop zoom, and method using the same - Google Patents

Abstract

Disclosed is an optical zoom acquisition image monitoring apparatus and method using crop zoom. An optical zoom acquisition video surveillance device using crop zoom according to an embodiment of the present invention includes a communication unit that receives an image through communication with an IP camera (Internet Protocol Camera), a user interface unit that receives an input of the location of an object from the received image, A region setting unit that sets the cropping area including the input object, a center point tracking unit that tracks the center point of the cropping area, a zoom factor determining unit that determines a zoom factor through vector quantization based on the tracked center point, Comprising a correction unit that performs shake correction according to whether or not image enlargement is successful by zoom factor, a cropping unit that enlarges an image frame by performing cropping on the corrected image, and an output unit that outputs the cropped image do. As a result, it is possible to implement a digital zoom in which shaking is corrected and image quality is excellent.

Description

Video Surveillance Apparatus and Method for Obtaining Optical Zoom Using Crop Zoom {OPTICAL ZOOM ACQUISITION IMAGE SENSING DEVICE WITH CROP ZOOM, AND METHOD USING THE SAME}

The present invention relates to a video surveillance apparatus and method for acquiring optical zoom using crop zoom, and more particularly, obtaining optical zoom using crop zoom capable of securing image quality while extending the limit of optical zoom through resolution change. It relates to video surveillance devices and methods.

A PTZ camera is a camera that can adjust electrical rotation (Pan), vertical tilt (Tilt), and zoom (Zoom). If you want to see it, perform the zoom function.

The zoom function is a function that allows the user to enlarge a specific area of the video captured by the PTZ camera to the entire screen. Pan and tilt performance are important in the PTZ camera, but zoom performance has a great impact on the specifications of the PTZ camera.

Zoom features include optical zoom and digital zoom. Optical zoom is performed by hardware using an array of lenses, and digital zoom is performed by software using a program built into the camera.

Optical zoom has a higher image quality than digital zoom, but has a disadvantage in that it can be used only within a magnification range that can be provided by hardware. On the other hand, the digital zoom can enlarge more than the optical zoom, but has a disadvantage in that the quality of the image is deteriorated.

Therefore, it is preferable to use optical zoom in terms of image quality, but since the expansion of optical zoom is directly related to equipment size and cost, it is not easy to add hardware to obtain a desired zoom factor.

Accordingly, there is a need for a correction technique capable of providing image quality comparable to optical zoom even when the magnification of the hardware optical zoom reaches its limit.

Korean Patent Publication No. 10-2020-0094309 (published on August 7, 2020)

In order to solve the above problems, the technical problem to be achieved by the present invention is to find an accurate center point for an image including an object, set a region to be enlarged, and compensate for shaking based on vector quantization, thereby providing an enlarged image of excellent quality. It is to propose an optical zoom acquisition video surveillance apparatus and method using crop zoom capable of acquiring .

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

As a means for solving the above-described technical problem, an optical zoom acquisition video surveillance device to which crop zoom is applied according to an embodiment of the present invention includes a communication unit for receiving an image through communication with an IP camera (Internet Protocol Camera), and a received image. A user interface unit that receives an input of the position of an object from, an area setting unit that sets a cropping area including the inputted object, a center point tracking unit that tracks the center point of the cropping area, and vector quantization based on the tracked center point. A zoom factor determination unit that determines a zoom factor, a correction unit that performs shake correction according to whether or not the image has been successfully enlarged by the determined zoom factor, a cropping unit that enlarges an image frame by performing cropping on the corrected image, and an output unit outputting the cropped image.

The region setting unit sets an object-containing region including objects, sets a point corresponding to maximum values in the width and height directions of the set object-containing region as a conic point, and extends from the conic point to the outside of the object-containing region. By setting a Euclidean area to be, the cropping area may be set using the Euclidean distance.

The area setting unit may calculate the area ratio of the distance proportional surface according to the Euclidean polynomial by the following formula:

According to the present invention, it is possible to implement digital zoom capable of overcoming the limitations of optical zoom by setting a cropping area by a center point traced in an area including an object, determining a zoom magnification according to the cropping area, and applying it to image enlargement. There is an effect of providing an optical zoom acquisition video surveillance apparatus and method to which crop zoom is applied.

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

1 is a network configuration diagram of a digital zoom realization system capable of shake correction according to a preferred embodiment of the present invention;
2 is a block diagram of an optical zoom acquisition video surveillance device to which crop zoom is applied according to a preferred embodiment of the present invention;
3 is a diagram for explaining the operation of the area setting unit shown in FIG. 2;
4 and 5 are views for explaining the operation of the center point tracking unit shown in FIG. 2;
Figure 6 is a view for explaining the operation of the correction unit shown in Figure 2, and,
FIG. 7 is a flowchart for explaining a video surveillance method for acquiring optical zoom to which crop zoom is applied according to a preferred embodiment of the present invention.

The above objects, other objects, features and advantages of the present invention will be easily understood through the following preferred embodiments in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments described herein and may be embodied in other forms. Rather, the embodiments introduced herein are provided so that the disclosed content will be thorough and complete and the spirit of the present invention will be sufficiently conveyed to those skilled in the art.

In this specification, when an element is referred to as being on another element, it means that it may be directly formed on the other element or a third element may be interposed therebetween.

Also, when it is said that a first element (or component) operates or is executed on (ON) a second element (or component), the first element (or component) means that the second element (or component) It should be understood that it is operated or executed in an environment in which it is operated or executed, or operated or executed through direct or indirect interaction with the second element (or component).

Where an element, component, device, or system is referred to as including a component consisting of a program or software, even if not explicitly stated otherwise, that element, component, device, or system means that the program or software executes or operates. It should be understood that it includes hardware (eg, memory, CPU, etc.) or other programs or software (eg, operating system or driver required to drive hardware) required to do so.

In addition, it should be understood that, unless otherwise specified, the element (or component) may be implemented in any form of software, hardware, or both software and hardware.

In addition, terms used in this specification are for describing embodiments and are not intended to limit the present invention. In this specification, singular forms also include plural forms unless specifically stated otherwise in a phrase. The terms 'comprises' and/or 'comprising' used in the specification do not exclude the presence or addition of one or more other elements.

1 is a network configuration diagram of a digital zoom realization system capable of shake correction according to a preferred embodiment of the present invention.

Referring to FIG. 1, a digital zoom realization system capable of shake correction according to a preferred embodiment of the present invention includes an IP camera 100 and a digital zoom realization device 200 connected to each other to enable communication through a network. .

The IP camera 100 is a camera having a function of transmitting an image to an electronic device such as a PC or mobile device by being connected to the Internet by wire or wirelessly. The IP camera 100 according to the present embodiment captures an image of a preset shooting area and transmits the image to the digital zoom implementation device 200 through a predetermined network.

The type of camera applied to the IP camera 100 is irrelevant. For example, a dome type, a box type, a bullet type, and a PTZ (Pan Tilt Zoom) type may be applied, and in general, pan, tilt, and zoom controllable on the road PTZ cameras are mainly installed.

The device 200 for implementing digital zoom receives an image captured by the IP camera 100 by communicating with the IP camera 100 through a predetermined network. Thereafter, the digital zoom implementation device 200 may provide the image received from the IP camera 100 according to a user's (or manager's) request.

In addition, the device 200 for realizing digital zoom may provide an image captured by the IP camera 100 as it is, but may provide various functions including a zoom function to suit a user's request.

In addition, when the IP camera 100 is installed indoors, it is not greatly affected by the external environment, but when the IP camera 100 is used in the form of a closed-circuit television (CCTV), it is usually installed outdoors. As such, when the IP camera 100 is installed outdoors, shaking of the IP camera 100 may occur due to the influence of external environments such as wind and vibration of a pier. (100) is reflected as it is in the video taken, and it appears as shaking of the video.

Therefore, when the IP camera 100 captures an image in which shaking occurs due to the influence of the external environment, the device 200 for implementing digital zoom provides a function of correcting the shaking image. That is, the device 200 for realizing digital zoom provides a digital zoom function beyond the limit of optical zoom and a function of correcting a shaken image. The digital zoom realization device 200 will be described in more detail in FIG. 2 to be described later.

In this embodiment, it is illustrated that there is only one IP camera 100 connected to the digital zoom implementation device 200 through a network, but this is merely for convenience of description and is not limited thereto. That is, the apparatus 200 for implementing digital zoom can communicate with the plurality of IP cameras 100 through a network and receive images from various points.

2 is a block diagram of an optical zoom acquisition video monitoring apparatus to which crop zoom is applied according to a preferred embodiment of the present invention.

Referring to FIG. 2 , the device 200 for realizing digital zoom according to a preferred embodiment of the present invention includes a communication unit 210, a user interface unit 220, an area setting unit 230, a center point tracking unit 240, and a zoom factor. It includes a determination unit 250, a correction unit 260, a cropping unit 270, an output unit 280, a storage unit 290, and a control unit 295.

The communication unit 210 supports the network interface of the digital zoom implementation device 200, and in this embodiment, can receive an image captured by the IP camera 100 from the IP camera 100, and input by a user. Various control signals may be transmitted to the IP camera 100 .

The user interface unit 220 supports an interface between the digital zoom implementation device 200 and the user, and can receive various control signals input by the user. To this end, the user interface unit 220 may be provided to receive input signals through hardware input devices such as a keyboard, a mouse, and a control keyboard (joystick). For example, the user may input the position of the object by moving the mouse, and the position of the mouse cursor may be the position of the object input by the user.

The area setting unit 230 sets a cropping area including the object based on the position of the object input through the user interface unit 220 . More specifically, the area setting unit 230 sets an object-containing area including an object, sets a point corresponding to the maximum values in the width direction and height direction of the set object-containing area as a conic point, and sets the object from the conic point. By setting a Euclidean region extending outside the included region, the cropping region may be set using the Euclidean distance.

The area setting unit 230 may calculate the area ratio of the distance proportional surface according to the Euclidean polynomial by Equation 1.

은 유클리어 다항식에 따른 거리 비례 표면의 면적비, i는 유클리어 이동거리 중심점으로의 이동좌표, j는 현재 벡터의 이동거리 좌표, M은 코닉 포인트 중심 기약 다항식 기반의 저면적 시프트 비트병렬 비교식 모델, u는 유클리어 거리 변수, v는 현재 객체의 벡터 위치이다.here,

is the area ratio of the distance proportional surface according to the Euclidean polynomial, i is the movement coordinate to the center point of the Euclidean movement distance, j is the movement distance coordinate of the current vector, M is a low-area shift bit-parallel comparison model based on the conic point central irreducible polynomial , u is the Euclidean distance variable, v is the vector position of the current object.

The center point tracking unit 240 tracks the center point of the cropping area set by the area setting unit 230 . More specifically, the center point tracking unit 240 forms a circle including the cropping area, sets a point at which the circle and the cropping area contact each other as a reference point, and then sets the maximum number of rows and columns of each side corresponding to the cropping area. The center point can be tracked by the value.

Also, the center point tracking unit 240 may calculate the center point by Equation 2.

는 중심점의 좌표, x는 이미지 방향성 벡터의 위치값, μ는 x가 중심점으로 이동될 확률 벡터, σ는 이동확률 벡터 반복문이다. here,

is the coordinates of the center point, x is the position value of the image direction vector, μ is the probability vector that x will move to the center point, and σ is the movement probability vector repeat statement.

The zoom magnification determiner 250 determines the zoom magnification through vector quantization based on the center point tracked by the center point tracker 240 . Memory cropping may be performed during the operation of the zoom factor determining unit 250 . At this time, the zoom magnification determiner 250 may determine the zoom magnification through vector quantization using a convolutional neural network (CNN) algorithm.

Among many deep learning models, CNN algorithms are capable of powerful frame understanding, decomposition, and interpretation for image recognition. Since the CNN algorithm is already known, a description thereof will be omitted.

The correction unit 260 performs shake correction according to whether or not the image enlargement is successful according to the zoom factor determined by the zoom factor determiner 250 . At this time, the correction unit 260 may determine whether the image enlargement is successful by comparing the zoom factor determined in the current process with the previous zoom factor. More specifically, the correction unit 260 determines that image enlargement has failed if the current zoom factor is less than or equal to the previous zoom factor, and determines that image enlargement has succeeded if the current zoom factor exceeds the previous zoom factor. Thereafter, the correction unit 260 performs shake correction only when it is determined that image enlargement is successful.

In addition, when the correction unit 260 determines that the image enlargement is successful and wants to perform shake correction, it calculates coordinates for shake correction and performs shake correction using the calculated coordinates for shake correction. Here, the correction unit 260 may calculate coordinates for shaking correction by Equation 3.

는 상기 흔들림 보정을 위한 좌표, x는 흔들리는 이미지의 방향성 벡터, k는 차원 벡터, μ는 x가 상기 중심점으로 이동될 확률에 대한 벡터, ∑는 벡터의 반복된 공 분산 행렬,

는 흔들림이 지속된 시간이다.here,

is the coordinate for the shake correction, x is the directional vector of the shake image, k is a dimension vector, μ is a vector for the probability that x is moved to the center point, ∑ is a repeated covariance matrix of vectors,

is the duration of shaking.

The cropping unit 270 enlarges an image frame by performing cropping on the image corrected by the correction unit 260 . Cropping is a term used to describe a method of trimming unnecessary parts of a photo or the like, usually to reveal objects you want to highlight.

The output unit 280 outputs the cropped image. To this end, the output unit 280 may be implemented such that a hardware output device is connected to a display device, and outputs a corresponding image to the display device.

The storage unit 290 stores all information necessary for the operation of the digital zoom implementation device 200 . For example, the storage unit 290 stores Equation 1, Equation 2, and Equation 3 so that the area setting unit 230, the center point tracking unit 240, and the correction unit 260 operate smoothly. you can make it

The control unit 295 controls overall operations of the digital zoom realization device 200 . That is, the control unit 295 includes the communication unit 210, the user interface unit 220, the region setting unit 230, the center point tracking unit 240, the zoom ratio determination unit 250, the correcting unit 260, the cropping unit ( 270), the output unit 280, and the storage unit 290 control signal input and output.

FIG. 3 is a diagram for explaining the operation of the area setting unit shown in FIG. 2 .

The area setting unit 230 sets the object-containing area (A) using the location of the object designated by the user, and sets a point corresponding to the maximum values in the width and height directions in the object-containing area (A) as a conic point (Conic point). Point) (B).

Then, a Euclidean region C extending from the conic point B to the outside of the object-containing region A is set. The Euclidean area (C) corresponds to the area specified in this embodiment to use the Euclidian distance used when calculating the distance between two points.

The Euclidean region (C) is partitioned by lines extending diagonally from each conic point (B), whereby regions corresponding to I, II, III, and IV are created. As shown, in the Euclidean region C, four triangular regions having vertices located at the center of the Euclidean region C may be created.

)를 수학식 1에 의해 산출할 수 있다.As a result, regions I, II, III, and IV corresponding to the Euclidean region (C) and region V corresponding to the object-containing region (A) exist. The area setting unit 230 sequentially calculates the area ratio of the five areas centered on the conic point B, so that the area ratio of the distance proportional surface according to the Euclidean polynomial (

) can be calculated by Equation 1.

은 코닉 포인트(B)를 중심으로 코닉 면적 다항식에 대한 계산식을 각 영역(Ⅰ 내지 Ⅴ)별로 5번 순차 연산함을 의미한다. 또한,

는 각 영역(Ⅰ 내지 Ⅴ)의 면적에 대한 누적 값이다. 수학식 1은, 코닉 포인트(B) 중심의 기약 다항식 기반의 저면적 시프트 비트 병렬 비교식이라 할 수 있다.In the above Equation 1,

means that the calculation formula for the conic area polynomial centered on the conic point (B) is sequentially calculated 5 times for each area (I to V). also,

is a cumulative value for the area of each region (I to V). Equation 1 can be referred to as a low-area shift bit parallel comparison equation based on an irreducible polynomial centered on the conic point (B).

)를 산출한 후, 이를 이용하여 객체가 포함되는 크롭핑 영역을 설정한다. IP 카메라(100)로부터 입력받은 영상에 따라서 객체의 위치가 중심이 아닐 수 있지만, 영역 설정부(230)에서는 객체의 위치과 무관하게 객체를 포함하는 크롭핑 영역을 설정하도록 한다.Finally, the area setting unit 230 uses Equation 1 to determine the area ratio of the surface proportional to the distance according to the Euclidean polynomial (

) is calculated, and the cropping area in which the object is included is set using this. Although the position of the object may not be the center according to the image received from the IP camera 100, the area setting unit 230 sets the cropping area including the object regardless of the location of the object.

4 and 5 are diagrams for explaining the operation of the center point tracking unit shown in FIG. 2 .

The center point tracking unit 240 tracks the center point E of the cropping area D set by the area setting unit 230 . According to the image illustrated in FIG. 4, an object exists in the center of the image input from the IP camera 100, and in this case, the cropping area D is set so that the object is in the center, and the center point tracking unit 240 A center point E corresponding to the center of the cropping area D is found. The cropping area D may be cropped like D' after cropping is performed later.

The operation of the center point tracking unit 240 will be described in more detail with reference to FIG. 5 . The center point tracking unit 240 forms a circle (F) including the cropping area (D), sets a point where the circle (F) and the cropping area (D) come into contact with each other as a reference point, and then crops the area ( The central point (E) is traced by the maximum value of the row and column of each side corresponding to D).

에서

는 중심점 위치의 열(Rows)에 해당하는 값이고,

는 중심점 위치의 행(Cols)에 해당하는 값이다. 또한,

는 원주율을 중심으로 이미지의 중심을 찾기 위해 사용되며,

는 원주율을 중심으로 이미지 프레임 중심을 찾기 위한 4개의 기준점 포인트를 정하고 각 변의 행과 열의 최대값을 구하기 위해 사용된다.The center point tracking unit 240 may determine the center point by Equation 2 above. In Equation 2, corresponding to the coordinate values of the center point

at

Is a value corresponding to the rows of the central point location,

Is a value corresponding to the row (Cols) of the central point position. also,

is used to find the center of the image around the circumference,

is used to determine the four reference points for finding the center of the image frame around the circumference of the circle and to obtain the maximum values of the rows and columns of each side.

FIG. 6 is a diagram for explaining the operation of the correction unit shown in FIG. 2 .

When shaking occurs in the IP camera 100 due to external environments such as wind and vibration of a bridge pier, shaking also occurs in an image captured by the IP camera 100. Since the image with such shaking is not the image desired by the user, it is necessary to correct the image to a shake-free image.

According to the present embodiment, the correction unit 260 determines whether or not image magnification is successful according to the zoom magnification determined by the zoom magnification determination unit 250, and performs shake correction when the image magnification is successful.

The correction unit 260 calculates the coordinates for shake correction using Equation 3, and performs shake correction using the calculated coordinates for shake correction.

)를 산출하는 수학식 3에서,

은 2K의 크롭핑된 이미지의 좌표 즉,x=x1, y=x2, w=x3, h=x4에 해당한다. 또한, 수학식 3에서 흔들리는 방향성 벡터 x가 타임라인에 따라 흔들림이 지속되는 시간이 확률 벡터

에 도달하면 시계열 무한 루프(

)가 "

"를 만족함으로 시계열 변수를

로 초기화시킨다.

은 도 6에서 도시한 화살표 방향으로 도시한 방향으로의 흔들림이 지속된 시간일 수 있다.Coordinates for shake compensation (

) In Equation 3 that calculates,

corresponds to the coordinates of the 2K cropped image, that is, x=x1, y=x2, w=x3, and h=x4. In addition, in Equation 3, the duration of the shaking direction vector x along the timeline is a probability vector

is reached, the time series infinite loop (

)go "

" by satisfying the time series variable

initialize it with

may be the duration of shaking in the direction indicated by the arrow shown in FIG. 6 .

은 이미지 면적 벡터만큼 차원 벡터 k를 만족할 때까지 무한 반복시킨다. 또한, 수학식 3에서

은 피칭 또는 롤링되는 이미지 타임라인을 중심으로 피칭의 경우 상,하 각 1/2, 롤링의 경우 좌,우로 각 1/2로 분할하는 역할을 한다.Also, in Equation 3

is infinitely repeated until the dimension vector k is satisfied as much as the image area vector. Also, in Equation 3

serves to divide the pitching or rolling image timeline into upper and lower 1/2 in case of pitching and left and right 1/2 in case of rolling.

FIG. 7 is a flowchart for explaining a video surveillance method for acquiring optical zoom to which crop zoom is applied according to a preferred embodiment of the present invention.

In this embodiment, referring to FIGS. 1 to 7 , a video surveillance method for obtaining an optical zoom using crop zoom according to a preferred embodiment of the present invention will be described.

When an image is captured by the IP camera 100, the IP camera 100 transmits the captured image to the digital zoom implementation device 200 through network communication. Accordingly, the digital zoom implementation device receives an image captured by the IP camera 100 (S300). The image transmitted by the IP camera 100 may be stored in the storage unit 290 for a predetermined period of time.

The image received from the IP camera 100 is output through the output unit 280 and provided to the user. Through this output image, the user can input the location of the object by manipulating an input device such as a mouse. The location of the object input by the user is input through the user interface unit 220 (S310). If there is no location information of the object specially input from the user, the current mouse cursor location may be determined to be the location of the object or the center of the current image may be determined to be the location of the object according to a preset setting value. there is.

The area setting unit 230 sets a cropping area D including the location of the object input by the user (S320). The cropping area D in the area setting unit 230 may be set as described in FIGS. 3 and 4 .

When the cropping area D is set by the area setting unit 230, the center point tracking unit 240 tracks the center point of the cropping area D (S330). The method of tracking the center point in the center point tracking unit 240 is as described in FIGS. 4 and 5 .

When the center point is determined by the center point tracker 240, the zoom magnification determiner 250 determines a zoom magnification (S340). Thereafter, the correction unit 260 compares the current zoom factor determined by the zoom factor determining unit 250 with the previous zoom factor to determine whether the image enlargement is successful (S350).

In step S350, the correction unit 260 corrects shaking of the image when the image enlargement is successful, that is, when the current zoom factor exceeds the previous zoom factor (S360).

Then, by performing cropping, the image frame is enlarged by the current zoom magnification (S370). If, in step S350, the correction unit 260 fails to enlarge the image, that is, if the current zoom magnification is less than or equal to the previous zoom magnification, cropping is not performed.

Finally, when the enlargement of the image frame is completed, the output unit 280 outputs the enlarged image in a form that the user can check (S380).

By this procedure, according to the video surveillance method for obtaining optical zoom using crop zoom according to the present invention, the limitation of optical zoom can be overcome by using cropping zoom, but the advantage of overcoming the limitation of digital zoom that causes image quality degradation there is

Those skilled in the art to which the present invention pertains will understand that the present invention may be embodied in other specific forms without changing its technical spirit or essential features. Therefore, the embodiments described above should be understood as illustrative in all respects and not limiting. The scope of the present invention is indicated by the following claims rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and equivalent concepts should be construed as being included in the scope of the present invention. do.

200: digital zoom implementation device 210: communication unit
220: user interface unit 230: area setting unit
240: central tracking unit 250: zoom magnification determining unit
260: correction unit 270: cropping unit
280: output unit 290: storage unit
295: control unit

Claims (10)

A communication unit for receiving an image through communication with an IP camera (Internet Protocol Camera);
a user interface unit that receives a location of an object from the received image;
an area setting unit configured to set a cropping area including the input object;
a center point tracking unit for tracking a center point of the cropping area;
a zoom factor determiner determining a zoom factor through vector quantization based on the tracked center point;
a correction unit performing shake correction according to whether or not the image enlargement by the determined zoom magnification is successful;
a cropping unit enlarging an image frame by performing cropping on the corrected image; and
Including; an output unit for outputting the cropped image;
The area setting unit,
An object-containing region including the object is set, a point corresponding to the maximum values in the width direction and height direction of the set object-containing region is set as a conic point, and a conic point extending outside the object-containing region An optical zoom acquisition video surveillance device using crop zoom, characterized in that a Euclidean area is set and the cropping area is set using a Euclidean distance.
delete According to claim 1,
The area setting unit calculates the area ratio of the distance proportional surface according to the Euclidean polynomial by the following formula.

here,

is the area ratio of the distance-proportional surface according to the Euclidean polynomial, i is the movement coordinate to the center point of the Euclidean movement distance, j is the movement distance coordinate of the current vector, M is the low-area shift bit parallel comparison equation based on the conic point centered irreducible polynomial model, u is the Euclidean distance variable, v is the vector position of the current object.
According to claim 1,
The center point tracking unit forms a circle including the cropping area, sets a point at which the formed circle and the cropping area contact each other as a reference point, and determines the maximum value of each row and column of each side corresponding to the cropping area. An optical zoom acquisition video surveillance device using crop zoom, characterized in that by tracking the center point by the.
According to claim 4,
The center point tracking unit determines the center point by the following formula:

here,

is the coordinate of the center point, x is the position value of the image direction vector, μ is a probability vector that x will move to the center point, and σ is a movement probability vector repeat statement.
According to claim 1,
The correction unit determines that the image enlargement has failed if the determined zoom factor is less than or equal to the previous zoom factor, and determines that the image enlargement has succeeded if the determined zoom factor exceeds the previous zoom factor. An optical zoom acquisition video surveillance device using crop zoom, characterized in that the shake correction is performed only when it is determined.
According to claim 6,
The correction unit, when it is determined that the image enlargement is successful, calculates coordinates for shake correction, and performs the shake correction using the calculated coordinates for shake correction. Acquisition Video Surveillance Device.
According to claim 7,
The correction unit calculates the coordinates for the shake correction by the following formula. An optical zoom acquisition video surveillance device using crop zoom:

here,

is the coordinate for the shake correction, x is the directional vector of the shake image, k is a dimension vector, μ is a vector for the probability that x is moved to the center point, ∑ is a repeated covariance matrix of vectors,

is the duration of shaking.
In the video surveillance method for obtaining an optical zoom using crop zoom applied to a device capable of communicating with an IP camera,
receiving an image through communication with the IP camera;
receiving a position of an object from the received image;
setting a cropping area including the input object;
tracing a center point for the cropping area;
determining a zoom magnification through vector quantization based on the tracked central point;
performing shake correction according to whether or not the image is successfully enlarged by the determined zoom magnification;
Enlarging an image frame by performing cropping on the corrected image: and
Including; outputting the cropped image;
In the step of setting the cropping area,
setting an object-containing region including the object;
setting a point corresponding to the maximum values in the width and height directions of the set object-containing region as a conic point; and
setting a Euclidean area extending from the conic point to the outside of the object-containing area, and setting the cropping area using the Euclidean distance; Acquisition video surveillance method. delete

Images