KR20110044492A - An image monitoring system with target axis of position sensing and tracking device - Google Patents

Abstract

PURPOSE: A target image monitoring system and a tracking device thereof are provided to perform automatic tracking having reliability with a low cost. CONSTITUTION: An optical unit(20) comprises a camera(21), a gyro sensor(22), a GPS antenna, and a GPS sensor. A first rotation driver(30) is rotationally installed on the axis line of the camera and detects the location of a rotary motor(31) by equipping with a position detecting sensor(32). A port(35) is loaded on the rotation driver and drives an optical unit through call connection. A second rotation driver(40) is in associated with the lower part of the first rotation driver and detects the location of a rotary motor(41). A second port(45) is loaded on the second rotation driver.

Description

An image monitoring system with target axis of position sensing and tracking device}

The present invention relates to a target coordinate tracking video surveillance system by position recognition for tracking a moving object using a camera image coordinate point, and more particularly, a movement for detecting a moving direction of a moving object in an image frame received in real time. The present invention relates to a control technique that responds to position detection results in position detection of an object, coordinate values by a gyro sensor, and position recognition values by a GPG sensor, and a tracking video surveillance apparatus accordingly.

In general, image processing can be classified into low level, middle level, and high level processing. Low level image processing performs a basic image to remove noise or extract a feature from an input image. It can be said that a process of extracting a combination from these extracted features. In addition, the upper level is a process of performing an object recognition through an inference process from such an extracted object.

The basic principle of detecting the position of the moving object is to compare the data of the previous frame and the data of the current frame received from the camera in real time and extract the changeable part to detect the position of the moving object.

In general, 3D model-based, area-based, active contour-based, and feature-based methods are used for tracking moving objects. The 3D model-based method is an object-oriented method that restores highly accurate models and trajectories for a small number of vehicles. The area-based method secures an area connected to a continuous image and tracks a moving object through correlation measurement. The area-based method detects an object according to an image difference from an input image by measuring a background of a photographing time point. In addition, there is an active contour-based method for expressing and tracking a boundary contour of an object while dynamically updating it, and a feature-based method for tracking a local feature rather than tracking the entire object is a representative object tracking method.

Recently, object tracking methods do not depend on one specific method and apply only the advantages of these methods. That is, it is generally composed of a process of detecting a motion in a surveillance area, separating and isolating a motion region from a background screen, extracting feature points, and tracking a direction of motion vector by tracking feature points of continuous images. The following is the process of implementing each step algorithm.

Background Extraction (Backgroud Extraction) is a process of obtaining a background image by separating moving objects from several consecutive images. An analog image signal input from a CCD camera is acquired through a frame grabber, and then converted into a digital image signal. If necessary, the frames are separated into appropriate sizes and filtering is performed to remove the contained noise.

Binarization extracts the contour of an image and calculates a more accurate motion vector by calculating the position of the moving object through calculation with the already binarized and filtered difference image, from which the motion region in the image is separated from the background image. You can get a video.

Convex polygonal contours extracted from contour extraction can't be separated separately from the moving parts of moving objects because they are affected by changes in the surrounding environment. Therefore, it is necessary to extract the feature inside the contour to distinguish the object being tracked.

The background image is extracted from the input image and binarized, and the target image is tracked and sent to the screen through the recognition and control of the input feature according to the feature of the target image. ), So a detailed description thereof will be omitted.

However, the edge detection method of conventional image processing has a lot of computation in hardware, and in some cases, there is a complexity that a learning algorithm requiring a complex mask is required.

For example, complex algorithms and computations such as image recognition to judge cars or people, and cognitive science and artificial intelligence to recognize car license plates were needed.

The present invention has been made to solve the above-described problems, and the algorithm for the precision tracking by calculating the coordinate data recognition function by the gyro sensor, the camera position by the GPS sensor and the data value of the correct image by the rotation drive device Its purpose is to simplify the image acquisition process steps.

In addition, it is possible to detect the moving direction of a moving object by introducing only the processing of the coordinate values by the gyro sensor and the position value of the GPS sensor in the image frame input in real time, so that it does not require the performance of many processes as before. There is another purpose.

A camera capable of moving according to a control signal;

An image detector which acquires motion coordinates from an image provided by the camera and determines a motion coordinate different from a moving direction of the camera;

A gyro sensor unit configured to provide position adjustment information for positioning an object moving according to an input movement degree at a center of the image;

In the target coordinate tracking video surveillance system by position recognition comprising a GPS sensor unit for checking the position of the camera to recognize the exact position even in the zoom ratio of the camera,

The image detection unit calculates and outputs a position and movement coordinate value of a moving target by comparing an image provided by the camera with a previous or subsequent image in coordinate units,

The gyro sensor unit recognizes the coordinates of the moving target to calculate and output the coordinates of the position of the target,

The GPS sensor unit calculates and outputs a coordinate value of the camera position,

The control unit receives the moving position and coordinate values of the target from the image detector, the gyro sensor unit, and the GPS sensor unit to classify the noise and the monitoring target based on the position and the direction of the moving target, and thus the center position and the moving coordinate of the monitoring target. It provides a target coordinate tracking image monitoring system by position recognition further comprising an automatic tracking control unit for calculating a.

According to an embodiment of the present invention, the target coordinate tracking image surveillance system using a single camera detects the movement of an object of a set size as judging a target in a different direction from a background movement while monitoring a large area using a single camera. By continually centering, reliable auto-tracking is possible at low cost, resulting in a wide range of applications.

According to an embodiment of the present invention, the target coordinate tracking video monitoring system according to the position recognition can effectively perform monitoring on subsequent detected objects by changing the range of the automatic tracking area when other moving objects are detected during the automatic tracking. There is an effect that can alleviate the limitation of the tracking method using a camera.

According to an embodiment of the present invention, the target coordinate tracking video surveillance system by subdividing the cell size when the moving object is detected to filter the noise or to filter the noise differently according to the monitoring level to increase the reliability of the tracking. It can be effective.

According to an embodiment of the present invention, the target coordinate tracking image monitoring system based on position recognition acquires and stores an enlarged image of a monitoring target through zoom-in when the moving speed of the target under automatic tracking decreases or stops, and also records the coordinate information. It is effective to increase the reliability of surveillance information and to use it as evidence when necessary.

The target coordinate tracking video surveillance system based on location recognition according to an embodiment of the present invention adds geographical information about a surveillance region to distinguish a main surveillance region from a non-surveillance region, thereby reducing noise and concentrating surveillance. There is an effect that can increase the efficiency.

BRIEF DESCRIPTION OF THE DRAWINGS The above and other objects, features and advantages of the invention will become more apparent from the following detailed description of the present invention when taken in conjunction with the accompanying drawings.

As shown in FIGS. 2 and 3, the tracking image monitoring apparatus 10 used in the present invention includes a camera 21 having one end fixed and photographing in the axial direction, and a gyro sensor 22 recognizing a target coordinate set in front of the tracking image monitoring apparatus 10. An optical means unit 20 having a GPS sensor 23 associated with a GPS antenna 24 received from a satellite and at the rear;

A first rotational drive unit 30 rotatably installed on an axis of the camera 21 and having a position detection sensor 32 for detecting a position of the rotation motor 31;

A power source, an image, and a communication connection mounted on the first rotation driving unit 30 to drive the camera 21, the GPS sensor 22, and the gyro sensor 23 of the optical means unit 20. 1 port part 35;

A second rotation driver 40 having a position detection sensor 42 for detecting a position of the rotation motor 41 to rotate left / right in association with a lower portion of the first rotation driver 30;

The camera 21 generates driving signals of the first rotation driver 30 and the second rotation driver 40 based on the coordinate data set for photographing the target, and transmits the driving signals to the corresponding rotation drivers 30 and 40, respectively. A storage means (50) provided with a driving signal output unit (51) and an image processor (52) for storing an image of the camera;

And a second port part 45 mounted on the second rotation driving part 40 to connect power, video, and communication of the storage means to the outside. .

The first rotational drive part has an annular shape as a whole, and is provided with a five-phase stepping motor and a brake device therein, and is configured to be rotatable within a range of 0 ° to 90 ° in the vertical direction of the camera. Therefore, the first rotation driving unit may rotate the camera circumference by a predetermined angle (for example, 10 ° clockwise) according to the driving signal received from the outside.

The second rotary drive unit is mounted below the first rotary unit, and is configured to rotate within the range of about 0 ° to 180 ° independently of the rotary motion of the first rotary unit. To this end, a five-phase stepping motor and a brake device are provided inside. Therefore, the second rotating part may rotate by a predetermined angle (for example, 10 ° in the ascending direction) according to the driving signal received from the outside.

The present invention is to explain the operation method of the target coordinate tracking video surveillance system by position recognition to monitor a large surveillance area using a single camera capable of movement.

Since the target monitoring of the area secured to the image of the camera must be performed using a single camera, the monitoring is performed in such a manner that the monitoring screen of the image is scanned along the surveillance area while moving the camera. In this way, all surveillance areas can be scanned at a uniform level, but the limitations of using a single camera can reduce surveillance efficiency. Accordingly, the scan schedule of the camera may be adjusted to reduce the scan rate of the image area having a low monitoring level and to increase the scan rate of the image area having a high monitoring level.

In other words, it is possible to monitor a high-monitored video area for a long time in a stationary state, reduce the scanning speed to increase the monitoring time, or set a path such that a high-monitored video area is included in the scanning path multiple times. In addition, the low-monitoring video area can be scheduled to be monitored for a relatively short time.

When the monitoring target appears in the image obtained according to the monitoring schedule of the camera in which the stop and the movement are alternated, the camera keeps the monitoring target until the monitoring target moves out of the surveillance area or completely stops in accordance with the movement of the monitoring target. To track.

As shown in FIG. 1, the driving unit is driven according to a control signal of the automatic tracking control unit to operate a camera and provides a zoom function when necessary, and the level of the image is fixed to facilitate identification of the image provided by the camera. An image processor for converting an image of a camera into an image format to be maintained at a level or stored in a storage unit, an image detector for acquiring a motion coordinate from an image provided by the camera, and determining a motion coordinate different from a moving direction of the camera; Including a gyro sensor unit for providing position adjustment information for the moving object to be located in the center of the image according to the degree of movement input, and a GPS sensor unit for checking the position of the camera to recognize the exact position even in the zoom ratio of the camera The target coordinate tracking video surveillance system For example, the image detecting unit compares an image provided by the camera with a previous or subsequent image in coordinate units, calculates and outputs a position and movement coordinate value of a moving target, and the gyro sensor unit recognizes a coordinate of a moving target. Calculates and outputs a coordinate value of the target position, and the GPS sensor unit calculates and outputs a coordinate value of the position of the camera, and the automatic tracking control unit of the target is detected from the image detection unit, the gyro sensor unit, and the GPS sensor unit. It consists of an automatic tracking control unit that receives the movement position and coordinate values and calculates the center position and movement coordinates for the monitoring target by classifying the noise and the monitoring target based on the position and direction of the movement target. The driving unit, the camera, and the image processing unit may be collectively referred to as a camera unit (or simply a camera).

Using this basic configuration, the movement of the image can be automatically tracked, and the necessary image can be stored in the storage unit.

The image input from the camera is divided into a predetermined target area, the movement of each target is detected, the coordinate information is extracted, and the target determined as an example of the case where there is a moving object in the still image of the image processor from which the target is extracted However, in the subsequent image, it is an example of a method of acquiring the movement coordinates by detecting the movement to another position through similar target detection. In this case, the moving speed of the target may be obtained through the acquisition speed of the image.

Based on the information obtained, the tracking target is always centered on the screen, so that the tracking target can be continuously and safely tracked.

For example, when a target moves in a still image, the target of a plurality of targets knows that the target moves to a target of a different position in the next image, and the image detection unit detects the moving target. If the camera is stationary compared to the previous or subsequent image in units of targets, the position and movement coordinate values of the moving target are calculated and output.

That is, since the size of the monitoring target can be determined through the number of adjacent targets moving in the same direction based on the position information and the coordinate information of the plurality of moving targets output through the image detector, the desired monitoring target can be limited to the target. In this case, by limiting the number of such targets, unnecessary noise can be blocked. For example, if the number of targets is 20 or more, noise or noise caused by people or animals and the natural environment (indirect image change caused by wind, rain, snow, hail, etc.) that are less than or equal to the number of targets is opened or closed. Can also block noise.

It detects the target image from the image detector through the gyro sensor and detects the target displacement with respect to the moving target by calculating the three-dimensional coordinates of the coordinates using the gyro sensor values measured according to the movement of the target in three-dimensional space. This makes the location of the target accurate.

If the position coordinates can be acquired by interlocking with the GSP satellite through the GPS sensor, the position of the camera is determined by performing a mode of obtaining the position coordinates again using the GPS sensor.

On the other hand, as the target is set to a larger size, it is easier to calculate, so in a basic monitoring situation, a method of increasing the size of the target set in the image detector and detecting a noise by reducing the size of the target when motion is detected may be applied. If the monitoring target is selected through such noise filtering, the monitoring target will be continuously tracked, thereby increasing the size of the target again, thereby reducing the amount of computation required for real-time motion detection.

For example, since the number of targets required for vehicles, humans, and animals are different for image output, monitoring targets can be classified based on the number of targets, and the computational load for this is not large. In order to apply this method, an average value of center positions and coordinates of moving targets obtained through the image detector, the gyro sensor, and the GPS sensor should be calculated and provided to an automatic tracking controller for generating position adjustment information for actual automatic tracking.

Therefore, the automatic tracking control unit according to the present invention receives the moving position and coordinate values of the target from the image detector, the gyro sensor unit and the GPS sensor unit and provides the center position and the moving coordinates for the monitoring target to the automatic tracking control unit to average the average values. It is provided to calculate coordinate values. In addition, the automatic tracking controller sets the target size of the image detector to be large in the monitoring state, sets the target size to be small when motion detection is confirmed by the image detector, and sets the target size and the moving target provided from the image detector. It is desirable to filter the noise based on the number and the direction of?.

On the other hand, the automatic tracking control unit generates position adjustment information for determining the orientation of the camera to the predicted movement position of the monitoring object based on the coordinates and target information of the monitoring object, the moving speed and zoom state of the camera and the driving restriction information. . Herein, the driving limit information means a limit speed or a rotation angle of the actual driving unit. Of course, the control unit may process some of these functions instead, so that the position and direction of the moving target provided by the image detection unit to the camera may be determined based on the moving speed and the zoom state of the camera at the time when the image is acquired. State, the motion is not operated, etc.) and converts the movement to the automatic tracking control unit, and converts the position adjustment information provided by the automatic tracking control unit according to the current moving speed and the zoom state of the driving unit. It will generate a signal to control.

In the above described and illustrated with respect to preferred embodiments according to the present invention. However, the present invention is not limited to the above-described embodiments, and various changes and modifications may be made by those skilled in the art without departing from the scope of the present invention. .

1 is a block diagram of a target coordinate tracking video surveillance system by position recognition according to the present invention.

2 is a cross-sectional view of the tracking video surveillance apparatus according to the present invention,

Figure 3 is a separated configuration of the tracking image surveillance apparatus according to the invention of FIG.

<Description of the symbols for the main parts of the drawings>

Optical means 20, camera 21, gyro sensor 22, GPS sensor 23, GPS antenna 24, the first rotary drive unit 30, the rotary motor 31, position detection sensor 32, the first port unit 35, the second rotary drive unit 40, rotary motor 41 Position detection sensor 42, second port portion 45, storage means 50, drive signal output portion 51, image processing portion 52

Claims (2)

A camera capable of moving according to a control signal; An image detector which acquires motion coordinates from an image provided by the camera and determines a motion coordinate different from a moving direction of the camera; A gyro sensor unit configured to provide position adjustment information for positioning an object moving according to an input movement degree at a center of the image; In the target coordinate tracking video surveillance system by position recognition comprising a GPS sensor unit for checking the position of the camera to recognize the exact position even in the zoom ratio of the camera, The image detection unit calculates and outputs a position and movement coordinate value of a moving target by comparing an image provided by the camera with a previous or subsequent image in coordinate units, The gyro sensor unit recognizes the coordinates of the moving target to calculate and output the coordinates of the position of the target, The GPS sensor unit calculates and outputs a coordinate value of the camera position, The control unit receives the moving position and coordinate values of the target from the image detector, the gyro sensor unit, and the GPS sensor unit to classify the noise and the monitoring target based on the position and the direction of the moving target, and thus the center position and the moving coordinate of the monitoring target. Target coordinate tracking video surveillance system by position recognition, characterized in that it comprises an automatic tracking control unit for calculating a. One end is fixed to the camera 21 to shoot in the axial direction, the gyro sensor 22 to recognize the coordinates of the target set in the front and the GPS sensor 23 associated with the GPS antenna 24 received from the satellite in the rear is installed Optical means unit 20; A first rotational drive unit 30 rotatably installed on the axis of the camera 21 and having a position detection sensor 32 for sensing a position of the rotation motor 31; A power source, an image, and a communication connection mounted on the first rotation driving unit 30 to drive the camera 21, the GPS sensor 22, and the gyro sensor 23 of the optical means unit 20. 1 port part 35; A second rotation driver 40 having a position detection sensor 42 for detecting a position of the rotation motor 41 to rotate left / right in association with a lower portion of the first rotation driver 30; The camera 21 generates driving signals of the first rotation driver 30 and the second rotation driver 40 based on the coordinate data set for photographing the target, and transmits the driving signals to the corresponding rotation drivers 30 and 40, respectively. A storage means (50) provided with a driving signal output unit (51) and an image processor (52) for storing an image of the camera; And a second port part (45) mounted on the second rotation driving part (40) to connect power, video and communication of the storage means to the outside.

Images