KR20170001872A - Statistical analysis apparatus and method using pattern analysis in omni-directional camera images - Google Patents
Statistical analysis apparatus and method using pattern analysis in omni-directional camera images Download PDFInfo
- Publication number
- KR20170001872A KR20170001872A KR1020150091217A KR20150091217A KR20170001872A KR 20170001872 A KR20170001872 A KR 20170001872A KR 1020150091217 A KR1020150091217 A KR 1020150091217A KR 20150091217 A KR20150091217 A KR 20150091217A KR 20170001872 A KR20170001872 A KR 20170001872A
- Authority
- KR
- South Korea
- Prior art keywords
- image
- smart
- unit
- specific object
- statistical analysis
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
- G08B13/19608—Tracking movement of a target, e.g. by detecting an object predefined as a target, using target direction and or velocity to predict its new position
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19617—Surveillance camera constructional details
- G08B13/1963—Arrangements allowing camera rotation to change view, e.g. pivoting camera, pan-tilt and zoom [PTZ]
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19654—Details concerning communication with a camera
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19665—Details related to the storage of video surveillance data
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B3/00—Audible signalling systems; Audible personal calling systems
- G08B3/10—Audible signalling systems; Audible personal calling systems using electric transmission; using electromagnetic transmission
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B5/00—Visible signalling systems, e.g. personal calling systems, remote indication of seats occupied
- G08B5/22—Visible signalling systems, e.g. personal calling systems, remote indication of seats occupied using electric transmission; using electromagnetic transmission
- G08B5/36—Visible signalling systems, e.g. personal calling systems, remote indication of seats occupied using electric transmission; using electromagnetic transmission using visible light sources
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Electromagnetism (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
In the present invention, in the case of the conventional CCTV surveillance system, since only a fixed radius is localized, a plurality of cameras must be installed, a rectangular area is generated according to the environment of the surveillance area, There is a problem in that the operation of tracking the object is delayed by the camera due to an increase in the amount of computation for the object and there is no function of transmitting the automatic warning broadcasting according to the situation alarm information generated in the field, The smart omnidirectional camera device 100, the data communication interface unit 200 and the smartimage and statistical analysis server 300. In this case, instead of a plurality of CCTVs installed in one place, The object can be generated as a 360-degree donut-type panoramic image and can be monitored in real time, and the remote smart image / statistical analysis server It is possible to continuously detect and track specific objects that are behaving abnormally in real time continuously without any interruption. It can accumulate moving trajectory patterns of specific objects in real time and select one of image color map and statistical graph, It is possible to control the automatic warning broadcasting to be transmitted according to the event situation of the specific object generated in the field, and it is possible to control only the program in addition to the existing system so that it is compatible with other neighboring smart It is possible to track the position of a specific object in real time by linking with statistical analysis device and ubiquitous sensor network (USN), and it is possible to improve the security effect by omnidirectional camera based on intelligent image analysis technology by 80% To provide a smart statistical analysis apparatus and method using the moving trajectory pattern analysis To have its purpose.
Description
In the present invention, instead of a plurality of CCTVs installed in one place, a 360-degree donut-type omnidirectional image can be generated with one camera and real-time monitoring can be performed. It is possible to continuously detect and track a specific object in real time continuously without interruption and to accumulate moving trajectory pattern of a specific object in real time to select one of image color map and statistical graph to display statistical analysis report by place and time And more particularly, to an apparatus and method for smart statistical analysis using motion trajectory pattern analysis in an omnidirectional camera image.
Recently, due to the frequent occurrence of violent crime, social anxiety has increased and interest in individuals and public safety has increased.
For this reason, the importance of the development and application of video surveillance system for prevention, post-analysis and quick resolution of crime is increasing.
In image surveillance system, image quality improvement, image transmission and storage technology have developed much more than in past systems.
Also, by developing a network camera connected to the Internet, it is possible to perform video surveillance remotely.
These systems are widely applied to the peripheral technology, and as the price is relatively low, it is possible to widely distribute the effective video surveillance system.
CCTV (Closed Circuit Television) is installed in a place requiring security such as a public place, a hunting area, and an access control area using a wide-area camera, and the screen image is directly monitored by the monitoring personnel or recorded in a storage device, And the like.
However, since the conventional CCTV surveillance system only performs local area surveillance for a fixed radius, a large number of cameras must be installed for a wide area surveillance, and a rectangular area is generated depending on the environment of the surveillance area.
In addition, even if a plurality of CCTV cameras are installed, there is a problem that the operation amount of the CCTV cameras is interlocked and the operation schedule is increased, so that the operation of tracking the object by the camera is delayed.
In addition, in the case of a conventional CCTV camera, there is a problem that it is impossible to prevent a criminal act in advance because there is no function of transmitting an automatic warning broadcast according to situation alarm information generated in the field.
In order to solve the above problems, instead of a plurality of CCTVs installed in one place, an object can be generated as a 360-degree donut-type omni-directional image with a single camera and can be monitored in real time, The analysis server can continuously detect and track specific objects that are behaving abnormally in real time continuously without interruption and select statistical analysis report by selecting one of image color map and statistical graph by accumulating moving trajectory pattern of specific object in real time. The present invention provides an apparatus and method for smart statistical analysis using motion trajectory pattern analysis in omnidirectional camera images that can be displayed by place and time zone.
In order to achieve the above object, a smart statistical analysis apparatus using motion trajectory pattern analysis in an omnidirectional camera image according to the present invention comprises:
The identification ID is set and the object is photographed 360 degrees in all directions with one camera while being located at a specific place and then the taken donut type omni-directional image is transmitted to the remote smart image / statistical analysis server through the data communication interface unit A smart
A data
And a data communication interface unit for receiving a donut-type omnidirectional image transmitted from the smart omnidirectional camera device to convert the imaged omnidirectional image, detecting a specific object, tracking the position of the specific object based on the movement trajectory pattern of the specific object, And a smart image /
As described above, in the present invention,
First, in place of installing a plurality of CCTVs in one place, a 360 ° donut type omnidirectional image can be generated and monitored in real time by a single camera, so that the surveillance range can be improved by 70% The installation cost can also be reduced by 80% compared to the existing one.
Secondly, the distorted image of the rectangular panoramic image conversion unit can be corrected through the image distortion correction unit, the image quality can be improved, and a high-quality image can be provided.
Third, after detecting a specific object, it is possible to track the position of the ideal object based on the movement trajectory pattern of the specific object and to control the automatic warning broadcast according to the event situation of the specific object generated in the field, The crime rate can be reduced by 70%.
Fourth, it is possible to interoperate only the program in addition to the existing system, and it is excellent in compatibility.
Fifth, the moving trajectory pattern of a specific object can be stacked in real time, and the statistical analysis report can be displayed in each place and time zone by selecting one of the image color map and the statistical graph, and the whole security effect based on the intelligent image analysis technology Can be improved by 80%.
FIG. 1 is a block diagram showing components of a smart
FIG. 2 is a block diagram showing components of a smart omnidirectional camera device according to the present invention;
FIG. 3 is a perspective view showing an outer appearance of an omnidirectional camera module including a lens unit and a parabolic reflector according to the present invention,
FIG. 4 is a block diagram illustrating the components of the omnidirectional camera module according to the present invention.
FIG. 5 is a block diagram illustrating components of a smart image / statistical analysis server according to the present invention.
FIG. 6 is a block diagram illustrating components of a rectangular panoramic image conversion unit according to the present invention.
FIG. 7 is a block diagram showing a configuration of a 4-channel image converting unit according to the present invention;
8 is a diagram illustrating an embodiment of converting a donut-type omni-directional image into a rectangular panoramic image having four channels by using a coordinate transformation formula through a four-channel image conversion unit according to the present invention.
FIG. 9 is a diagram illustrating an embodiment of mapping a pixel value in the nearest position among a total of nine pixel values calculated through the coordinate operation unit in the 9-point interpolation algorithm engine unit according to the present invention.
FIG. 10 is a block diagram showing the components of the smart control unit according to the present invention.
11 is a diagram illustrating a histogram of the distribution of brightness of a local region through a HOG object detection unit according to an embodiment of the present invention,
FIG. 12 is a diagram illustrating an object moving trajectory analyzing unit according to an embodiment of the present invention that detects an object in a real-time input image using a Gaussian Mixture Model and extracts a noise-canceled foreground image using a morphology operation In an embodiment showing a result screen,
FIG. 13 is a diagram illustrating an example of accumulating movement trajectories of an object movement trajectory analysis unit according to the present invention.
14A is a diagram illustrating a result of a heat map (calculated in units of 10 minutes) in a binary image calculated through an accumulation value of each pixel through an object movement trajectory analysis unit according to an embodiment of the present invention.
FIG. 14B is a diagram illustrating a result of a heat map expressed in a background image to be displayed to a user through an object movement trajectory analysis unit according to an embodiment of the present invention.
FIG. 15 is a diagram illustrating a result of a heat map 30 minutes after the beginning of an image analysis expressed in a background image of a divided image to be displayed to a user through an object movement trajectory analysis unit according to the present invention.
FIG. 16 is a graph showing the result of a graph obtained in 10-minute increments based on the start of image analysis through the object movement trajectory analysis unit according to the present invention.
17 is a block diagram showing the components of the statistical analysis control unit according to the present invention,
FIG. 18 is a block diagram showing the components of the movement trajectory accumulation forming unit according to the present invention,
FIG. 19 is a diagram illustrating an example of accumulating movement trajectories of a specific object through the moving trajectory accumulation forming unit according to the present invention.
20 shows an embodiment showing accumulating pixels through the first accumulation mode of the moving trajectory accumulation forming unit according to the present invention,
FIG. 21 is a diagram illustrating accumulation of pixels through the second accumulation mode of the moving trajectory accumulation forming unit according to an embodiment of the present invention.
22 is a diagram illustrating accumulation of pixels through the third accumulation mode of the trajectory accumulation forming unit according to the present invention,
FIG. 23 is a diagram illustrating a method of actually accumulating pixel units in a third cumulative formation mode of the moving trajectory accumulation forming unit according to the present invention,
FIG. 24 shows an embodiment in which each RGB channel is allocated and configured in the JET spectrum through the JET spectrum
25 shows an embodiment showing the display of a color map on an image through the image color map
FIG. 26 is a diagram illustrating a second statistical analysis report formed by an RGB statistic graph by accumulating specific object movement trajectory information in real time through a statistical graphical display unit according to the present invention,
27 is a flowchart showing a smart statistical analysis method using an analysis of a moving trajectory pattern in an omnidirectional camera image according to the present invention.
Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram illustrating components of a smart
First, the smart
The smart
2, the
First, the
The
Second, the
The
As shown in FIGS. 3 and 4, the
The
As shown in FIG. 3, one aspheric surface single optical lens capable of wide-angle photographing is formed on one side of an inner space of a body formed of a longitudinal dome structure.
The
The
This forms an AR coating on the surface.
Here, the AR coating removes the infrared region band to prevent light saturation, thereby increasing the transmittance of the lens.
Using the coherence of light, it improves the transmission characteristic in a desired wavelength region by attaching a film having a thickness smaller than the wavelength.
The
This is a device that converts light into an electrical signal. Since the charges generated in the light receiving part sequentially move through the MOS capacitors connected in series, the voltage changes from the source follower to the voltage, so a high voltage power source is used.
A separate chip is included in the light receiving unit.
The CCD sensor unit according to the present invention may cause incidence of whitening and smear due to incidence of an excessive amount of light.
In order to prevent this, a filter for blocking infrared rays in the vicinity of 650 nm to 1500 nm is used.
As an infrared cut filter, an IR cut filter and an IR absorption filter are constituted.
The IR cut filter is formed by an optical coating method and generates a wavelength shift according to an incident angle, thereby blocking infrared rays.
The IR absorption filter serves to absorb an infrared wavelength.
Third, the
The
This is because a first event mode in which a warning sound is transmitted to the surroundings together with a voice message indicating "thief" when the event situation of a specific object is an event related to the theft and theft, A second event mode in which a warning sound is transmitted to the surroundings together with a voice event that "an external person has entered", and a second event mode in which when the event situation of a specific object is an event related to sexual crime, Quot ;, and a third event mode in which a warning sound is transmitted to the surroundings is selected and transmitted.
Fourth, the toroidal omnidirectional
The donut-type omnidirectional
Next, the data
The data
This is configured by selecting either a WiFi communication interface, an RS 485 communication interface, or an RS 232 communication interface.
Next, the smart image and
The smart image and
5, the apparatus includes a donut type panoramic
First, a description will be given of a donut type omnidirectional
The donut type omnidirectional
Second, the rectangular panorama
The rectangular panorama
As shown in FIG. 6, the image forming apparatus includes a 4-channel
[Four-Channel Image Conversion Unit 321]
The 4-channel
This traces the concentric circle line of each pixel of the toroidal omni-directional image to form a coordinate mapping relationship between the toroidal omni-directional image and the rectangular panoramic image.
That is, FIG. 7 is a block diagram illustrating a configuration of a 4-channel image converting unit according to the present invention, wherein R denotes a radius of an inner circle, and 2R denotes a donut-type omnidirectional image having a radius of an outer circle.
An area between the inner radius R and the
As shown in Fig. 8, the donut-type omnidirectional image is converted into a rectangular panoramic image composed of four channels using a coordinate conversion formula.
The mapping relation between the donut type panorama image and the rectangular panorama image exists in a transformation relation from the polar coordinate system to the rectangular coordinate system.
The coordinate transformation relation between two images can be represented by the parameters (r i , θ i ) and (U i , V i ) and can be expressed by the following equations (1) and (2).
The conversion relations between the coordinates (X i , Y i ) and (U i , V i ) of the two images can be expressed by Equations (3) and (4) using Equations (1) and
[Coordinate operation unit 322]
The coordinate
This determines the center coordinate value of the toroidal omnidirectional image, divides the toroidal omnidirectional image into four symmetric regions, divides the corresponding rectangular panoramic image into four regions, And calculate the coordinate value.
If the coordinate value is calculated using Equations (3) and (4), the amount of calculation becomes large and the hardware complexity becomes very large.
At this time, in the present invention, the hardware complexity is reduced through the following process, and the realization is realized while reducing the amount of calculation.
First, the center coordinate value of the donut type panoramic image is determined, and the size of the rectangular panoramic image is determined.
Second, the donut-type panoramic image is divided into four symmetric regions, and the corresponding rectangular panoramic image is divided into four regions.
Third, one of the four regions is determined as a calculation region, and coordinate values are calculated using Equations (3) and (4).
Here, the remaining three regions are calculated by the symmetric coordinate method using the calculated coordinate values again, and can be expressed by the following equation (5).
Since the coordinate values of the remaining three areas can be simultaneously calculated through the coordinate calculation unit according to the present invention and the calculation formula is very simple, it is possible to quickly convert the coordinates of the coordinate conversion unit, and the size of the memory used for the look- There is a number.
[Look-up table generating unit 323]
The lookup
This generates LUT_X that stores the value associated with X and LUT_Y that stores the value associated with Y in equations (3) and (4).
The lookup
Further, only a quarter of the rectangular panoramic image size is stored using the symmetry property of the toroidal omnidirectional image.
Third, the image
The image
This consists of a 9-point interpolation algorithm engine unit that maps 9 pixel values to pixel values nearest to 9 pixel values.
That is, all interpolated pixel values can be calculated as an average of neighboring actual pixel values.
For example, the interpolated pixel value (i, j + 0.5) can be represented by an average value of the actual pixel values (i, j) and (i, j + 1).
The pixel value i + 0.5, j + 0.5 is expressed as an average value of the actual pixel values i, j, i + 1, j, i, j + 1, .
The interpolated pixel value is expressed by the following equation (6).
The coordinate values calculated through the coordinate calculation unit 220 according to the present invention are mapped to pixel values closest to the nine pixel values.
In Fig. 9, the red triangle is mapped to a value of (i + 0.5, j) in the closest position.
The 9-point interpolation algorithm engine unit according to the present invention can be represented by only two pixel values or an average value of four pixel values.
As described above, the engine part of the 9-point interpolation algorithm can reduce the computational complexity without degrading the quality of the converted panoramic image of the rectangle, and has a very easy characteristic in realizing real-time low-cost hardware.
Fourth, the
The
10, the HOG type
[
HOG
Type object detection unit 341]
The HOG-type
As shown in FIG. 11, the distribution direction of the brightness of the local region is expressed by a histogram, and is used as a feature vector, which is suitable for expressing the shape of an object.
In order to obtain the Histogram of Gradient (HOG) feature, it is necessary to first calculate the slope magnitude value and the directional value.
Assuming that the brightness value of each pixel is f (i, j), the slope magnitude m (i, j) and the direction? (I, j) are expressed by Equation (7).
If a slope size and directional information are known, a directional histogram for the brightness change of a cell region composed of 8x8 pixel regions can be calculated.
Considering only the directionality of the tilt magnitude, θ (i, j) is divided into 20 ° intervals of 0 ° to 180 °, and 9 directions per cell are used to calculate m (i, j) And is configured to have a histogram Bin (Bin).
Then, the gradient magnitude values are accumulated in each histogram bin to calculate the distribution.
Also, one block includes 3 x 3 cells, and each block is configured to normalize.
Since there are nine directional histogram bins in one cell and nine cells in one block, the number of histogram bins in one block is 81. The feature values for each block are calculated while sliding one cell in the scanning direction.
For each block, the feature p is normalized using the following equation (8).
Here, k is the number of features in the block, and ε is set to 1 to prevent calculation failure when the denominator becomes zero.
[Object movement locus analyzer 342]
The object movement
Specific objects related to abnormal behavior, person, and vehicle among the objects are newly appearing and disappearing on the screen in the image, and numerous position shifts occur.
Accordingly, in the object movement trajectory analyzing unit according to the present invention, specific objects related to abnormal behavior, people, and vehicles should be assigned to each unique ID through labeling, and each object should be tracked.
Certain objects with unique IDs will lose their unique ID if they disappear from the image or if tracking fails.
At this time, in the object movement trajectory analysis unit according to the present invention, if the same specific object fails to be traced in the middle, it is recognized as a new specific object and a new ID is given.
FIG. 12 is a diagram illustrating an object moving trajectory analyzing unit according to an embodiment of the present invention that detects an object in a real-time input image using a Gaussian Mixture Model and extracts a noise-canceled foreground image using a morphology operation And also to an embodiment showing a result screen.
Next, in order to analyze the movement trajectory of a specific object, a space in which a specific object is located is set in advance.
The location of a particular object means the pixel location in the image.
The object movement trajectory analysis unit according to the present invention performs tracking using the location, size, and type information of a specific object performed through the smart object tracking control unit, and stores the trajectory information until the tracking fails.
FIG. 13 is a diagram illustrating an accumulation of movement trajectories through an object movement trajectory analysis unit according to an embodiment of the present invention. In FIG. 13, an
The motion trajectory is obtained by continuously connecting the blob center blocks of each calculated frame, and the movement trajectory of the object is calculated by connecting the center of the foreground portion through the same result as (c).
The trajectory is a straight line because it is the connection between the center coordinates of the object.
Therefore, in the present invention, the Bezier curve method of approximating the curve shape is applied.
The Bezier curve is an n-th order curve obtained from n points, and a curve of a cubic Bezier is used as shown in Equation (9) below to generate a smooth curve.
The object movement
For example, in the present invention, binary images of one frame per second are accumulated in consideration of the image size of an omnidirectional camera module of at least 3 Megapixels and the amount of computation required for image analysis, and cumulative values of trajectories are calculated based on a maximum of one hour.
For example, if the heat map is accumulated one time per second, the maximum is 3600 for one hour.
If the color map is divided by 360, which is the value of the color map, the heat map has a color map value of 10 units.
The heat map calculated in units of 10 minutes by applying this is shown in FIG.
FIG. 14 (a) relates to an embodiment showing a heat map result in a binary image calculated through the cumulative value of each pixel, and FIG. 14 (b) And in one embodiment showing a heat map result.
FIG. 15 is a diagram showing an example of a heat map result 30 minutes after the start of image analysis expressed by being displayed on a background image of a divided image to show to a user.
As described above, when the image is represented by a color map through the object movement trajectory analyzing unit 420 according to the present invention, it is possible to express at which position the movement trajectory is large.
However, it is difficult to know whether there are a lot of specific objects at a certain time in the place where the camera is installed as a heat map which expresses colors in the image.
Therefore, it is possible to provide the analysis data about the number of the floating population at a certain time by expressing the information which is difficult to express by the color map.
FIG. 16 is a diagram illustrating graphical results obtained in units of 10 minutes based on the start of image analysis through the object movement trajectory analysis unit according to the present invention. Since the graph is also a connection between integer values, see.
Therefore, the Bezier curve method is applied to approximate this to a curve shape.
The results of the graph using the Bezier curves show that the floating population was the largest for 10 minutes after the start of the analysis.
[Smart Object Tracking Control Unit 343]
The smart object
In tracking a specific object, there is no restriction according to the tracking form of a specific object, and by using the color information, the tracking characteristic of a specific object is stronger than that of other tracking methods.
Table 1 below is a table describing object tracking result data and contents through the smart object tracking controller according to the present invention.
Number of frames
Tracking
Number of frames
error
Number of frames
Error rate
[Statistical analysis control unit 344]
The statistical
As shown in Fig. 17, this is constituted by a moving locus
The movement trajectory
This accumulates the data of a specific object in order to know the movement trajectory of a specific object for image analysis.
In order to accumulate the movement trajectory information of the specific object, the cumulative data of the unit of time and the unit of the day are displayed in order to accumulate the partial pixels passing through the object in the whole image.
The moving trajectory accumulation forming unit according to the present invention includes a first
FIG. 19 is a diagram illustrating an embodiment for accumulating movement trajectories of a specific object through the movement trajectory accumulation forming unit according to the present invention. Referring to FIG. 20, a moving trajectory in which a specific object moves in a first input image and a second input image, FIG. 21 relates to an embodiment showing accumulating pixels through a second accumulation mode, FIG. 22 relates to an embodiment showing accumulation of pixels through a third accumulation mode Lt; RTI ID = 0.0 > pixel < / RTI > mode.
FIG. 23 is a view showing how a pixel is actually accumulated on a pixel-by-pixel basis through the third accumulation mode according to the present invention, in which a certain block in the central portion of the block moves over time Accumulate one pixel for the block.
The initial value of each pixel is initialized to 0, and the central block of the blob accumulates 1 to the past pixels. As the pixel passes through the central block of the blob over time, the cumulative value of the pixel increases do.
If the entire blob is used, the blob is accumulated by 1 based on all the pixels passing by. In the case of using the blob center pixel, the cumulative value is applied to the past pixels according to the movement locus of the central pixel.
These cumulative values are used when a cumulative value of each pixel is converted into a color map in the case where the cumulative value is displayed in a color map by statistical analysis of the moving trajectory.
When representing cumulative data of a moving trajectory on a two-dimensional space, it can be represented by more effective data using color mapping.
The movement locus
In the JET spectrum of the JET spectrum
That is, the JET spectrum
The part where the movement trajectory of the specific object is not present is not represented by the color but the part where the movement trajectory of the specific object is present but the movement trajectory is the smallest is represented as dark blue having only the blue channel 127, Most places are represented in dark red, so that they can be seen at a glance.
The image color map
This is done by applying an arbitrary threshold to the cumulative value of each pixel in the cumulative database so that the excluded pixel does not process anything except for a value that does not meet this threshold and sets the pixel with the minimum value to dark blue, Is set to dark red, the difference between the minimum value and the maximum value is divided into a predetermined range, and the color map is superimposed on the image in the color of the JET spectrum formed in the JET spectrum color distribution portion.
The image color map
25 is a view showing an embodiment of displaying a color map on an image through an image color map
The threshold value is set to 5 so that no pixel is assigned to the pixel accumulation value having a pixel value of 5 and a color ranging from dark red to dark blue is given to a pixel having a value ranging from 6 to 15, To display a color map on the image.
The statistical graph
This is configured to display the RGB statistic graph calculated in units of time and to display the number of movement trajectories at any time.
The RGB statistical graph calculation method is calculated by calculating the sum of the accumulated data through the movement trajectory analysis in real time and displaying it on the graph.
If there is no specific object moving in the space, the Y axis value of the graph is set to 0, and if there are many moving objects, the Y axis value is set high.
In this case, the Y-axis value of the data is added in units of one hour based on the time of the image analysis server, so that the cumulative value accumulated for one hour is calculated and displayed in a graph.
FIG. 26 is a diagram showing a second statistical analysis report formed by real-time accumulation of specific object movement trajectory information through a statistical graph type display unit according to the present invention and formed by an RGB statistical graph.
[Warning broadcast transmission control unit 345]
The alert
Here, a 1: 1 customized automatic warning broadcast sends a warning message "Catch thieves" when it is a specific object that runs away, and when it is a specific object of garbage dumping, a warning message "Do not dump the garbage ~ In case of a specific object to be roamed, it is configured to send a warning message "This is a place where the security monitoring system is operated, please move to another place ".
Hereinafter, a specific operation process of the smart statistical analysis method using the motion trajectory pattern analysis in the omnidirectional camera image according to the present invention will be described.
First, as shown in FIG. 27, a specific object is photographed 360 degrees in all directions with a single camera through a smart omnidirectional camera device (S100).
Next, the donut type panoramic image taken by the smart panoramic camera device is transmitted to the smart image / statistical analysis server through the data communication interface unit (S200).
Next, the smart image and statistical analysis server receives the donut-type panoramic image transmitted from the smart omnidirectional camera device, detects the specific object by image conversion, and then tracks the position of the specific object based on the movement trajectory pattern of the specific object (S300).
Next, the smart image / statistical analysis server traces the position of a specific object based on the movement trajectory pattern of a specific object, accumulates the movement trajectory pattern of the specific object in real time, and displays the statistical analysis report by place and time zone (S400 ).
Next, the warning broadcast transmission control unit of the smart image / statistical analysis server transmits a customized warning broadcast signal according to the event status of the specific object to the smart omnidirectional camera device in the field (S500).
Finally, the smart omnidirectional camera device receives the warning broadcast signal from the alarm broadcast control unit of the smart image / statistical analysis server and transmits a warning broadcast according to the event status of the specific object (S600).
1: Smart statistical analysis device 100: Smart omnidirectional camera device
110: main body 120: omnidirectional camera module
130: Speaker unit 140: Donut type omnidirectional image transmitter
200: Data communication interface unit 300: Smart image / statistical analysis server
310: a donut type panoramic image receiving unit 320: a rectangular panoramic image converting unit
330: image distortion correction unit 340: smart control unit
Claims (7)
A data communication interface unit 200 for connecting the smart omnidirectional camera device and the smart image / statistical analysis server to each other through wired / wireless communication lines to form a bidirectional data communication network,
And a data communication interface unit for receiving a donut-type omnidirectional image transmitted from the smart omnidirectional camera device to convert the imaged omnidirectional image, detecting a specific object, tracking the position of the specific object based on the movement trajectory pattern of the specific object, And a smart image and statistical analysis server (300) for accumulating the movement trajectory pattern of the object in real time and displaying the statistical analysis report by place and time zone. The smart statistical analysis using the trajectory pattern analysis in the omnidirectional camera image Device.
A main body 110 formed in a longitudinal dome structure for protecting and supporting each device from external pressure,
An omnidirectional camera module 120 positioned on the head portion of the main body and generating a toroidal omnidirectional image,
A speaker unit 130 positioned at a rear end of the main body and driven according to a control signal of the smart image and statistical analysis server to transmit a warning broadcast according to an event condition of a specific object,
And a donut-type omnidirectional image transmitting unit (140) located at one side of the inner space of the main body and connected to the data communication interface unit to transmit a donut type omnidirectional image taken by the omnidirectional camera module Smart statistical analysis device using motion trajectory pattern analysis.
A lens unit 121 for introducing the light coming from the 360 degree field of view into the parabolic reflector,
A parabolic reflection plate 122 for reflecting the light introduced from the lens unit to the CCD sensor unit,
And a CCD sensor unit (123) that focuses the light reflected from the parabolic reflector to generate a donut type panoramic image.
A donut type omnidirectional image receiving unit 310 connected to the data communication interface unit and receiving a donut type omnidirectional image transmitted from the smart omnidirectional camera device,
A rectangle panorama image converting unit 320 for converting a donut type panorama image received by the donut type panorama image receiving unit into a panorama image of a rectangle by panorama unrolling,
An image distortion correction unit 330 for correcting the distorted image of the rectangular panorama image conversion unit and improving the image quality,
After detecting a specific object on the image-corrected image through the image distortion correction unit, the position of the specific object is tracked based on the movement trajectory pattern of the specific object, and the movement trajectory pattern of the specific object is accumulated in real time, , And a smart control unit (340) for controlling the smart control unit (340) to control the smart control unit (340) to display the images according to time zones.
A four-channel image conversion unit 321 for converting a donut-type panoramic image into a rectangular panoramic image composed of four channels using a coordinate conversion formula,
The toroidal omnidirectional image data is divided into four regions, one of which is used to calculate coordinates using a coordinate conversion formula, and the remaining three regions are used to calculate three coordinate values simultaneously using the symmetry property to generate a lookup table A coordinate calculating unit 322 for storing the data in advance,
A lookup table generating unit 323 for calculating coordinate values between the toroidal omnidirectional image data and the rectangular panoramic image in advance and storing them in a lookup table and loading a lookup table value corresponding to the currently input donut type omnidirectional image data value Wherein the motion vector of the camera is determined based on the motion trajectory pattern.
An HOG-type object detection unit 341 for detecting an object in an image using a Histogram of Gradient (HOG)
An object movement trajectory analysis unit 342 for analyzing repetitive movement trajectory information of a specific object related to an abnormal behavior, a person, and a vehicle as a result of object detection after removing the background of the object among the objects detected by the HOG type object detection unit, ,
A smart object tracking control unit for tracking a Kernel Based object in a hybrid method comprising ROI (Region of Interest), which is a result of detection by the HOG type object detection unit, and specific object movement trajectory information of the object movement trajectory analysis unit 343,
A statistical analysis control unit 344 for accumulating the specific object movement trajectory information tracked by the smart object tracking control unit 153 in real time and controlling the statistical analysis report to be displayed in each place and time zone,
And a warning broadcast transmission control unit (345) for controlling the customized warning broadcast signal according to an event situation of a specific object tracked by the smart object tracking control unit to be transmitted to the smart omnidirectional camera device in the field. Smart statistical analysis device using motion trajectory pattern analysis.
A step S200 of transmitting a donut-type omni-directional image taken by a smart omnidirectional camera device to a remote smart image / statistical analysis server through a data communication interface unit,
Receiving a donut type omni-directional image transmitted from a smart omni-directional camera device in a smart image / statistical analysis server, detecting a specific object by image conversion, and tracking the position of a specific object based on a movement trajectory pattern of the specific object (S300 )Wow,
A step S400 of displaying a statistical analysis report by location and time zone by tracking the position of a specific object on the basis of a moving trajectory pattern of a specific object in the smart image and statistical analysis server and accumulating moving trajectory patterns of a specific object in real time, ,
A step S500 of sending a customized warning broadcast signal according to an event situation of a specific object to a smart omnidirectional camera device in the field,
(S600) of receiving a warning broadcast signal from a warning broadcast control unit of a smart image / statistical analysis server in a smart omnidirectional camera device and transmitting a warning broadcast according to an event condition of a specific object Smart statistical analysis device using motion trajectory pattern analysis.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020150091217A KR101729906B1 (en) | 2015-06-26 | 2015-06-26 | Statistical analysis apparatus and method using pattern analysis in omni-directional camera images |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020150091217A KR101729906B1 (en) | 2015-06-26 | 2015-06-26 | Statistical analysis apparatus and method using pattern analysis in omni-directional camera images |
Publications (2)
Publication Number | Publication Date |
---|---|
KR20170001872A true KR20170001872A (en) | 2017-01-05 |
KR101729906B1 KR101729906B1 (en) | 2017-05-04 |
Family
ID=57835549
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
KR1020150091217A KR101729906B1 (en) | 2015-06-26 | 2015-06-26 | Statistical analysis apparatus and method using pattern analysis in omni-directional camera images |
Country Status (1)
Country | Link |
---|---|
KR (1) | KR101729906B1 (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101881524B1 (en) * | 2018-04-23 | 2018-07-24 | 박영기 | The apparatus and method of video surveillance system using intelligent based location information and action cognitive information in subject for video |
KR101884611B1 (en) * | 2017-02-23 | 2018-08-02 | 이노뎁 주식회사 | Method of extracting objects-of-Interest out of CCTV video using metadate filtering of moving objects |
KR101939096B1 (en) | 2018-06-07 | 2019-01-16 | 쿠도커뮤니케이션 주식회사 | Apparatus for Providing Trajectory and Driving Method Thereof |
KR101956008B1 (en) | 2018-06-07 | 2019-03-08 | 쿠도커뮤니케이션 주식회사 | Apparatus for Providing Trajectory and Data Building Method Thereof |
KR20190060161A (en) * | 2017-11-24 | 2019-06-03 | 주식회사 리얼타임테크 | Apparatus for analyzing Multi-Distributed Video Data |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10769906B2 (en) | 2017-08-31 | 2020-09-08 | Hanwha Techwin Co., Ltd. | Doorbell, security system including doorbell apparatus, and method of operating the same |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101360246B1 (en) | 2013-08-12 | 2014-02-11 | 김규미 | The multi-functional anti-crime camera system |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101442669B1 (en) * | 2013-03-22 | 2014-09-24 | 에프엔엔(주) | Method and apparatus for criminal acts distinction using intelligent object sensing |
KR101377029B1 (en) * | 2013-11-29 | 2014-03-20 | 주식회사 에보시스 | The apparatus and method of monitoring cctv with control moudule |
-
2015
- 2015-06-26 KR KR1020150091217A patent/KR101729906B1/en active IP Right Grant
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101360246B1 (en) | 2013-08-12 | 2014-02-11 | 김규미 | The multi-functional anti-crime camera system |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101884611B1 (en) * | 2017-02-23 | 2018-08-02 | 이노뎁 주식회사 | Method of extracting objects-of-Interest out of CCTV video using metadate filtering of moving objects |
US10271018B2 (en) | 2017-02-23 | 2019-04-23 | Innodep Co., Ltd. | Method of detecting critical objects from CCTV video using metadata filtering |
KR20190060161A (en) * | 2017-11-24 | 2019-06-03 | 주식회사 리얼타임테크 | Apparatus for analyzing Multi-Distributed Video Data |
KR101881524B1 (en) * | 2018-04-23 | 2018-07-24 | 박영기 | The apparatus and method of video surveillance system using intelligent based location information and action cognitive information in subject for video |
KR101939096B1 (en) | 2018-06-07 | 2019-01-16 | 쿠도커뮤니케이션 주식회사 | Apparatus for Providing Trajectory and Driving Method Thereof |
KR101956008B1 (en) | 2018-06-07 | 2019-03-08 | 쿠도커뮤니케이션 주식회사 | Apparatus for Providing Trajectory and Data Building Method Thereof |
Also Published As
Publication number | Publication date |
---|---|
KR101729906B1 (en) | 2017-05-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR101729906B1 (en) | Statistical analysis apparatus and method using pattern analysis in omni-directional camera images | |
US8471892B2 (en) | Wide field-of-view reflector and method of designing and making same | |
KR101729907B1 (en) | Security surveillance apparatus and method using pattern analysis in omni-directional camera images | |
AU2012255691B2 (en) | Surveillance system | |
US7366359B1 (en) | Image processing of regions in a wide angle video camera | |
CN114093142B (en) | Object-perceived temperature anomaly monitoring and early warning by combining visual sensing and thermal sensing | |
US20050036036A1 (en) | Camera control apparatus and method | |
US20030071891A1 (en) | Method and apparatus for an omni-directional video surveillance system | |
KR100976441B1 (en) | Parking lot management system using omnidirectional camera and management method thereof | |
US9576335B2 (en) | Method, device, and computer program for reducing the resolution of an input image | |
US20120038776A1 (en) | Automatically Expanding the Zoom Capability of a Wide-Angle Video Camera | |
KR101502448B1 (en) | Video Surveillance System and Method Having Field of Views of 360 Degrees Horizontally and Vertically | |
CN101119482A (en) | Overall view monitoring method and apparatus | |
CN103929592A (en) | All-dimensional intelligent monitoring equipment and method | |
US20180367739A1 (en) | Rotation-adaptive video analytics camera and method | |
WO2018032270A1 (en) | Low complexity tamper detection in video analytics | |
EP3745718B1 (en) | Method of controlling pan-tilt-zoom camera by using fisheye camera and monitoring system | |
JP6456039B2 (en) | Surveillance camera system | |
Pawłowski et al. | Visualization techniques to support CCTV operators of smart city services | |
CN116052004B (en) | Bidirectional monitoring method and device for abnormal events, electronic equipment and storage medium | |
Tamaki et al. | An automatic compensation system for unclear area in 360-degree images using pan-tilt camera | |
US11825241B2 (en) | Camera and a method therein for facilitating installation of the camera | |
KR20180017329A (en) | Image Pickup Apparatus | |
US10102432B2 (en) | Image recognition method | |
TW201308219A (en) | Ultra-wide-angle imaging method and system using the same |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
A201 | Request for examination | ||
E902 | Notification of reason for refusal | ||
E701 | Decision to grant or registration of patent right | ||
N231 | Notification of change of applicant | ||
GRNT | Written decision to grant |