WO2017043933A1

WO2017043933A1 - Method for transmitting multiple image streams for improving background and object identification capability of ip camera

Info

Publication number: WO2017043933A1
Application number: PCT/KR2016/010202
Authority: WO
Inventors: 곽한섭; 하성식
Original assignee: 주식회사 유니시큐; 곽한섭
Priority date: 2015-09-11
Filing date: 2016-09-09
Publication date: 2017-03-16
Also published as: KR101598959B1

Abstract

The present invention takes photographs by setting parameters (40) of a camera (10) to be different for each real time frame (100) in one camera (10), and then classifies the frames (100) captured by the parameters (40) in the same condition so as convert the same into a plurality of image streams and transmit the same. Therefore, a method for transmitting multiple image streams for improving a background and object identification capability of an IP camera, according to the present invention, has remarkable effects of: enabling a method of a conventional server to be compatibly used without a change by using an IP address used by the conventional server; classifying the image streams on a per port basis so as to take photographs through one camera such that all objects such as a rapidly traveling vehicle and a slow moving person and bicycle can be identified without degrading image quality compared with that of a conventional camera; identifying both a bright region and a dark region, and controlling the regions through parameters different from each other so as to capture a plurality of image streams, in regions having a definite difference in the brightness of captured screens; and enabling a server to classify the captured image streams according to IP address and port and manage the same.

Description

Transmission Method of Multiple Video Streams for Improving Background and Object Identification of IP Cameras

The present invention relates to a method for transmitting multiple video streams for improving the background and object identification of an IP camera. By setting different parameters in the frame, you can get bright and dark images from a single camera, and adjust the shutter speed of the frame to identify the background of the IP camera that can identify both the number and the person running in the dark area. And a transmission method of multiple video streams for improving object identification capability.

According to the prior art, distributed processing-based camera image service system and method of Patent No. 10-1420006, Image storage DB (DataBaes) for providing a space for storing the image; An image storage server configured to receive images photographed by cameras and divide and store the images photographed by the cameras in the image storage DB; An image retrieval server for distributing and retrieving images stored in the image storage DB; An image analysis server for distributing and analyzing the images stored in the image storage DB; And a camera management server for transmitting a camera related command input through the user's mobile device to the camera of the user. And the image storage server stores an image photographed by the camera of the user according to a storage environment set by a user, wherein the storage environment includes a storage space distribution, and the storage space distribution is recently. The storage space allocated to the image is distributed larger than the storage space allocated to the past image, and the image storage server reconstructs the past image to reduce the capacity and then resaves the data. Is reduced to reduce the capacity of the past image, the camera management server automatically controls the camera operation of the user based on the location of the user, the location of the user is the user's mobile device is the camera of the user Is determined based on whether it was scanned by an AP (Access Point) located in a location where The image retrieval server retrieves a captured image of a camera associated with the user from the image storage DB and provides the image to a mobile device of the user. The cameras transmit images and location information together and store the image. The server stores the images and location information received from the cameras in the image storage DB, the image search server provides the image retrieved from the image storage DB with the location information to the user's mobile device, the mobile The device is described as a distributed processing-based camera video service system, characterized in that it displays the image and the map showing the location information together.

According to another prior art, an illegal parking automatic unmanned automatic enforcement method and system having a function of a crime prevention-based Citizen using multiple cameras of Korean Patent Application Publication No. 10-0862398, the road in an illegal parking vehicle and an unmanned security enforcement method for crime prevention Acquiring an image of the surveillance region set in the plurality of analog images with a plurality of fixed fixed cameras; Converting the obtained plurality of analog detection images into one analog image in an image mixer; Converting the analog image of the image mixer into a digital image capable of image analysis by an image transmitting apparatus and transmitting the analog image to a field control module via a hub; Detecting an object of interest such as a vehicle or a person by analyzing an image of the object transmitted from the field control module and generating a 2D position coordinate of the detected object region of interest; The 2D position coordinates are converted to 3D position coordinates from the position mapping module and then controlled by an on-site control module to control a control or control PTZ camera module to move a PTZ camera to an index coordinate to track an object of interest. It is described as an illegal automatic parking control method for illegal parking having a security CCTV function using multiple cameras.

However, the conventional systems as described above have the disadvantage that only the vehicle number of the speeding vehicle can be checked, and that the image quality is deteriorated or photographed by several cameras in order to identify a person other than the vehicle.

Therefore, through the transmission method of multiple video streams for improving the background and object identification of the IP camera of the present invention, it is possible to use the IP address used in the existing server without changing the method of the existing server. In addition, by classifying video streams by PORT, one camera can be used to identify objects such as fast-moving vehicles, slow-moving people, and bicycles without any deterioration in quality than conventional cameras. In a very different area, both the bright and dark areas can be identified and controlled through different parameters to capture multiple video streams, and the server can manage the recorded video streams by managing IP addresses and PORTs. Provides a method of transmitting multiple video streams for improving background and object identification I would want.

According to the present invention, after setting the parameters 40 of the camera 10 differently for each of the real-time frames 100 in one camera 10, the frames 100 captured by the parameters 40 of the same condition are photographed. It is characterized in that the classification and conversion to multiple video streams.

The multiple video stream transmission method for improving the background and object identification of the IP camera according to the present invention can be used interchangeably without changing the method of the existing server using the IP address used in the existing server, By classifying video streams by PORT, one camera can be used to identify objects such as fast-moving vehicles, slow-moving people, and bicycles without any deterioration of quality than conventional cameras. In different areas, both bright and dark areas can be identified and adjusted through different parameters to capture multiple video streams, and the server can manage the recorded video streams separately by IP address and PORT. .

1 is a conceptual diagram of a method of transmitting multiple video streams for improving background and object identification of an IP camera of the present invention;

2 is a detailed diagram of a method for transmitting multiple video streams for improving background and object identification capability of an IP camera according to the present invention.

3 is a conceptual diagram of parameter application of a method for transmitting multiple video streams for improving background and object identification of an IP camera according to the present invention.

10: camera

20: server

30: monitor 31: window

40: parameter

50: IP address

60: PORT

100: frame

200: video

In addition, the plurality of video streams transmitted from the camera 10 to different port numbers are restored to the original video through the decoder in the server 20, and stored in the DB server or output to the monitor 30.

In addition, the monitor 30 is characterized in that the window 31 is divided and displayed by the number of the set camera 10 parameter 40.

In addition, the plurality of windows 31 divided into the monitor 30 is characterized in that the size can be adjusted.

In addition, the camera 10 is assigned a unique IP address 50, the camera 10 assigned the IP address 50 is assigned to each PORT (60) to the video stream classified according to the conditions of the parameter 40 ) Is assigned.

In addition, the present invention includes a camera (10) for photographing a plurality of video streams having different parameters (40); A server 20 for managing an image stream captured by the camera 10; A monitor (30) for outputting an image stream received from the server (20); Characterized in that it comprises a.

Hereinafter, described in detail by the accompanying drawings of the present invention.

1 is a conceptual diagram of a method for transmitting multiple video streams for improving background and object identification of an IP camera of the present invention, and FIG. 2 is a detailed view of a method for transmitting multiple video streams for improving background and object identification of an IP camera of the present invention. 3 is a conceptual diagram of parameter application of a method of transmitting multiple video streams for improving background and object identification capability of an IP camera of the present invention.

The present invention will be described in detail, and the present invention sets the parameters 40 of the camera 10 differently for each of the real-time frames 100 in one camera 10, and then photographs the parameters 40 of the same condition. The frame 100 photographed by FIG. 1 is classified into multiple image streams and transmitted.

In addition, the plurality of video streams transmitted from the camera 10 to different port numbers are restored to the original video through the decoder in the server 20 and stored in the DB server or output to the monitor 30.

The camera 10 includes a lens and lighting.

The camera 10 photographs the frame 100 by applying continuous parameters, and generates the frames 100 in order by the parameters 40 applied for each category, and the generated frame 100 sets the set parameters. (40) Divided in order by category.

The camera has an encoder function, and an encoder method uses an H.264 or MPEG codec.

The video stream is divided into different video streams according to the IP address 50 and the PORT 60 number.

The IP address 50 is different for each camera.

The PORT 60 includes several in one IP address 50, and one PORT 60 includes one video stream.

The PORT 60 is assigned a different number from one IP address 50.

The number of the PORT 60 is the same as the number of video streams photographed and classified by one camera.

The monitor 30 is characterized in that the window 31 is divided and displayed by the number of parameters of the camera 10 set 40.

The parameter 40 is controlled by a micro control unit (MCU) controlling the inside of the camera 10.

The parameter 40 is to speed up signal processing such as calculation by a digital signal processor (DSP) inside the camera 10.

The parameter 40 is divided by category, and configures the photographing order, shutter speed, sensitivity, etc. for each category.

The photographing order is a photographing order according to the classification.

The shutter speed is the speed at which the shutter opens and closes to give the frame a suitable light.

The sensitivity is such that the frame feels light.

The parameter 40 is automatically set in order by the camera 10, and the degree of light input is adjusted by the parameter 40 set for each frame 100 generated when the camera is photographed.

In one embodiment of the parameter, the parameter includes a gain, the gain is an ISO method to amplify the light exposure of the frame with less light exposure to recognize the shape of the object, when shooting in the dark place ISO method The image can be amplified so that it can be recognized so that the object can be recognized even in the dark.

The frame 100 generally refers to a single still image in the image 200 which is sequentially arranged.

By checking the frame 100, if the average brightness is less than the reference value, the brightness value is raised as a whole, and if the average brightness exceeds the reference value, the brightness value is lowered as a whole.

As an embodiment of the frame 100, the frame 100 may obtain a frame for photographing a slow object and a frame for photographing a fast object, and the frame for photographing the slow object is a frame at which a shutter speed is slowed. (100) The shutter speed is slow and the overall light exposure is high, so the brightness is bright but fast moving objects are hard to see, and the frame of the fast object is shot with the fast shutter speed so that the number of fast moving vehicles can be seen well but the exposure of light This lowers the brightness.

As another embodiment of the frame 100, the frame 100 is a frame for photographing the fast object is to check the license plate of the car, the frame is photographed for the slow object is to identify the appearance of a person or a bicycle.

As an embodiment of the photographing order, the photographing order is to repeatedly photograph the set categories in turn.

As another embodiment of the photographing order, the photographing order photographs the set classification in the designated pattern order, and the order of the frame 100 may be that the frames 100 having the same parameter 40 are photographed continuously.

As an embodiment of the classification, the classification is to divide the frames having the same parameter 40 differently, and the fps quota may be set differently according to the classification.

As another embodiment of the above classification, by assigning fps according to the classification, it is possible to allocate a lot of fps to a frame that needs to be shot quickly and to assign a few fps to a frame that needs to shoot a slow shooting speed, and vice versa. You can assign a lot of fps to frames that need to be shot at slower speeds.

The fps is a number of frames 100 per second, and the lower the fps, the easier it is to shoot, but the image 200 appears to be broken. When the number of frames 100 increases, it is appropriate to use 30 to 90 fps due to insufficient storage space.

When the shutter speed is fast, the exposure of light decreases and the slow object is blurred, but it is suitable for shooting a fast moving object. When the slow shooting is performed, the exposure of light increases. will be.

The frame 100 transmits the video stream classified by the parameter 40 of the same condition to the server through the network.

The server 20 receives the video streams transmitted from a plurality of cameras separately through the IP address 50 and the number of the PORT 60, and each video stream received by the server 20 is configured DB. The image 200 converted by the decoder to the monitor 30 is output to the monitor 30.

The server divides the video stream through the numbers of the IP address 50 and the PORT 60, and stores them in the DB or outputs them to the monitor 30. If the IP addresses 50 and the PORT 60 have the same numbers, If the same video stream is different from any one of the IP address 50 or the PORT 60, it is divided into different video streams and stored in the DB or output to the monitor 30.

As an embodiment of the DB, the DB is an internal DB inside the server or an external DB connected via a network.

The decoder restores the original image 200 converted to the H.264 or MPEG codec by the encoder.

The image 200 is a continuous frame 100, and is transmitted through a stream method so that the monitor 30 can see it.

The stream method is a method of transmitting continuous data, and the stream method may use TCP / IP, RTSP, and the like. The stream method may be transmitted to the server 20 so that the monitor 30 may check it by any one of the stream methods. will be.

The data transmitted in the stream method through the server 20 may be output through the monitor 30.

The monitor 30 is capable of outputting the frame 100 which is divided into a window 31 and delivered in a stream manner, and each frame 100 delivered in the stream manner is automatically and manually imaged before being output. Adjustment is possible.

The window 31 outputs a continuous frame 100 divided into categories according to each position and size.

In one embodiment of the window 31, the window 31 is adjustable in size, it is possible to change the size through the size adjustment button selection, drag and drop (Drag and Drop).

As another embodiment of the window 31, the window 31 can change the position where each image 200 is output, and the image output to the window 31 by a double click, drag and drop, etc. 200) is to change the output position.

In addition, the present invention includes a camera (10) for photographing a plurality of video streams having different parameters (40); A server 20 for managing an image stream captured by the camera 10; A monitor (30) for outputting an image stream received from the server (20); It will include.

As an embodiment of the present invention, the present invention is applied by applying different parameters 40 in the sorting order to one camera 10 having different parameters 40 is set, and classifies the photographed frame 100 By transmitting to the server 20 in a stream manner, the frame 100 transmitted to the server 20 is stored in the DB or output to the monitor 30.

In another exemplary embodiment of the present invention, the monitor 30 monitors the brightness information of the vehicle frame and the human frame, and monitors the frame 100 adjusted by adjusting the brightness by applying the parameter 40 to the frame 100. By dividing the 30 into the window 31, one frame 100 classified into one window 31 is output.

A typical IP camera captures only one video stream and transmits it to a server. When classifying video streams in the server, each camera is classified based on a different IP address and one window is assigned to one IP monitor. In the present invention, the camera is not only distinguished by different IP addresses, but also by using a plurality of video streams captured by the camera by PORT, so that more video streams can be obtained and used in existing servers. You can use it interchangeably without changing the method.

Therefore, the transmission method of multiple video streams for improving the background and object identification of the IP camera according to the present invention can be used interchangeably without changing the method of the existing server using the IP address used in the existing server, In addition, by classifying video streams by PORT, one camera can be used to identify objects such as fast-moving vehicles, slow-moving people, and bicycles without any deterioration in quality than conventional cameras. In the very different areas, both bright and dark areas are identified and controlled through different parameters to capture multiple video streams, and the server has a remarkable effect of managing the captured video streams separately by IP address and PORT. have.

Claims

After shooting by setting the parameters 40 of the camera 10 differently for each real time frame 100 in one camera 10, the frames 100 captured by the parameters 40 of the same condition are classified and several In the multiple video stream transmission method for improving the background and object identification of the IP camera converting and transmitting to two video streams,

The multiple video streams received from the camera 10 with different port numbers are restored to the original video through the decoder in the server 20 and stored in the DB server or output to the monitor 30,

The window 30 is divided and displayed on the monitor 30 by the number of parameters of the camera 10 set 40.

The window 31 divided into a plurality of monitors 30 are each adjustable in size,

The camera 10 is assigned a unique IP address 50, and the camera 10 assigned the IP address 50 assigns each PORT 60 to a video stream classified according to the conditions of the parameter 40. Assign them,

The video stream is divided into IP addresses 50 and PORT 60 and stored in a DB or output to the monitor 30. If the IP addresses 50 and PORT 60 are all the same number, the same video is used. If the IP address 50 or the PORT (60) number is different, it is divided into different video streams.

If the average brightness is less than the reference value by checking the frame 100, the MCU of the camera 10 raises the brightness value of the frame 100 as a whole. If the average brightness exceeds the reference value, the MCU increases the brightness value of the frame 100. Transmission method of multiple video streams for improving background and object identification of IP cameras