RU2371880C1 - Panoramic video surveillance method and device for implementing thereof - Google Patents

Abstract

FIELD: information technologies.

SUBSTANCE: panoramic video surveillance is performed with a television system by means of television camera with the view in the area close to semi-sphere, i.e. in spatial angle of 360 degrees in azimuth and tens of degrees as to the place angle. Image acquisition in the television camera is performed with field angle in the object space of 360 degrees in azimuth, and circular panoramic television image is converted to a common image; at that, one current "circular" frame is read off the server memory by means of n "square" frames, the number of which meets the ratio:

where γ_h - horizontal angle of vision field in degrees of the image surveyed by the operator, and optical assembly of television camera is a panoramic catadioptric objective, and to the server there introduced is the unit converting "circular" frame to "square" frames (UCCF), the input of which is connected to the output of RAM unit for a frame, and the output - to "network" output; at that, this conversion is performed in a programmable way.

EFFECT: eliminating losses of resolution capability of panoramic image by eliminating movement of light flux during exposure.

3 cl, 4 dwg

Description

The present invention relates to panoramic television surveillance, which is performed by a television system using a television circular camera in an area close to the hemisphere, i.e. in a spatial angle of 360 degrees in azimuth and tens of degrees in elevation.

The closest in technical essence to the claimed invention should be considered a method of panoramic television surveillance [1], which consists in installing the television camera in a fixed position, moving the light flux directively with repeating cycles towards the photodetector, generating an analog image signal in the television camera observed space, convert the analog video signal into a digital video signal, transmit the video signal through the interface to the server, record the video signal l server RAM carried selective control panoramic image stored in the server memory to the video monitor two or more connected to the server PC.

The prototype device [1] contains a series-connected television camera and a server, which is a node of a local area network, the television camera consists of sequentially located and optically coupled optical nodes and a digital television signal sensor, and the lens of the stationary outside optical node is covered with a transparent dome inside which there is a mirror rotating by means of an engine, and two or more personal computers are connected to the server, each of which contains system unit, the manipulator "mouse" and a video monitor.

The spatial arrangement of the mirror, the focal length of the lens, and its relative aperture in the aggregate determine the collecting ability of the optical unit and its angular field of view when the photodetector is exposed to the digital television signal sensor during image capturing in a television camera. Each individual capture image corresponds to a segment of space with a limited angle of field of view of the television camera at the current mirror position. The combination of these fragmentary images forms the desired panorama. Due to the fact that the process of exposure (accumulation) of charges in the photodetector of the television camera of the prototype is carried out not for a static, but for a moving image, blurring of the generated image in the horizontal direction is inevitable.

The disadvantage of the prototype is the loss of resolution generated by the television camera panoramic image presented to the operator on the screen of a video monitor.

The objective of the invention is the elimination of loss of resolution of the panoramic image by eliminating the movement of the light flux during the exposure.

The problem is solved in that in the claimed method of panoramic television surveillance, which consists in installing the television camera in a fixed position, forming an analog image signal of the observed space in the television camera, converting the analog video signal to a digital video signal, transmitting the video signal through the interface to the server, recording video signal to the server’s RAM; they selectively control the panoramic image stored in the server’s memory on the screens monitors of two or more personal computers connected to the server, according to the proposed invention, image capture in a television camera is performed with an angular field in the space of objects 360 degrees in azimuth, and a panoramic panoramic television image is converted into a conventional image, and one current “circular” frame is read from memory server using n "rectangular" frames, the number of which satisfies the ratio:

,

,

where γ _g is the horizontal angle of the field of view in degrees observed by the operator of the image.

The problem in the inventive panoramic television monitoring device, designed to implement the inventive method, is solved by the fact that in this device containing a series-connected television camera and a server, which is a node of the local area network, the television camera consists of sequentially located and optically connected optical node and a digital television signal sensor, and two or more personal computers are connected to the server, each of which contains the system unit, the mouse and video monitor, according to the invention, the optical node of the television camera is a panoramic mirror-lens lens, and a unit for converting a “ring” frame into “rectangular” frames (BPKP) is inserted into the server, the input of which is connected to the output of the operational unit memory per frame, and the output - to the output of the "network", and this conversion itself is performed not by new hardware, but by software.

Comparative analysis with the prototype shows that the inventive method is characterized by the presence of new features, namely the presence of the following actions:

- an angular field of image capture of 360 degrees in azimuth in a television camera;

- converting the annular panoramic image into a regular image with the condition for this action to be taken when the number n of "rectangular" frames corresponds to one current "annular" frame.

The combination of known and new features is not known from the prior art, therefore, the claimed method meets the requirement of novelty.

Comparative analysis with the prototype shows that the inventive device is characterized by the presence of new features, including:

- a new constructive element, which is BPCP - a block for converting a "ring" frame into "rectangular" frames;

- electrical connections BPKP as part of the server;

- the form of the elements, namely the form of the optical node of the television camera, which is a panoramic mirror-lens lens, and the form of the BODC, which is implemented in software.

The combination of known and new features of the claimed device is not known from the prior art, therefore, the proposed technical solution meets the criterion of novelty.

In the claimed solution, panoramic observation is carried out without directive movement of the light flux towards the photodetector of the television camera, there are no more mechanical moving elements in the television camera, therefore, blurring of the television image generated by the camera in the horizontal direction is eliminated, as a result of which the loss of resolution inevitable in the prototype in each fragment is eliminated panoramic image displayed on computer monitors.

Therefore, according to the technical result and the method of its achievement, the proposed solution meets the criterion of the presence of an inventive step.

Figure 1 shows the structural diagram of the inventive device that implements the inventive method; figure 2 shows a ring image and the possible placement of areas of television recording; figure 3 schematically shows a panoramic image proposed to the operator of a personal computer, in the form of a sequence of six "rectangular" frames; figure 4 shows an example of the possible placement of the photodetector sections on a matrix chip made by CMOS technology for panoramic observation.

The inventive panoramic television monitoring device, see figure 1, contains a series-connected television camera 1 and a server 2, which is a node of a local area network, while the television camera 1 consists of sequentially located and optically connected optical node 1-1, which is a panoramic mirror a lens, and a 1-2 digital television signal sensor, and server 2 includes a BPKP — a unit for converting a “ring” frame into “rectangular” frames (not shown in FIG. 1), the input of which It is connected to the output of the RAM block per frame, and the output is to the “network” output, to which two or more personal computers are connected, each of which contains a system unit 3, a “mouse” 4 manipulator and a video monitor 5.

Panoramic mirror-lens as an optical node 1-1 of a television camera is designed to form an optical image of a circular view (ring image) and design it on the target photo sensor 1-2. The technical solution of the panoramic mirror-lens lens has been patented in Russia by Russian specialists [2]. The angular field in the space of objects for this lens is 360 ° in azimuth and can reach (75-80) ° in elevation, depending on the design. It should be noted that the size of the scattering circle of the panoramic lens should correspond to the size of the decomposition element (pixel) of the photodetector.

Suppose that the horizontal angle of the field of view (γ _g ) of the image observed from the screen of the video monitor 5 is sixty degrees. Then the annular panoramic image projected onto the target of the sensor 1-2 (see figure 2) is conditionally divided into 6 areas in accordance with the required number n equal to 6 for the number of "rectangular" frames when reading one "ring" frame. Each “rectangular” area of television recording (capture) will occupy in the corresponding ring sector not the entire area, but a certain part of it. It is possible to select the area of these areas when they have mutual overlaps.

Let the “rectangular” capture regions with a 4: 3 format be located in the ring sectors according to the criterion for the equidistance of their geometric centers from the center of the optical ring, as shown in FIG. 2. Suppose that as a 1-2 sensor, a VEA-535 open-circuit television camera (VEI-535), offered by the Russian company EVS, is used [3]. This camera can transmit a digital video signal via USB 2.0 at a frame rate of photoelectric conversion of 4 Hz with the number of pixels of the CMOS sensor matrix 2592 (H) × 1944 (V) and the target format 4: 3.

When inscribing in a vertical matrix size ring image (see figure 2) this number of elements is enough to read six images of a "rectangular" frame with a standard resolution of each 500 (H) × 582 (V) pixels, as well as to read the rest of the array pixels that are useless in this application.

In the future, for new matrix photodetectors made according to CMOS technology, it is quite possible to place six photodetector sections on a crystal with a 1: 1 format, as shown in Fig. 4, which ensures a more rational use of the crystal area. Thanks to this, the frame rate of photoelectric conversion can be increased, which will provide the opportunity to observe dynamic panoramic scenes.

Considering that the CMOS technology, in contrast to the CCD technology, allows in principle quite simple means to organize a photodetector containing tens of millions of pixels on a crystal [4, p.75], we can also hope for the implementation of 16 photodetector sections on its area, which will provide the operator with a viewing angle horizontal 22.5 °.

An important circumstance should be noted regarding the use of a photodetector on a CMOS sensor in a television camera. The horizontal field of view of the photodetector horizontally at 360 ° leads much more often to an assessment of the plot situation as observation of objects in conditions of complex illumination and / or brightness in one television frame. Under these conditions, a television camera should have an exceptionally wide dynamic range, which is achieved by organizing a multi-channel light-to-signal conversion in a photodetector [5, 6], which is not available for CCD arrays. When using CMOS sensors, the dynamic range of a television camera can cover 120 dB, because controlling the steepness of this conversion to maintain the optimal exposure time (accumulation) is fundamentally possible in each pixel of the photodetector. The basis for this is the diagram of the topological organization of the device - “coordinate addressing”, providing access to each element.

BPKP is designed to convert the "ring" frame of the image formed by the television camera 1, in the usual "rectangular" frames. In our example, 6 areas of the "circular" frame (see figure 2) correspond to 6 "rectangular" frames at the output of the BODC (see figure 3). This conversion is performed in the server 2, in which the digital video signal is directly recorded in the RAM block. The organization of reading each of the six "rectangular" frames is also assigned to the server, and its implementation is carried out programmatically in real time. The read operation includes for each "rectangular" frame the following procedures:

- individual turn of the electronic image relative to the geometric center by an angle in the range from minus 90 to plus 90 degrees;

- correction of geometric distortions of the corresponding section of the optical panoramic image.

The method of panoramic television surveillance is as follows. We use the structural diagram of the claimed device for panoramic television surveillance (see figure 1) that implements the claimed method.

A television camera 1, comprising a panoramic mirror-lens 1-1 and a sensor 1-2, is mounted in a fixed position, for example, using a photographic tripod (not shown). For the image of the circular view of the space observed by the television camera, the sensor 1-2 via the USB 2.0 interface transmits a digital video signal to server 2 and writes it to the RAM block.

The digital video recording signal for each “ring” image frame with the help of the BPC is converted into n “rectangular” frames and is provided to users of personal computers at the output of the “network” server 2.

Assuming that the current angle of view (γ _g ) of the presented panoramic image is 60 ° horizontally, the “annular” recording frame includes 6 conditional areas (see FIG. 2), which are read in the BODC 6 by “rectangular” frames (see FIG. .3).

The local computer network connected to server 2 is constituted by personal computers, each of which is equipped with a system unit at position 3, a mouse manipulator at position 4, and a video monitor at position 5. Each operator of a personal computer can select using the mouse manipulator 4 the proposed image and its output to the screen of the video monitor 5.

The technical result of the invention is to eliminate horizontal blur in each fragment of the observed panoramic image due to lack of movement during exposure of the photodetector in a television camera, which ensures the preservation of the original resolution.

Currently, all the elements of the structural diagram of a panoramic television monitoring device that implement all the actions according to the claimed method and determine the claimed device, mastered or can be mastered by domestic industry. Therefore, the present invention should be considered as meeting the requirement for industrial applicability.

INFORMATION SOURCES

1. Whiting Joshua B., Barker Alex C. Panoramic digital camera system. US patent No. 6034716, H04N 7/00, 348/36, 348/37, claimed 18L2.1997, published 03/07/2000.

2. Kurtov A.V., Solomatin V.A. Panoramic mirror lens. Russian patent No. 2185645, G02B 13/06, G02B 17/08, filed December 22, 1999, published October 20, 2001.

3. www.evs.ru. Unpacked megapixel television USB 2.0 cameras.

4. Berezin V.V., Umbitaliev A.A., Fakhmi Sh.S., Tsytsulin A.K., Shipilov N.N. Solid State Revolution in Television: Television systems based on charge-coupled devices, systems on a chip, and video systems on a chip / Ed. A.A. Umbitalieva and A.K. Tsytsulina - M.: “Radio and Communications”, 2006

5. Smelkov V.M. Analytical evaluation of the multichannel method of expanding the dynamic range of a television system // Special equipment. - 2007. - No. 2. - p. 25-29.

6. Smelkov V.M. I go to the raster: An essay on inventions in the class H04N / Veliky Novgorod, Ed. NovSU named after Yaroslav the Wise, 2007 .-- 176 p.

Claims (3)

1. The method of panoramic television surveillance, which consists in installing the television camera in a fixed position, forming an analog image signal of the observed space in the television camera, converting the analog video signal to a digital video signal, transmitting the video signal via the interface to the server, recording the video signal into the server’s RAM, carry out selective control of the panoramic image stored in the server’s memory on the screens of two or more video monitors connected to the server sonal computers, characterized in that the image capture in the television camera is performed with an angular field in the space of objects 360 ° in azimuth, and the panoramic panoramic television image is converted into a normal image, and one current "circular" frame is read from the server’s memory using n "rectangular »Frames, the number of which satisfies the ratio:

where γ _g is the horizontal angle of the field of view in degrees observed by the operator of the image. 2. A panoramic television surveillance device comprising a television camera and a server in series with a local area network, and the television camera consists of a serial and optically connected optical node and a digital television signal sensor, and two or more personal computers are connected to the server, each of which contains a system unit, a mouse and a video monitor, characterized in that the optical node of the television camera is pa a mirror-lens lens, and the server has a block for converting a “ring” frame into “rectangular” frames (BPKP), the input of which is connected to the output of the main memory block per frame, and the output to the “network” output, and this conversion is performed programmatically. 3. The panoramic television observation device according to claim 2, characterized in that the photodetector of the digital television signal sensor is made on a matrix with a complementary metal-oxide-semiconductor structure (CMOS matrix).

Images