JP2000092475A - Rotary tv camera monitoring device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To secure the reliability by performing the bidirecrional transmission of control signal etc., of a rotating TV camera monitoring device in non- contacting system. SOLUTION: The camera control signal to a rotatable camera 46 is transmitted to a camera control circuit 44 via a rotary transformer fixation part 51 and a rotating part 52 without contacting and the camera status signal after control is detected by a sensor 47, time-multiplexed with a camera video signal 45, and put back to the rotary transformer fixation part 51 through an FM modulator 50 to return to a block C through a coaxial cable. Camera status data are separated at 76 to optimize the camera control by the automatic decision of a CPU 77 or in collaboration with an operation part 78. For a video signal 71, an image measuring circuit 80 optimizes camera image-pickup operation corresponding to a temporally changing input image automatically or semiautomatically as mentioned above. The algorithm of the image-pickup operation always has camera angle, position, field angle (wide, telephoto), etc., which are optimized by closed loop system constitution according to image measurement data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

FIELD OF THE INVENTION The present invention relates to a rotary TV monitoring system used for monitoring purposes. A TV surveillance camera device linked to a rotatable rotating body that is mechanically non-contacting, a technology for realizing bidirectional transmission of electric signals of the camera inside and outside the rotating body, and an image using the video signal of the camera. The technology involved in measuring camera characteristics and optimizing camera control using the information.

[0002]

2. Description of the Related Art In a TV monitoring apparatus, there is a rotation monitoring camera system in which a dome-shaped TV camera covered with, for example, a translucent plastic cover is installed on a ceiling or the like and the TV camera is rotated. However, in a relatively large space, there are many system configurations for monitoring a relatively wide area such as a vicinity of an entrance of a specific place, an ATM corner of a bank or the like, and a system design is a single function. It was not well suited for the purpose.

In a conventional surveillance system, a surveillance area is usually covered by setting a wide angle of view. However, it is difficult to identify a criminal's face due to insufficient resolution of a TV image. It was incomplete and could not serve its intended purpose. In order to supplement the above drawbacks, it is necessary to rotate the camera, use the pan and tilter functions, etc., and accurately point the camera angle to the object to be monitored. In this case, in order to be able to move freely, an electric signal, particularly a coaxial cable, needs to move around with a radius of rotation, and it is necessary to secure a wide space. This is a restriction on the installation location, and, for example, has been an obstacle for installation on a ceiling or the like. Moreover, the surveillance camera was a single-function control, and was an open-loop control that was operated by a command from the operation unit.

As a means for avoiding this, there is a method of outputting an electric signal to the outside of the rotating body via a contact type slip ring. However, the life of the slip ring is short, and the maintenance work is difficult in a system installed on a ceiling or the like. Met. Conventional AT
The TV camera monitoring system that monitors M, etc. has a learning function that uses the image content determined by the initial setting as it is, learns the image content that changes temporally and spatially, and uses that information for camera optimal control sequentially. I didn't. Therefore, it did not have a function capable of sufficiently achieving the monitoring purpose.

[0005]

SUMMARY OF THE INVENTION An electric connection required for taking out an output of a rotary TV surveillance camera installed in conjunction with a rotatable rotating body and a camera control signal outside the rotating body is maintenance-free. Realized with a long life and stable method. An image recognition quality that can sufficiently match the TV monitoring image with the original monitoring purpose is realized. From the one-way, open-loop control operated by the command of the operation unit, the status of the camera after the control is always monitored two-way, and a function for updating the control method is provided.

[0006]

A video signal and a control signal of a camera, which are electrically connected to a rotary TV surveillance camera installed in association with a rotatable rotating body, are realized by a non-contact rotary transformer system. . In such a case, bidirectional transmission of a plurality of types of electric signals inside and outside the rotating body is realized by performing time multiplexing on one coaxial cable. Multiple compact and lightweight cameras are mounted on a rotating body, their functions are independent of each other, and parameters that are essential for monitoring when operating them are acquired by image measurement means using camera output images, and are therefore reflected in camera control Have means to make it. Therefore, it has the feature that the system was constructed in a closed loop configuration.

[0007]

FIG. 1 shows an example of an external view of a dome-shaped camera. The CCD camera 10 has a function of freely rotating up and down (tilt) and left and right (pan). For the purpose of avoiding complexity, an example in which one camera is mounted on a rotating body will be described.
The camera 10 is rotated by a mechanism 8 to which a shaft 6 and a case 9 are attached and which can move up and down, (tilt) left and right (pan), and a bearing 7. Means for rotating the shaft 2 includes, for example, a stepping motor 4 and a bearing 5.
It is rotated by 60 degrees. Stepping motor 4
There are several bolts fixed to the ceiling, but they are not shown. The opening 1 is a wire for transferring an electric signal to and from the outside, and a hole for taking out an AC cord for supplying power of the wire and the stepping motor 4.

The translucent plastic cover 11 has a hemispherical shape and is attached to the holder 3. This is processed to a thickness of 2 mm and a uniform thickness, and consideration is taken so that the refractive index becomes uniform, and the geometric distortion in the dome shape, that is, optical aberration is suppressed to a practical level.

FIG. 2 shows a side view of the structure of the rotary transformer described in the present invention. For convenience of explanation, the signal winding 23 is 1
Individual, electrostatic or electromagnetic shield windings 24, 25
2 illustrates the case where the number of the power supply windings 26 is one. However, the present invention is not limited to this configuration. When a plurality of different signal frequency windings are provided, a short ring for preventing magnetic or electrostatic coupling between them is provided as necessary.

In comparison with a conventional rotary transformer, the primary winding is formed on a fixed part, on the stator side, and the secondary winding is formed on a rotating rotor, which keeps a narrow gap and faces spatially. And placed. The rotor / stator refers to a disk-shaped ferrite core having a high magnetic permeability, in which grooves 21 and 22 are provided with predetermined windings and embedded. The thin disk made of these two ferrites is a transformer having a structure spatially opposed to each other with a gap of about 30 mm.

The equivalent circuit of the rotary transformer is shown in FIG.
2 and 33. L1 is the inductance of the primary winding,
L2 is the inductance of the secondary winding, M is the mutual inductance, and k is the coupling coefficient. In this proposal, power supply and video signal, and camera control pulse signal,
After that, pulse signals such as camera status signals (information such as position and angle and various parameters for camera operation) and the like are controlled bidirectionally by spatial electromagnetic coupling of the transformer, that is, from the stator side to the rotor side and in the opposite direction. It is transmitted between. Therefore, a Mn-Zn ferrite core material is selected, and the frequency characteristics of the magnetic permeability, the winding diameter of the wire, the number of turns, and the like are optimized so as to have desired characteristics as appropriate.

The windings of the video and pulse signals transmitted and received by the rotary transformer are arranged in a flat coil shape in grooves formed in the radial direction of the ferrite core. FIG. 4 schematically shows a power supply winding T1 and an optional signal winding T2,
Between the T3 windings, at least one, if necessary, a plurality of windings, a shielding winding, and SR1 and SR2 having a short ring shape are arranged between the windings. If necessary, a plurality of ribbon-shaped conductors (not particularly shown) are arranged for shielding. Thus, a structure capable of preventing magnetic and electrostatic interference and interference between two arbitrary signal windings is adopted.

In addition to the method of embedding the windings and ribbon-shaped conductors in the grooves in the rotary transformer as shown in FIG. 2, it is possible to realize a shielding effect equivalent to or even better than these. It is easy to form a thin film of a metal, an alloy thereof, or a composite material thereof, such as copper or permalloy, at a location corresponding to a groove by vacuum evaporation, and electromagnetic and electrostatic shielding can be further improved. According to this, it is possible to omit the work of forming a groove in the ferrite core for applying the winding. Furthermore,
Since the gap between the rotor and the stator can be reduced, the coupling force of the transformer can be improved.

As means for supplying power to the rotating body from outside the rotating body, it is necessary to supply power to the primary side of the transformer, that is, to the stator side. For this reason, a method of superposing a camera power supply on the camera video signal coaxial cable 67 is adopted. Both ends of the coaxial cable are terminated with a matching resistor of 75Ω for each of Z1 and Z2. Z from the camera power supply circuit 72
1 to a DC / AC converter 62 via a constant current circuit 61.

The constant current circuit 61 keeps the current flowing through the coaxial cable 67 constant even if the load current fluctuates on the rotor 52 side, so that there is no noise disturbance in the video signal. It is supplied from the DC / AC converter to the rotor 52 via the winding T1 of the stator 51. A rectifier 41 and a stabilized power supply circuit 42 supply power to all circuit blocks arranged in the rotating body. The input power supplied from the stator is determined in consideration of the required power of the rotating body. In this case, the waveform may be a sine wave signal, a square wave having appropriate rising characteristics, a triangular wave, or the like. It is selected from the viewpoint of rectification efficiency, interference with frequency characteristics of a signal transmitted and received via a transformer, and the like.

A method of controlling the rotating camera 46 shown in FIG. 4 from outside the rotating body will be described. In the camera control console, block C, a process of multiplexing the camera control signal during the vertical blanking period of the camera video signal is performed by a command from the operation unit 78. In this case, the CPU 77 always refers to the camera status signal separation circuit 76, performs measurement and judgment, and automatically operates.
It is also possible for humans to add and supplement information, and any means will generate a camera control signal to be updated.

FIG. 5 shows the allocation of multiplexed signals.
Is shown schematically in FIG. The camera video signal 83, the camera control signal 81, and the camera status signal 82 after the control are multiplexed, and they are multiplexed by the non-contact control and the bidirectional transmission system via the rotating stator 51 and the rotor 52. It is configured to achieve a functional monitoring purpose.

The types of digital signals for controlling the rotary camera 46 are for expanding camera functions such as pan, tilt, tele, wide, zoom, and focus. A digital signal for controlling these is composed of a plurality of pulse trains and is multiplexed as 81 in a vertical blanking period 84. On the other hand, reference numeral 82 denotes a multiplexed pulse signal for returning to the operation unit 78 the newly updated status of the camera under control, that is, pan, tilt, tele, wide, zoom, and focus information. For example, if the pulse train has a 4-bit sequence configuration, 16 types of control signals can be independently obtained. However, in the present invention, a 16-bit sequence is used in consideration of expandability of various parameters.

Serial data is transmitted during transmission, and parallel processing is performed when the data is processed. Thereafter, the data is decoded (not shown). The output signal is determined by the CPU 77, and further, an image measurement method described later is referred to for monitoring purposes. Perform optimal control of the matched camera. 81, 82, and two types of 16-bit sequence pulse signals for bidirectional transmission are time-superimposed on the camera video signal 83 so that they can be transmitted through the coaxial cable 67 for video signals. Can be built.

The camera video signal from the console block C, which is time-multiplexed with the camera video signal, is supplied to the coaxial cable 6
7 and transmitted to the rotator block B. This signal is supplied to the control signal separating circuit 63 via the capacitor C2 and the surge absorber 65, and then the camera control digital signal 81 is separated from the coaxial cable 67. This signal passes through a rotary transformer, passes through a slice circuit and a serial / parallel conversion circuit 43, enters a camera control circuit 44, is decoded into a signal for each control item, and controls the rotary camera 46.

The camera status signal 82 which is updated successively
Are time-multiplexed in the vertical blanking period 84 of the camera video signal 83. It is inserted in a time division manner among a plurality of horizontal synchronization signals 86 existing in the vertical blanking period.
The voltages representing the updated camera status information are individually obtained as analog outputs of a sensor 47 disposed integrally with the camera, such as a turntable, a camera position, and an angle. These are individually subjected to A / D conversion 48 to be binarized. A parallel-to-serial conversion 49 is performed on these data, and a synthesis circuit 45
Send to Then, time multiplexing is performed by the camera video signal 83 and the synthesizing circuit 45.

The time multiplexed signal is converted by an FM modulator 50 into an F
Receive M modulation. This signal is sent to the FM
After demodulation by the demodulator 64, via the surge absorber 66,
An anti-reflection terminator, Z2, 75Ω is added.
The signal is transmitted to the camera control console and the block C via the coaxial cable 67 via the condenser C2. The video output signal for monitoring is always monitored at the terminal 71.

Most of surveillance cameras that move with rotating bodies on the market are simple and single-function. In this proposal, these conventional systems are expanded. That is, two or more cameras having independent functions, such as tele and wide, are installed on the same rotating body and individually imaged at the same time to constantly obtain the entire image, details, and an enlarged image of the monitoring area. Recently, an extremely lightweight image sensor such as an ultra-small, artificial retinal camera, or a multifunctional CMOS camera can be used, and there is no adverse effect of mounting a plurality of imagers on the same rotating body.

After the area to be monitored by the monitoring camera is set, the monitoring area is imaged by cameras having different functions. For example, in FIG. 6A, an image including the ATM corner of a bank and a user standing in front of the ATM area, for example, an ATM 95, a floor 96, and a user 97 are acquired as an image 91. Important monitoring spots are shown in Fig. 6.
It is enlarged as shown by b. The image 92 may be an enlarged image obtained in a time series of the image 91, or may be obtained independently and simultaneously. This is a wide area with one TV
The resolution of a TV surveillance system that is monitored by a camera is generally insufficient to identify a criminal or the like, and is intended to compensate for this and to realize a substantial improvement in the surveillance function.

Therefore, the enlarged image of FIG. 6b is necessary, but at the same time, from the viewpoint of special circumstances of surveillance and court evidence,
The captured environment of FIG. 6b needs to be integrated with the background image. Therefore, as shown in FIG. 6C, the image 91 of FIGS. 6A and 6B is keyed to the image 92. Image 93 shows a state where the two images are superimposed. Image 93
There are two cases where the image 94 in the image is the image 91 itself or the size thereof is arbitrarily processed. That is, these sequences, the size of the image 91, the insertion position, etc.
Subject to software controlled by the PU.
Two cameras with different functions are rotated integrally with the rotating body, and the outside thereof is covered with a translucent plastic dome-shaped cover, so that the surveillance entity cannot be seen from the outside.

A method of realizing such a function in a time sharing manner with only one camera will be described below. The sequence for switching these camera functions between time, wide at t1, time, and tele at t2 is determined by the operation unit 78 of the control console and the CPU 77 of the microcomputer in consideration of the situation of the monitoring area in advance. It is equipped with a means for optimizing the image that constantly changes. Therefore, the camera image is output from the output terminal 71 to the capture circuit 79 and the image measurement circuit 80.
The CPU 77 measures and judges data, and the camera control is always optimized. In this case, what is needed is a software that implements a sequence that always changes over the subject, that is, a wide image, a background image, and FIG. 6A from the viewpoint of emphasizing the resolution for identifying the culprit. The reason why the background image, FIG. 6A, is required is that the image for monitoring purpose is required to be falsified. Therefore, unless the background (environment) from which the image is captured can be recognized, it cannot be used for surveillance purposes or as forensic evidence.

A system similar to the present invention can be constructed by using a single camera system without using a plurality of cameras and using a system in which the image is switched between time-division or discrete-tele and wide, for example. However, deterministic scenes cannot be synchronized in principle. The wide image, the image 91 in FIG. In this respect, in this case, FIG.
As shown in the figure, an image 91 is superimposed on both screens in tele and wide, that is, an image 92, on the screen of the TV monitor.

When a single camera is used, the superimposed (Picture-in-Picture) screen 91 is temporarily stored in a digital memory at a discrete time t2 at which a tele image is obtained, and then the tele image is stored. This image is refreshed at the acquisition interval, (t2 + tn).
In a system using two cameras independently, an image to be reduced and displayed is temporarily captured in a digital memory in real time, reduced, and superimposed at a predetermined position on the screen at an arbitrary size. That is, synchronization can be ensured only in the case of two or more camera systems.

In one surveillance camera system, the algorithm and sequence for selecting the tele / wide function of the same camera are important. From the viewpoint of maximizing the monitoring purpose, a luminance change area of the monitoring TV image is extracted. In a method in which an input image is subjected to digital image processing in 256 steps, a temporal change in luminance can be roughly classified into two types. (A) Approximately 15 levels /
Second, (b) When a person moves in front of a background such as a building, the speed is about 90 to 200 levels / second. The microcomputer determines data according to the luminance change speed of the extracted image, and uses the data as an optimal control signal for the tele / wide and switching sequence. Also, this is reflected in the position control information of the camera rotating body. That is, the camera moves to the optimum camera position and the camera angle is also optimized.

To determine the center of gravity of an image in a specific area, first, the first moment in the required ROI (Reasion Of Interest) shown in FIG.
Ask like. Here, PRF (x) and PRF (y) represent the gray value of each pixel in the image and ROI on the X axis and Y
This is the value projected on the axis and totaled.

[0031]

(Equation 1)

[0032]

(Equation 2)

Next, the barycentric coordinate on the x-axis, Pg (x), is
Is required.

[0034]

(Equation 3)

The coordinates of the center of gravity on the y-axis, Pg (y), can be obtained by equation (4).

[0036]

(Equation 4)

The pq-order moment mpq of the figure given the set R is given by equation (5). Here, 0th order, m00 is the area.

[0038]

(Equation 5)

Further, according to Equation 6, the moment of inertia can be obtained.

[0040]

(Equation 6)

The moving distance of the subject can be obtained from the center-of-gravity measurement data of this image, and these can be easily realized by microcomputer assist. Using this data,
The tele / wide automatic switching is realized, and further, the tele / wide switching sequence can be optimized.

The tele-screen is not always limited to a reduced screen.
The combination of the wide image and the time table are appropriately selected from the above-described image itself. Generally, the monitoring image has a system configuration in which the monitoring image is simultaneously recorded on a time-lapse VTR or the like. Therefore, it can be analogized that the variation of the imaging sequence is easily extended from the contents described here in order to increase the redundancy of the video information from the viewpoint of the monitoring purpose.

[0043]

According to the present invention, the output of a rotary TV surveillance camera installed in conjunction with a rotatable rotating body and the camera control signal are
The electrical connection required for taking out the rotating body has been realized without maintenance. The present invention has achieved a significant improvement over conventional imperfect monitoring systems, both in terms of monitoring functions and in terms of device reliability.

The bidirectional transmission and control of a plurality of types of electric signals inside and outside the rotating body of the rotary TV surveillance camera are realized by time-multiplexing one coaxial cable. As a result, since the system was designed in a closed loop configuration, the system was constantly updated up-to-date, and optimal control conditions could be maintained.

[Brief description of the drawings]

FIG. 1 is an external view of a rotary TV surveillance camera of the present invention.

FIG. 2 is a side view illustrating a configuration example of a rotary transformer.

FIG. 3 is an equivalent circuit of a rotary transformer.

FIG. 4 is a block diagram of an electric circuit of the rotary TV surveillance camera.

FIG. 5 is a waveform illustrating a pulse signal that is time-multiplexed in a vertical blanking period of a camera video signal.

6a, 6b, 6c are images taken by the surveillance camera.

FIG. 7 is a diagram illustrating measurement of a first moment and a center of gravity of an image.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 AC code outlet 2 Shaft 3 Holder 4 Stepping motor 5 Bearing 6 Shaft 7 Bearing 8 Moving mechanism 9 Case 10 CCD camera 11 Translucent plastic cover 21 Rotor of rotating transformer (rotating part) 22 Stator of rotating transformer (fixing part) 23 Video signal winding 24, 25 Short ring 26 Power supply winding 27 Gap between rotor and stator 31 Equivalent primary inductance of rotary transformer 32 Equivalent secondary inductance of rotary transformer 33 Mutual of rotary transformer Inductance 41, 42 DC power supply circuit 43 Slice / serial / parallel converter 44 Camera control circuit 45 Compositing circuit 46 Rotating camera 47 Sensor 48 A / D conversion circuit 49 Parallel / serial conversion circuit 50 FM modulator 51 Rotating transformer stator Block 52 rotary transformer Rotor block 61 Constant current circuit 62 DC / AC converter 63 Control signal separation circuit 64 FM demodulator 65, 66 Surge absorber 71 Video output terminal (for coaxial cable) 72 Camera power supply circuit for superimposing coaxial cable 73, 74 Surge absorber 75 Control Signal multiplexing circuit 76 Camera status data separation circuit 77 CPU 78 Operating section of rotary camera main body 79 Video capture circuit 80 Image measurement circuit 81 Camera control pulse signal composed of 16 bits 82 Pulse signal representing camera status composed of 16 bits 83 Camera video signal 84 Vertical blanking period 85 Vertical synchronization signal period 86 Horizontal synchronization signal period 91, 92, 93, 94 Surveillance camera image taken 95 ATM (Automatic Teller's Machine) 96 ATM corner floor 97 ATM corner user

Claims (4)

[Claims] 1. A surveillance TV camera mounted on a 360-degree rotatable rotating body, wherein a function of arbitrarily rotating and tilting a monitoring area of the TV camera can be controlled from outside the rotating body. A case where a video of the surveillance TV camera, a camera control signal and a camera status signal updated after the control are output to the outside of the rotating body, and a case where a bidirectional control means for inputting the rotating body portion from outside the rotating body is realized. In addition, a rotary TV camera monitoring device having an input / output transmission unit for an electrical signal in a mechanically non-contact manner with a rotating body. 2. A rotor comprising a grooved ferrite core for supplying power to an electronic circuit mounted on the rotating part from outside the rotating body to a rotating part of the monitoring TV camera device attached to the rotating body. There are stators, each of which is equipped with a plurality of rotating transformers, each of which has an electric winding in a ferrite groove in multiple blocks, enabling electrical power to be supplied to the electric circuit of the rotating body from outside the rotating body without mechanical contact. Rotary TV camera monitoring device with the following features: 3. A moment of a specific area of a subject image is measured in a timely and discrete manner, a center of gravity is calculated, and a moving body distance is obtained from a temporal change thereof.
Rotary T with means for obtaining automatic control signals that can optimize functions for wide, tele, and pan and tilt monitoring purposes
V camera monitoring device. 4. The TV monitor screen is displayed for at least two
There are images that are taken independently by two cameras with different focal lengths, or images that are taken by the same camera at different focal lengths in chronological order. The second image is characterized in that an enlarged image of a part is set in advance in the monitoring area, and both images are sequentially updated based on the image measurement data, and both images are integrally displayed on the same TV monitor. Rotary TV camera monitoring device with.