SECURITY CAMERA WITH SELECTIVE ZOOM: A NEW SOLUTION.

SECURITY CAMERA WITH SELECTIVE SCALING: A NEW SOLUTION.

Vyacheslav Mikhailovich SMELKOV, PhD in Engineering

SECURITY CAMERA WITH SELECTIVE ZOOMING: A NEW SOLUTION

The device of the proposed camera is based on the analog method of optical-electronic image scaling, published in [3].

The structural diagram of the new television camera in accordance with the patent description [2] is shown in Fig. 1. The camera contains a first lens (1), a first television signal sensor (2), a frame signal generator (3), a beam splitter (4), a second television signal sensor (5), a targeting unit (6), a sync pulse selector (7), a switch-mixer (8), a fixed delay unit (9), an adjustable delay unit (10), and a synchronization signal generator (11).

Fig. 1. Structural diagram of the television camera

1 – first lens; 2 – first TV signal sensor; 3 – frame signal generator; 4 – beam splitter; 5 – second TV signal sensor; 6 – aiming unit; 7 – sync pulse selector; 8 – switch-mixer; 9 – fixed delay unit; 10 – adjustable delay unit; 11 – synchronization signal generator.

For covert surveillance, it is preferable to use as the first lens (1) a wide-angle photographic lens of the “fish eye” type with a horizontal viewing angle of about 110 degrees and a frame format of (24 x 36) mm. The design of the beam splitter (4), the targeting unit (6) and the sync pulse selector (7) is completely borrowed from [3].

The beam splitter (4) uses a sequentially located and optically connected semitransparent mirror (4-1), a collective lens (4-2), a reflective mirror (4-3) and a second objective (4-4). The dimensions of the optical frame at the second output of the beam splitter relative to the dimensions of the frame at its first output determine the optical scaling factor Km and the television magnification of the camera. Km is measured by the ratio of the frame dimensions of the objective (1) to the corresponding dimensions of the sensor photo target (5). Obviously, when choosing a CCD format equal to 1/3 inch, the typical Km is about 6x .

The guidance unit (6) is designed to perform spatial positioning horizontally and vertically of the second sensor (5) and to retrieve current information about its position in the rectangular coordinate system C 0U using position sensors. The guidance unit (6) contains a horizontal movement drive (6-1), a vertical movement drive (6-2) and a first position sensor (6-3) and a second position sensor (6-4) kinematically connected to them, respectively.

The frame generator (3) generates two pulse signals:

• frame signal,
• window signal.

The sizes of the “frame” (a x b) and the “window” (A x B) are related by the following dependencies:

a=X/Km, A=kX/Km,
b=X/Km, B=kY/Km,

where X is the horizontal raster size;
Y is the vertical raster size;
K is the multiplicity coefficient, selected within the range (1…2).

The horizontal and vertical offsets of the “frame” and “window” are determined by the values ​​of the constant voltages supplied to the input of the corresponding analog-to-digital converter (ADC) and are adjusted separately by potentiometers on Rx and RY. The latter, installed in the guidance unit (6), are the operational controls for the “frame” and “window”, and at the same time – the horizontal (6-3) and vertical (6-4) position sensors.

It should be noted that for high-quality operation of the TV camera, high-precision adjustment (alignment) of the guidance unit (6) must be performed during its setup. As a result of the adjustment, at all set positions of the Rx and RY potentiometer sliders, such a spatial position of the sensor (5) relative to the beam splitter (4) must be ensured so that the enlarged optical image projected onto its photo target is centered horizontally and vertically. A prerequisite for performing the adjustment is a high precision class of manufacturing all mechanical elements of the drives and the use of Rx and RY potentiometers with a linear dependence of the change in resistance on the angle of rotation.

The sensor (5) is characterized by its operation in the external synchronization mode by feeding the receiver synchronization signal (RSS) from an external generator to the “sync” input. These requirements are met by the South Korean SBC-4KXE camera module, which is based on a CCD matrix with a 1/3-inch target format and a number of elements of 795(H) x 596(V). It is obvious that the SBC-4KXE module can also be used as a sensor (2).

Block (9) is designed to perform a constant time delay of the horizontal sync pulse by half the line period and the vertical sync pulse by half the half-frame period, i.e.:

t zsf = Ts /2, t zkf = Tk/2,

where t zs f is the line delay, t zkf is the frame delay, Ts is the line period, Tp is the half-frame period.

The unit (10) is designed to perform a controlled time delay of the synchronization pulses from the outputs of the unit (9).

The adjustable delay by line t зср satisfies the relationship:

0 < t зср < Tс , (1)

and the adjustable delay by frame t зкр satisfies another relationship:

0 < t зк р < Tп , (2)

Control of the change in the delays t зс р and t зк р is carried out, respectively, from the first and second position sensors of the guidance unit (6).

Let's distinguish two modes in the operation of the TV camera:

• “Fragment selection” (mode 1);
• “Combined image” (mode 2).

Regardless of the camera operating mode, the input optical image is projected onto the photo target of the first sensor (2) via the optical path: first lens (1), semi-transparent mirror (4-1), collective lens (4-2), reflective mirror (4-3), second lens (4-4). At the same time, a fragment of this image magnified by a factor of Km is projected onto the photo target of the second sensor (5) via another optical path: first lens (1), semi-transparent mirror (4-1). In addition, regardless of the camera operating mode, the shaper (3) generates “frame” and “window” signals.

Let the logical “1” signal be fed to the control input of the switch-mixer (8), then the TV camera operates in the “Fragment Selection” mode, and at its output the video signal of the normal-scale image from the sensor (2) is mixed with the “frame” signal.

At the same time, an optical image magnified by a factor of Km, the geometric center of which is aligned with the center of the “frame”, is projected onto the photo target of the sensor (5).

Let us assume that the frame is set by the operator to select the central fragment of the presented image. In this case, the sliders of the Rx and RY potentiometers in the guidance unit (6) will occupy the middle position, and in unit (10) the line and frame sync pulses coming from unit (9) will be delayed by half the line period and half the half-frame period, respectively. That is, in unit (10) in this case the delay durations are:

t зс р =T с /2, (1-1)
t зк р =T п /2, (2-1)

For illustration purposes, Fig. 2c shows the time position of the horizontal synchronization pulse at the output of the unit (10) relative to the input pulse of the lines shown in Fig. 2b. Then the synchronization signal at the output of the generator (11) coincides in phase with the synchronization signal at the output of the sensor (2). As a result, the raster of the normal image formed by the sensor (2) and the raster of the enlarged image from the sensor (5) act synchronously and in phase with each other, and coincide when spatially superimposed, as shown in Fig. 3a. Note that in Fig. 3 the raster of the normal image is marked with a single hatch.

Fig. 2. Timing diagram explaining the formation of delays by line

Fig. 3. Relative position of the rasters of the first and
second sensors depending on the values ​​of the sync pulse delays by line and frame

If a logical “0” signal is fed to the control input of the unit (8), the TV camera switches to the “Combined image” mode. As a result, an image is formed at the output of the unit (8), and consequently of the camera, consisting of a signal of an enlarged image of the selected fragment in the center and a signal of a normal image on its remaining part.

Let us assume that the operator needs to check the enlarged image of another fragment, for example, located in the lower left part of the video monitor screen. Then he should return to the “Fragment Selection” mode. Then the operator selects a new image fragment using the frame by aiming the sensor (5), using the aiming unit (6) for this. In this case, the sliders of the Rx and RY potentiometers in the unit (6) will take a new position, and in the unit (10) another delay of the input horizontal and vertical sync pulses will be performed, satisfying the relations:

0 < t зс р < Tс /2, (1-2)
0 < t зк р < Tп /2,  (2-2)

For illustration purposes, Fig. 2g shows the time position of the line sync pulse at the output of the unit (10) relative to the input line pulse shown in Fig. 2b. Then the synchronization signal at the output of the generator (11) is ahead in phase of the synchronization signal at the output of the sensor (2). As a result, the raster of the enlarged image formed by the sensor (5) is shifted relative to the raster of the normal image formed by the sensor (2), as shown in Fig. 3b. Therefore, after switching the TV camera to the “Combined Image” mode, the enlarged section will be located on the video monitor screen in the initial selection zone, i.e. where the frame was installed.

Let's consider a situation where the operator needs to check the enlarged image of another fragment, marked with a frame in the upper right part of the monitor screen. In this case, the sliders of the Rx and RY potentiometers in block (6) will take a different position, and in block (10) a different delay of the input horizontal and vertical sync pulses will be performed, satisfying the relations:

Tс/2 < t зс р < Tс , (1-3)
Tp/2 < t zk p < Tп , (2-3)

For illustration, Fig. 2d shows the time position of the line sync pulse at the output of the unit (10) relative to the input line pulse shown in Fig. 2b. Then the synchronization signal at the output of the generator (11) lags in phase behind the synchronization signal at the output of the sensor (2). As a result, the raster of the enlarged image formed by the sensor (5) is shifted relative to the raster of the normal image formed by the sensor (2), as shown in Fig. 3c. Therefore, in this case, the enlarged image will also be located on the video monitor screen in the initial selection zone.

Note that relations (1-1), (1-2), (1-3) are special cases of relation (1), and relations (2-1), (2-2) and (2-3) are special cases of relation (2).

It is absolutely similar that if any other frame position is selected in the TV camera, the enlarged fragment will be located in the combined image only in the zone of the initial selection. The resolution indicator remains unchanged within the entire combined image, since the optical image of the fragment is “perceived” by the light-sensitive elements of the sensor (5) at the same density of their arrangement per unit length as the optical image of the rest of the part, recorded by the elements of the sensor (2).

In conclusion

The proposed TV camera has the potential to perform a spectrozonal or polarization analysis of the transmitted images. To do this, it is necessary to additionally introduce filters with different spectral transmission or with different orientation of the polarization plane into the optical channels of the camera. For example, to analyze vegetation, a filter with a transmission across the spectrum up to and including the red region can be used at the first output of the beam splitter, and an infrared filter can be used at the second output of the beam splitter. When the enlarged fragment of the image records a section of the cover with high reflection in the infrared region, its contrast against the background of the rest of the image is significantly higher. Therefore, it is advisable to set the TV camera constantly to mode 2, and to perform the analysis process using the “electronic magnifying glass” method.

In another example, to analyze water bodies, it is preferable to use polarizing filters whose polarization planes are mutually perpendicular.

Literature

1. Smelkov V.M. PiP method for a 24-hour security camera//Special equipment. 2002, No. 3, pp. 21–24.
2. Patent 2199828 RF. MKI7 HO4N 5/225, 5/238. Television camera with selective scaling/V.M. Smelkov, Yu.A. Smolyakov, V.E. Antonov and A.P. Ogarkov//B.I, 2003, No. 6.
3. Smelkov V.M. Television camera for covert surveillance and automated security//Special equipment. –2001, No. 3, pp. 20–23.