RU2683944C1 - Method of controlling the sensitivity of a television camera on a ccd matrix in conditions of complex illumination and / or complex brightness of objects - Google Patents

Abstract

FIELD: television technology.SUBSTANCE: invention relates to television technology and is oriented to use in television cameras, made on the basis of matrix television sensors based on the technology of charge-coupled devices (CCD), in which electronic sensitivity control is provided due to changes in the intraframe accumulation time. Duration of accumulation of information charges per frame is determined separately for each said individual target from the peak value of the video signal generated at the output of the photodetector during the corresponding time interval within the forward stroke along the frame, wherein frequency of element-by-element transfer of charge packets is reduced twice, and output register is made in the form of two adjacent linear registers acting in turn.EFFECT: technical result is to automatically increase sensitivity for fragments of a television frame with low illumination (brightness) of corresponding objects with simultaneous implementation of saving power of the photodetector.1 cl, 8 dwg

Description

The present invention relates to television technology and is focused on the use in television cameras made on the basis of matrix television sensors using charge-coupled device (CCD) technology, in which electronic sensitivity adjustment is provided by changing the intraframe accumulation time.

The closest in technical essence to the claimed invention should be considered a method of controlling the sensitivity of a television camera on a CCD matrix [1], which consists in the fact that on the target (storage section) of the CCD matrix with the organization "personnel transfer" with the period of frames accumulate information charges, transfer information charges with a frame transfer frequency from the accumulation section to the memory section, transferring information charges line by line from the memory section to the output register in the reverse scan interval, and in the interval of the forward horizontal scan, the information charges are transferred element-wise from the output register to the output block of the CCD matrix with the simultaneous conversion of the charge into the video signal voltage (CPS), and in the accumulation section in the interval between the frame transfer of the current frame and the accumulation cycle of the subsequent frame, excess charges are transferred to the substrate of the photodetector by technological organization of the anti-blooming region and the electronic shutter in the accumulation section, while the personnel transfer of information charges from section accumulation in the memory section is carried out at the final interval of the interval of the reverse motion of the frame scan with the corresponding time delay of the duration of accumulation of information charges, and in the interval of the interval of the reverse motion of the frame scan, prior to the transfer of information charges, the memory section is cleaned of stray charges due to their movement with the frequency of personnel transfer to the output register or due to their removal into the substrate of the photodetector by technological organization and antiblyumingovoy region and the electronic shutter in the memory section.

Note that for the prototype [1], the current loading by the charge packets of the output register of the CCD is performed within the time interval, which in the television scan takes the interval τ _o.h.s - the length of the reverse stroke along the line. The transfer of these charge packets along the output register and reading in the SPZN is performed with a frequency of element-wise transfer ƒ _e .

This method of controlling sensitivity is able to solve the problem of adapting the photodetector to the conditions of light overloads on the target.

It is assumed that the composition of the television camera that implements this sensitivity control method includes an automatic accumulation time adjustment device (ARVN), and the control voltage generated by this unit via the feedback circuit determines the exposure time (accumulation) of the CCD by the CCD per frame depending on the illumination (brightness) of the observed plot. Obviously, the memory section of the CCD prototype [1] is shielded from light.

However, when working in conditions of complex illumination and / or complex brightness of objects, when high illumination (brightness) in some parts of the field of view is accompanied by low illumination (brightness) in other parts of it, the prototype sensitivity control method implemented in a television camera [1] cannot fundamentally cope with the situation without avoiding image quality degradation throughout the frame.

This is explained by the fact that under these conditions, the automatic accumulation time adjustment (ARVN) of the television camera, working on this sensitivity control signal, counts its voltage using an amplitude detector according to the peak or average value of the video signal generated at the output of the photodetector. But it spreads its result, namely: the duration of the accumulation time, to all elements (pixels) of the target.

The disadvantage of the sensitivity control method in the prototype [1] is that there is a mode of limited accumulation (in time) for those portions of the image that are observed at low illumination (brightness) of the objects corresponding to them.

The objective of the invention is the organization in the automatic mode of increasing sensitivity for these fragments of a television frame by increasing the accumulation time for them while realizing the energy savings of the photodetector.

The problem in the claimed method of controlling the sensitivity of a television camera on a CCD matrix, based on the fact that information charges are accumulated in the frame of the CCD matrix with the organization "personnel transfer" with a frame period in accordance with the control voltage for the ARVN; transfer information charges with a frame transfer frequency from the accumulation section to the memory section shielded from light; line-by-line transfer information charges from the memory section to the output register in the interval of the reverse stroke horizontal scanning; and in the interval of the forward horizontal scan, the information charges are transferred element-wise from the output register to the output block of the CCD matrix with the simultaneous conversion of the charge into the video signal voltage (CPS), and in the accumulation section in the interval between the frame transfer of the current frame and the accumulation cycle of the subsequent frame, excess charges are allocated the substrate of the photodetector by technological organization of the anti-blooming region and the electronic shutter in the accumulation section, while the personnel transfer of information charges from section accumulation in the memory section is carried out at the final interval of the interval of the reverse motion of the frame scan with the corresponding time delay of the duration of accumulation of information charges, and in the interval of the interval of the reverse motion of the frame scan, prior to the transfer of information charges, the memory section is cleaned of stray charges due to their movement with the frequency of personnel transfer to the output register or due to their removal into the substrate of the photodetector by technological organization and the anti-blooming area and the electronic shutter in the memory section, it is decided that the accumulation section of the CCD matrix is horizontally divided into n isolated targets of the same format, which have parallel control of the photo-reception and scanning processes, while the control voltage is equal to the ARVN and, accordingly, the duration of the accumulation of information charges per frame is determined separately for each of this individual target according to the peak value of the video signal generated at the output of the photodetector during the corresponding time interval within the forward stroke in the frame, while the frequency of the element-wise transfer of the charge packets ƒ _{e is} reduced by half, and the output register is made in the form of two adjacent linear registers acting alternately on the two-channel overvoltage protection device, each of these linear registers contains half the elements of the number of pixels for each photodetector line, and in the interval τ _o.h.s. they charge both linear "registers sequentially in time and separately for the odd and even pixels of this line with charge packets of the current information line, and the number of phase electrodes for an individual pixel in both linear registers should be even, amounting to 2 or 4.

Comparative analysis with the prototype [1] shows that the inventive method is characterized by the presence of the following features:

- a condition for the implementation of preliminary actions with the CCD sensor of a television camera, namely: dividing the storage section of the photodetector into n isolated from each other targets with the same format, which are controlled in parallel;

- performing in a television camera parallel steps to set the duration of the accumulation of charges per frame for each of the n targets of the sensor;

- the implementation of energy saving matrix photodetector by halving the frequency of the element-by-element transfer ƒ _e as well as by the new organization of its constituent units: the output register and the overhead current detection device. According to the claimed method, the output register consists of two parallel operating registers, and the SPS from a single-channel block becomes two-channel. In this case, the charge signals will be recorded by the SPS in the correct phase ratio due to the choice of an even indicator for the number of phase electrodes as applied to an individual element of these registers.

It is important to note the following. From the monograph [2, p. 153] it is known that in an n-channel CCD with a cell size of 30 μm operating at a frequency of 1 MHz, a charge packet of 0.5 pC consumes a specific power of about 2.8 nW / cell. And this value grows like a square of the operating frequency!

When organizing such a control mode for element-wise transfer of charge packets for a CCD photodetector, the resolution of the video signal of the observed plot remains unchanged, and the power consumption of a television camera based on such a sensor is “braked”, compensating for all or part of the associated energy costs.

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

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

In FIG. 1 shows the circuit organization of a CCD matrix with four isolated targets (n = 4); in FIG. 2 is a structural diagram of a device explaining the implementation of the proposed method for controlling sensitivity, and is, in fact, the design scheme of the new ARVN photodetector; in FIG. 3 is an example of a circuit diagram of a peak detector, four of which are accepted (indicated) in FIG. 2; in FIG. 4b-4d with respect to the temporary position of the blanking pulse of the rows shown in FIG. 4a, plots of control signals are shown to obtain the necessary areas (“windows”) of photometric measurements of the charge relief of the photodetector; in FIG. 5b is a diagram illustrating the temporary position of the output signal of the ARVN device (relative to the frame blanking pulse shown in FIG. 5a); in FIG. Figure 6 shows a fragment of the cross section of the accumulation section or memory section of the CCD matrix, illustrating the physical processes that accompany the technological organization of the anti-blooming region and the electronic shutter in these sections of the photodetector; in FIG. 7 shows the position of the four “windows” of the photometry of the CCD array accumulation section under conditions of its complex illumination and / or complex brightness; in FIG. 8 is a diagram explaining the control of the first and second linear registers of the photodetector, in which the control of the element-wise transfer is carried out according to the claimed method.

Matrix 1 at the CCD (Fig. 1) with the organization of “personnel transfer” is made on a silicon crystal and consists of storage sections 1-1 connected in series with charge communication, memory sections 1-2, the first linear register 1-3- (1), the second linear register 1-3- (2) and BPZN 1-4. The dash-dotted lines in FIG. 1 show the selection of four isolated targets with the same format on the photodetector storage section 1-1, i.e. n = 4. We introduce the designation of these sensors, respectively, as: 1-1-1, 1-1-2, 1-1-3, and 1-1-4.

Similarly to the prototype [1], according to the first embodiment of the CCD matrix, each of these targets contains an anti-blooming region and an electronic shutter GA built into its semiconductor structure.

According to the second embodiment of the CCD matrix, the anti-blooming area and the electronic shutter GB are additionally integrated in the memory section 1-2.

Note that the application of these technological features is due to the need to eliminate spurious charges in the photodetector under conditions of its light overloads. For a television camera, such overloads are often a concomitant phenomenon and observation in conditions of complex illumination and / or complex brightness of objects.

As shown in FIG. 6, the GA shutter is an “electronic” shutter of the accumulation section. For a photodetector with three-phase charge transfer and with an n-conduction channel, if there is a low potential (relative to the substrate) on the GA gate, the latter is closed, and potential wells under the target phase electrodes are isolated from the sink region due to this barrier bias. Then, at the photographic target itself, the process of accumulation of charge photoelectrons under the Ф2Н electrodes is initiated (see Fig. 6a).

When a high potential is applied to the GA gate, the potential barrier is removed, the gate opens, and the accumulation of photoelectrons is excluded on the target. This is explained by the fact that the carriers, not lingering in potential wells under the Ф2Н phase electrodes, rush into deeper wells created by the potential DA in the drain region, and then recombine into the photodetector substrate (see Fig. 6b).

Obviously, the “electronic” shutter GB with the anti-blooming area controlled by the DB potential is realized in exactly the same way when they are built into the 1-2 memory section around the phase electrodes Ф2П, as shown in FIG. 6.

For all four isolated targets of the CCD matrix 1, parallel control is applied, which ensures the process of the current accumulation of the charge relief in accordance with the duration specified by the output pulse at the input of the electronic shutter of the sensor of each target, i.e. through GA1, GA2, GA3 and GA4, as shown in FIG. one.

The organization of this parallel control can be carried out by “multiplying” the pulse signals using external buffer stages for finished circuits that implement a set (set) of necessary control voltages.

The method for controlling the sensitivity of a television camera proposed in this technical solution can be implemented for two other technological options for a CCD matrix, i.e. for sensors manufactured according to the method of “line transfer” and “line-frame transfer” [3, p. 134-137]. In this case, the photodetector region of these devices, in which vertically arranged lines of photosensitive elements alternate with vertical lines of pixels isolated from light, should be technologically prepared in the same way as the accumulation section of the CCD matrix of personnel transfer, i.e. by dividing horizontally into n isolated targets of the same format, and the output register of these sensors is made in the form of two adjacent linear registers, acting alternately on a two-channel overvoltage detector.

Consider the previously announced block diagram of ARVN in FIG. 2. It contains a “window” signal shaper 2 for sensor photometry, a first peak detector 3 and a first pulse-width modulator (PWM) 4 connected in series, a second peak detector 5 and a second PWM 6 connected in series, a third peak detector 7 and a third connected in series PWM 8, as well as the fourth peak detector 9 and the fourth PWM 10 connected in series. The information inputs of all four six peak detectors (3, 5, 7, 9) are connected to the output of the photodetector 1, and the control inputs of these peak detectors - to first, second, third and fourth outputs of the driver 2. The fifth output of the generator 2 is connected to the reset inputs of all four peak detectors (3, 5, 7, 9).

The output of block 4 is connected to the electronic shutter GA1 of target 1-1-1, the output of block 6 is connected to the electronic shutter GA2 of target 1-1-2, the output of block 8 is to the electronic shutter GA3 of target 1-1-3, and the output of block 10 is to the GA4 electronic shutter target 1-1-4. It is assumed that in the television camera from its microcontroller to the input of the shaper 2 signals of the "windows" are triggered pulses and synchronization.

Shaper 2 is designed to implement four pulse signals (see Fig. 4b - 4d), which are fed to the control inputs of the peak detectors. To receive these signals, a three-digit binary counter can be used [see, for example, 4, p. 168-170].

By transmitting through the former 2 from the microcontroller of the television camera, a positive-polarity frame sync pulse is supplied to perform at the beginning of each frame zeroing (reset) of all four peak detectors (3, 5, 7, 9).

Peak detectors are designed to record the maximum level of an analog video signal supplied to their information inputs in the interval of the presence of a high level of a pulse signal at their control inputs.

Each of the peak detectors can be made on the basis of two operational amplifiers (op amps) according to the scheme proposed in [5, p. 301]. A feature of the circuit shown in FIG. 3, is the selection of the first (input) op-amp. This op-amp must additionally have a control input for implementing external control of the reserve power and operating point using an external bias voltage. An example of such an op-amp is the CA3078T microcircuit manufactured by RCA (USA).

The pulse signals that must be supplied to the control inputs of all four peak detectors (3, 5, 7, 9) are shown in the time diagrams shown respectively in FIG. 4b, 4c, 4g, 4d.

The output signals of peak detectors 3, 5, 7, and 9 are the control voltages for blocks 4, 6, 8, and 10 of a pulse-width modulator (PWM), the outputs of each of which will generate a "stand-alone" digital accumulation signal (see Fig. 5b )

до его минимального отсчета

в зависимости от величины входного управляющего напряжения.Note that this signal may vary during the frame from the maximum value of the reference

to its minimum countdown

depending on the magnitude of the input control voltage.

To fulfill its functional “duty”, a digital accumulation signal is fed to the control input of the sensor, which is its “electronic” gate GA.

Obviously, such a process of optimized accumulation of charges on sections 1-1 of the CCD matrix, depending on the level of illumination of the monitored scene, will occur in parallel and on all four targets 1-1-1, 1-1-2, 1-1-3 and 1- 1-4 by controlling through the corresponding valves GA1, GA2, GA3 and GA4 (see Fig. 1).

The devices shown in FIG. 1-2, work as follows.

An optical image of the observed scene under conditions of complex illumination and / or complex brightness of objects is projected onto the accumulation section 1-1 of the CCD matrix, and, therefore, onto all four of its targets (1-1-1, 1-1-2, 1- 1-3, 1-1-4).

Let in our example shown in FIG. 7, in conditions of high illumination of the observed plot, the area occupied by the “Window” 1 appears; in low light conditions - “Window” 2; in conditions of medium low light - “Window” 3, and in conditions of significantly reduced light - “Window” 4.

In the forward range of each television frame, the process of optimized accumulation of charges in the photosensitive pixels of all four targets 1-1-1, 1-1-2, 1-1-3 and 1-1-4 is proportional to the illumination of the controlled plot.

During the interval of the subsequent interval of the reverse scan of the frame scan, the charges of all the lines involved in the accumulation are transferred to the light-shielded pixels located on sections 1-2 of the memory.

Then, in a new personnel cycle, another charge “picture” is accumulated, and the charge packets accumulated in the previous frame are transferred from the memory section 1-2 to the periphery of the photodetector crystal, loading both charges in the horizontal scanning interval (τ _o.s. ) with new charges linear register. Note that in FIG. 1, the loading gates are shown by thickened lines.

Let us consider in more detail the “mechanism” of charge loading using the time diagrams of the signals shown in FIG. 8.

In FIG. Figure 8a shows a plot of a signal for a horizontal blanking pulse of a television scan, which is active during the interval τ _o.h.s with a period of lines T _s .

In FIG. 8b, FIG. 8c shows diagrams of pulsed signals that control the loading gates of the second linear register 1-3- (2) and the first linear register 1-3- (1), respectively.

Note that the first linear register 1-3- (1) is universal, providing transfer of charge packets in two directions, namely: both along the register and across (through), i.e. in the cells of the second linear register 1-3- (2).

To implement the second function, additional electrodes are installed in the gaps between the register elements 1-3- (1) having a pixel width (they are indicated by a dotted line in Fig. 1), interconnected and connected to a constant voltage, the value of which is not less than the control potential of the charge transfer. These additional electrodes simultaneously fulfill another important role, namely: they exclude charge losses during transfer in the register 1-3- (1). For this reason, exactly the same additional electrodes are installed in the gaps between the elements of the register 1-3- (2).

In FIG. 8g, FIG. 8d shows diagrams of pulsed signals that control the operation of both linear registers in parallel with respect to a two-phase charge transfer charge system, where T _e = 1 / - _e is the period of element-wise transfer of charge packets.

τ_о.х.с - интервале активного действия импульса на фиг. 8б через открытый затвор загрузки в регистр 1-3-(2), в ячейки под первыми фазными электродами, будут поступать заряды первого, третьего, пятого и других нечетных элементов этой строки.In the interim

τ _o.h.s - the interval of the active pulse in FIG. 8b, through the open loading gate to the register 1-3- (2), the cells under the first phase electrodes will receive the charges of the first, third, fifth and other odd elements of this row.

τ_о.х.с - интервале активного действия импульса, изображенного на фиг. 8в, через открытый затвор, в ячейки под первыми фазными электродами, будет загружаться зарядами регистр 1-3-(1), но применительно для второго, четвертого, шестого и других четных элементов этой строки.And in the subsequent interval

τ _o.h.s - the interval of active action of the pulse depicted in FIG. 8c, through an open gate, registers 1-3- (1) will be loaded into the cells under the first phase electrodes by charges, but with reference to the second, fourth, sixth and other even elements of this row.

Note that in this time interval, the charge packets loaded earlier in the linear register 1-3- (2) remain there “in their places”, while the potential wells of this register are in storage mode.

The charge packets of each line in each subsequent frame cycle are read element-wise in BPS 1-4, forming at its output an analog video signal of the observed image.

The technical result of the proposed solution is ensured by the fact that in a television camera the optimal indicators for the accumulation time (T _n ) will be obtained in a fully automatic mode for all n sections of the target area of the CCD matrix.

Therefore, in comparison with the prototype [1], an increased signal-to-noise ratio (ψ) of the generated video signal and, accordingly, an increase in sensitivity for those portions of the image that are recorded at low illumination (brightness) of the objects corresponding to them, and, no less important, will be achieved - with the implementation of energy saving sensor.

Currently, all the elements of the circuitry organization of the aforementioned matrix photodetectors on the CCD, as well as the blocks and elements of the commented block diagram of the ARVN sensor device that implements the proposed method for controlling the sensitivity of a television camera, are mastered or can be mastered by the domestic industry.

Therefore, the present invention should be considered as meeting the requirement for industrial applicability.

INFORMATION SOURCES

1. RF patent No. 2399164. IPC H04N 5/335; H04N 5/217. A method of generating an image signal. / V.M. Smelkov // B.I. - 2010. - No. 25.

2. Seken K., Thompset M. Instruments with charge transfer. Translation from English - "The World", 1978.

3. Vlado Damianowski. CCTV. Bible CCTV. Digital and network technology. Translation from English - M .: "IS-ES Press", 2006.

4. Tokheim R. Fundamentals of Digital Electronics. Translation from English - M.: “World”, 1988.

5. Peyton A. J., Walsh V. Analog electronics on operational amplifiers. Translation from English - M.: “BINOM”, 1994.

Claims (1)

A method for controlling the sensitivity of a television camera on a CCD under complex lighting conditions and / or complex brightness of objects, namely, that information charges are accumulated on the target (storage section) of the CCD with the “frame transfer” organization with a frame period in accordance with the control voltage for automatic adjustment of accumulation time (ARVN), transfer information charges with a frame transfer frequency from the accumulation section to the memory section shielded from light, transfer information line by line charge charges from the memory section to the output register in the interval of the flyback of the horizontal scan, and in the interval of the forward stroke of the horizontal scan, the information charges are transferred element-by-bit from the output register to the output block of the CCD with the simultaneous conversion of the charge into the voltage of the video signal (VCR), the interval between the personnel transfer of the current frame and the accumulation cycle of the subsequent frame, excess charges are diverted to the substrate of the photodetector by technological organization of the anti-blooming region and the electronic shutter in the accumulation section, while the personnel transfer of information charges from the accumulation section to the memory section is carried out at the final interval of the reverse scan interval of the personnel scan with the corresponding time delay of the duration of the accumulation of information charges, and in the interval of the reverse scan interval of the vertical scan, preceding the transfer information charges, carry out the cleaning of the memory section from spurious charges due to their movement with a frame transfer frequency and in the output register or due to their removal to the photodetector substrate by technological organization of the anti-blooming area and electronic shutter in the memory section, characterized in that the accumulation section of the CCD matrix is horizontally divided into n isolated targets with the same format, which have parallel operation control of photo-reception and scanning processes, while the value of the control voltage of the ARVN and, accordingly, the duration of the accumulation of information charges per frame are determined separately for each of this separately taken target according to the peak value of the video signal generated at the output of the photodetector during the corresponding time interval within the direct course of the frame, while the frequency of the element-wise transfer of charge packets ƒ _{e is} reduced by half, and the output register is made in the form of two adjacent linear registers acting alternately on a two-channel OTZN, with each of these linear registers containing half the elements of the number of pixels for each photodetector line, and in the interval τ _o.h.s. they charge both linear "registers sequentially in time and separately for the odd and even pixels of this line with charge packets of the current information line, and the number of phase electrodes for an individual pixel in both linear registers should be even, amounting to 2 or 4.

Images