Selection of exposure time for a security camera in the single-shot video signal generation mode.
Vyacheslav Mikhailovich SMELKOV,
Candidate of Technical Sciences
SELECTION OF EXPOSURE TIME FOR A SECURITY CAMERA IN THE SINGLE-SHOT VIDEO SIGNAL GENERATING MODE
The article analyzes possible methods for automatic selection of accumulation time depending on the parameters of the video signal in television cameras that provide the ability to photograph an intruder on television in the “MONOSHOT” mode.
A technical solution is proposed at the level of inventing a model of a television camera, which, in conditions of low illumination at a controlled object, allows for real-time observation (in the “TV” mode), as well as measuring the illumination of a scene from a video.
When a command is given to register an event, the television camera switches to the “MONOSHOT” mode, where the optimal exposure time is set based on the measurement result obtained in the “TV” mode.
The well-known mode of single-shot formation of a video signal (television photography) in TV cameras on matrix CCDs, called by its French inventors MONOSHOT” [1], creates certain difficulties in choosing the most important parameter from the image signal – the optimal exposure time of the photodetector.
The fact is that the choice of the accumulation duration in this mode cannot be performed in the classical way with the help of automatic sensitivity control devices (ASC), the operation of which is based on the analysis of the video signal of the previous frame via the feedback circuit, since the latter is excluded here in principle.
On the other hand, only the video signal selected as the control signal for the AFC ensures ideal matching of the input and output parameters of the machine according to the spectral characteristic.
However, an advanced assessment of the photoelectric process occurring on the CCD target can be made if we use the previously proposed method for generating a control signal for the AFC of small-frame television cameras according to a domestic invention [2].
The unique advantages of the control signal obtained by this method are:
- absence of time delay in relation to the accumulation process on the target;
- non-destructive nature of reading the information contained in it;
- the ability to obtain this information in a short period of time (up to 20 μs).
In the description of the invention [3], the author of this article published a method for automatically selecting the duration of accumulation of the CCD matrix according to the method [2] as applied to a TV camera operating in the single-shot image signal formation mode, which was named “MONOSHOT-AUTO”.
In this mode, the accumulation process in the CCD is carried out in time in separate “portions”, but with the measurement of the level of the charge relief of the photo target after each of them.
The charge relief is assessed by the current in the circuit of one of the phase electrodes of the accumulation section simultaneously for all of its elements (pixels).
When the total measurement result reaches the threshold that determines the optimal relief value, the process of accumulation of the information frame ends.
The undoubted advantage of this method is the wide range of possible selection of the exposure time of the photodetector depending on the illumination on the controlled object, which is estimated within the range from 100 μs to 10 s, i.e. determined by the physical limitations of the CCD matrix.
Unfortunately, the proposed method for generating a control signal dependent on the parameters of the video signal can only be implemented by coordinateless scanning, which excludes the choice of the “evaluated window” within the presented image frame.
In addition, due to the large capacities of the phase electrodes of the accumulation section, the noise variance (s2) of the video signal reading channel using this method is always greater than the noise variance of the matrix output device (s1) when reading the image signal in the traditional way, i.e. there is a large noise level in the control signal.
For a security camera used in the dark, when an adequately selected exposure of the photo receiver is more than one frame long, the work [4] proposes a mode of television photography of an intruder using the MONOSHOT-ZOOM-AUTO method, accompanied by another technical effect – an increase in the scale of the image of the recorded photograph.
The control signal, determined by the video signal, for automatic evaluation of the exposure duration is here formed in parallel from the second sensor, operating in the standard periodic TV mode, and therefore is the result of measurement due to the standard coordinate scan.
If the requirement to increase the image of the intruder is optional, then it is possible to significantly simplify the hardware of the TV camera operating in the “MONOSHOT-AUTO” mode, while maintaining the advantages of the control video signal.
The most important prerequisite for the implementation of this opportunity is the appearance on the market of frameless television cameras (camera modules), which provide two operating modes: “TV” and “MONOSHOT” with sequential switching from one to the other.
An example of such a product is the FTM-12 camera module from Philips (Netherlands) [5].
Below is a technical solution for a security television camera based on a similar sensor.
The structural diagram of the television camera, the device of which is recognized as an invention [6], is shown in Fig. 1.
The camera contains a lens (1), a video and sync pulse sensor (2), a peak detector (3), an analog-to-digital converter (ADC) in position (4), an accumulation duration generator (5), a single-shot multivibrator (6), the first RS trigger (7), the second RS trigger (8), an “OR” element (9), and an inverter (10).
Input signals: “Start”, “Reading” and output signals: “Video”, “Synchro” – provide the necessary interface between the TV camera and the user’s computer.
Fig. 1. Structural diagram of the TV camera
Let’s consider the operation of the TV camera using the timing diagram shown in Fig. 2. Note that it coincides with the diagram published in the article [4].
Fig. 2. Timing diagram explaining the operation of the TV camera
Until the moment of arrival of the pulse signal at the “Start” input, the logical “1” level is maintained at the inverse output of the RS trigger (8), and consequently at the “Mode Select” control input of the sensor (2). At the same time, a high logical level is maintained at the “Accumulation” control input of the sensor (2), and at its “Read” control input, the logical “0” level is present.
The TV camera operates in the “TV” mode, and a full television signal according to the television standard is generated at its output.
A logical “1” is maintained at the inverse output of the RS trigger (7), so the generator (5) is blocked by a high logical level at the control input and does not count the clock synchronization pulses coming from the sensor (2).
Let a trigger pulse arrive at the “Start” input at the moment t0, as shown in Fig. 2a. Then the RS triggers (7) and (8) go to the “0” state at the inverse output.
At the same time, the trigger pulse resets the peak detector (3) and triggers the monostable multivibrator (6).
The latter generates a pulse signal at the output (Fig. 2b), the duration of which is the resolution interval of the preliminary recording-installation operation in the generator counters (5).
During the interval t0 …. t1, the peak detector (3) measures the current value of the video signal. The constant voltage from the output of the peak detector (3) is then converted in the ADC (4) from analog to digital form and fed to the setup inputs of the shaper counters (5). By the time t1 (Fig. 2b), the recording and setup of this number in the shaper counters (5) must be completed.
Starting from the time t1, the logical level “0” is set at the output of the “OR” element (9), and, consequently, at the control input “Mode selection” of the sensor (2). Due to this, from the time t1, the TV camera switches to the “MONOSHOT” mode.
Note that from the moment t0, there is a low logical level at the control input of the shaper (5), i.e. the blocking at this input is removed. The counters of the shaper (5) count the data increment, and at its output, starting from the moment t1, a low logical level is set (Fig. 2d).
Therefore, the sensor (2) goes into a state of long-term accumulation of informative charges. The duration of accumulation in the sensor (2) is set to be optimal according to the criterion of the maximum signal/noise ratio of the video signal of the image being taken, which is achieved by preliminary calibration of the TV camera. After the end of the accumulation of charges in the sensor (2), the counters of the generator (5) are reset and the RS trigger (7) is set to the state “1” by the inverse output (moment t2 in Fig. 2d).
The counters of the shaper (5) are set to zero, and the sensor (2) goes into the “non-accumulation” state, since at its control input “Accumulation”, starting from this moment, a high logical level is formed. The “non-accumulation” state of the CCD matrix means that the collection of photoelectrons in the potential wells of the photodetector is excluded, and they all fall into the drain region and recombine.
Let us assume that at the subsequent moment t3 (Fig. 2d) at the control input of the camera “Reading” the logical “0” level is replaced by the logical “1” level.
Further, as in work [4], when the high level in the signal “Setting the reading duration” coincides with the end of the nearest frame blanking pulse (moment t4 in Fig. 2d), reading of the charge relief of the information frame begins, which continues during the interval t4 … t5.
As a result, an electrical signal of a single frame is formed at the output of the Video camera (Fig. 2g).
Note that the duration of this signal, taking into account the frame blanking pulse, is Tk = 20 ms and corresponds to the period of half frames according to the television standard.
At the moment t6 (Fig. 2d) the external pulse at the control input “Reading ends. Through the inverter (10) the resulting voltage drop sets the RS trigger (8) to the state “1” on the inverse output.
As a result, a high logical voltage level is applied to the control input “Mode selection”, and the camera, starting from the moment t6, is again set to the “TV” mode.
To implement the optimal accumulation of informative charges in the sensor (2), the camera must be calibrated according to the method described in [3], and the maximum capacity of the former counters (5) must be selected taking into account the growth of the collection of dark charge carriers and the permissible visibility of the background heterogeneity in the image.
Conclusion
There is no doubt that control determined by the video signal is preferable when choosing the optimal exposure time for television cameras operating both in the “TV” and “MONOSHOT” modes. The author hopes that the practical methods for solving this problem for the “MONOSHOT” mode described in this article will be in demand.
Literature
- French application No. 2589301 dated 10/28/85, IPC H04N3/15, 5/238. Electronic shutter device. Applicant – I2S (France).
- USSR Author's Certificate 1417210, IPC3 H04N 5/228. Method for generating a signal for controlling the sensitivity of a television camera on a CCD matrix/A.N. Kulikov, L.I. Khromov//B.I. 1988.- № 12.g.
- Patent 2146080 RF. MKI7 HO4N 3/14, 5/335. Device for single image signal formation/V.M. Smelkov, V.N. Mikhailov, V.Ya. Maklashevsky//B.I. 2000, № 6.
- Smelkov V.M. Security television camera: a new solution using the optical-electronic scaling method//Special equipment. 2002, № 6, p. 12 15.
- Image sensor module FTM-12.- Preliminary specification, April 1993.
- Patent 2199191 RF. MKI7 HO4N 3/14, 5/335. Television camera/V.M. Smelkov//B.I. – 2003.- No. 5.