DD245047A1 - ARRANGEMENT AND METHOD FOR CORRECTING THE OUTPUT SIGNALS OF A SEMICONDUCTOR IMAGE SENSOR CAMERA - Google Patents

Abstract

The invention relates to the real-time-capable correction of the output signals of semiconductor image sensors, such. For example, the dark current and shading correction in the application of CCD matrix cameras. The correction is carried out so that the corrected data are available on-line, so that they z. B. are directly displayed on a monitor. This is possible if the output of the solid-state image sensor camera is connected to an ADC which has two inputs for variable reference voltages and which are each controlled by a DAC with the correction values for the dark current and shading.

Description

Field of application of the invention

The invention relates to a circuit arrangement and a method for real-time correction of the shading and dark current influence on the output signal of semiconductor image sensor cameras. As semiconductor image sensors, CCD line sensors, CCD array sensors and the like are used in this connection. designated.

Characteristic of known technical solutions

Various methods for correcting electrical quantities are known. Often it is necessary to realize a mathematical function. In DE-OS 2430252 a device is described which corrects a digitally present after an analog-to-digital conversion measured value by a successively performed multiplication and addition by means of digital arithmetic and memory. The disadvantage here is above all the temporally successive addition and multiplication and the use of digital arithmetic units. The computing time increases with such arithmetic units with the word width used. Fast digital arithmetic units are also very expensive.

A special field of application is the correction of the output signal of a solid state image sensor camera with respect to the shading and dark current influence. With good approximation, the output signal A (n) of the ηth pixel applies

A (n) = D (n) + E (n) S (n)

where D (n) is the signal component caused by the dark current, E (n) is the illumination and S (n) is the sensitivity of the respective ηth pixel. D and S are constant for each picture element under constant operating conditions of the solid state image sensor camera. The correction takes place after the single determination of D (n) and S (n):

A (n) -D (n) ".

S (n)

In principle, a correction would be possible by means of the method described in DE-OS 2430252. The problem, however, is the time required. For on-line correction, matrix image sensors typically use only 100-200 ns per pixel.

For this reason, even known methods which perform a correction of the measured values by means of a computer, not applicable (100-1 000 times too slow) .-

However, methods are known in which the analog measured value is corrected before analog-to-digital conversion by means of analog computing circuits, such as logarithmic circuits and adder circuits. However, this results in the known linearity and stability problems at higher resolutions.

Object of the invention

The aim of the invention is to perform the correction of the output signal of a semiconductor image sensor camera real-time capable and with little effort.

Essence of the invention

The invention is based on the object to carry out the correction of the output signal of a semiconductor image sensor camera so that the corrected image data on-line, d. H. are available with the same clock as the image signal, so that they z. B. can be displayed on a monitor. At the same time, the aim is to reduce the effort required to a minimum.

This is possible by the use of an analog-to-digital converter, which makes it possible to limit its measuring range by two reference voltages U _re n and U _re f ₂ . Almost all high-speed analog-to-digital converters currently used have this capability / The voltage U _{ref1 limits} the range on the negative side and the voltage U _ref2 on the more positive side. These voltages can be varied in certain ranges specified by the circuit manufacturer.

The invention assumes that _ref2 mathematical calculations can be carried out by the selective change of U and U _re fi. With a known range of the measuring signal, it is possible to subtract or add a voltage by simultaneously increasing or decreasing the voltages U _re fi and U _ref2 by the same amount from the measured value with respect to the output signal of the analogue-digital converter. By changing U _re n or U _re f ₂ one achieves a multiplication or division of the measuring signal.

This will be explained using the example of the output signal of a semiconductor image sensor camera. The output signal of a semiconductor image sensor camera should be able to lie between Uo and Ui. However, the unlit semiconductor image sensor camera does not deliver U ₀ , but a higher voltage U _d , which corresponds to the dark current. The signal component U _d is independent of the illumination E and limits the possible dynamic range Ui - U ₀ to Ui - (U ₀ + U _d ). After the single determination of U _d , however, it is possible to increase U _{refl of} the analog-to-digital converter from U ₀ to U _d and U _ref2 from U ₁ to U ₁ + U _d .

Thus, the output signal of the analog-to-digital converter would assume the value 0 for the input value U _d , ie U _{d was} subtracted from the input signal.

If the semiconductor image sensor camera is now illuminated homogeneously with an intensity that is slightly lower than the saturation illuminance, the semiconductor image sensor camera supplies the output signal U _s .

If you now set U _ref2 to U _s , the analog-to-digital converter supplies the value for 100%.

In other, any lower illumination intensities of the analog-to-digital converter providing a normalized to U _s and U _d value with the specified settings for _re fi U and U _ref2. The output signal would be corrected according to the specified relationship. The consideration can be analogously transferred to all pixels of the solid state image sensor camera.

The weighted correction values U _d (n) and U _s (n) can be z. B. be stored in a semiconductor memory and U _re fi and U _re f ₂ are controlled via two digital-to-analog converter. The delay time for the correction results from the memory access time, the settling time for a digital-to-analog converter and the conversion time of the analog-to-digital converter. With the z. For available devices, there is no problem in achieving total times that are less than the pixel frequency.

Regardless of the used word width of the analog-to-digital converter, the cost of the solution depends mainly on the analog-to-digital converter, which is required in any known solution.

It goes without saying that the method described can also be used for other, any measured manipulations in which the values for U _re n and U _{ref2 have} a different origin than in the described solution.

embodiment

The method according to the invention is described below using the example of a circuit for correcting the output signal of a semiconductor image sensor camera according to FIG.

The correction circuit consists of a semiconductor image sensor camera 1, an analog-to-digital converter 2, a pixel counter 3, two digital-analog converters 4 and 5, three semiconductor memories 6, 7 and 8 and a data multiplexer 9. Controlled analog converter so that the maximum possible output signal range of the semiconductor image sensor camera results as an input range at the analog-to-digital converter. The image sensor is not illuminated. Thus, the output signal contains only the dark signal component U _d (n) of each picture element n. The data multiplexer 9 is controlled such that the measured values present as digital words by the analog-to-digital converter enter the semiconductor memory 6. By the pixel counter 3, a different address is applied to the memories 6, 7 and 8 for each picture element.

After determining the correction values for the dark signal, the solid-state image sensor camera is illuminated homogeneously. The data multiplexer is controlled so that the data enters the memory block 7. This completes the correction value determination.

Under constant operating conditions of the semiconductor image sensor camera, it is also possible to store the correction values after a single determination in a read-only memory, so that the correction values are not always to be redetermined.

For normal operation, the data multiplexer is controlled so that the measured values can enter the semiconductor memory 8. The dark signal values U _d (n) are thereby applied to the digital-to-analog converter 4 for generating U _ref , the light signal values U _s (n) to the digital-to-analog converter 5 for generating U _re f ₂ . Thus, the image signals behind the analog-to-digital converter 2 in the semiconductor memory 8 are corrected as digital words at your disposal.

The semiconductor memory 8 may be e.g. be a dual-ported RAM, accessing from a microcomputer or similar. allowed or can be replaced by another digital-to-analog converter that makes the corrected image signal visible on a monitor.

Claims (4)

Invention claim: An arrangement for correcting the output signals of a solid state image sensor camera, characterized in that the output of the solid state image sensor camera is connected to an analogue to digital converter having two variable reference voltage inputs each connected to a digital to analogue converter are connected and that the semiconductor image sensor camera is further connected via a pixel counter with at least two correction value memories, at whose outputs the inputs of the digital-to-analog converter are connected. 2. Arrangement according to item 1, characterized in that the output of the analog-to-digital converter is connected via a data multiplexer to the input of a correction value memory. 3. A method for correcting the output signals of a semiconductor image sensor camera using the arrangement according to item!, Characterized in that is controlled by changing the reference voltages of the analog-to-digital converter whose conversion range in dependence on the pixel clock with determined by known methods correction values. 4. The method according to item 3, characterized in that the first reference voltage is set so that it corresponds to the Dunkeistromsignal of the respective picture element of the solid-state camera and the second reference voltage is adjusted so that it is approximately the saturation signal of the respective picture element of the solid-state image sensor camera equivalent. For this 1 page drawing