WO1998033326A1

WO1998033326A1 - Digital colour camera for electronic photography

Info

Publication number: WO1998033326A1
Application number: PCT/EP1998/000425
Authority: WO
Inventors: Reimar Lenz
Original assignee: Reimar Lenz
Priority date: 1997-01-27
Filing date: 1998-01-26
Publication date: 1998-07-30
Also published as: JP2001522546A; EP0954928A1; AU6097898A; DE19702837C1

Abstract

Disclosed is a colour camera designed to make digital pictures, including moving objects, using one single CCD surface sensor with a colour mosaic filter, according to a method consisting in making two partial pictures in a quick sequence, for example by using a double flash and by modifying between both frames the spectral characteristic for the image points of the colour picture to be made, in such a way that, after the first partial picture, the CCD surface sensor is moved vertically by a screen distance. From the resulting colour separations for each image element information is obtained on the third missing colour separation, without the usual colour changes specific to a colour camera with a mosaic filter.

Description

Digital color camera for electronic photography

The invention relates to a digital color camera with an area sensor, in particular a CCD area sensor, consisting of light-sensitive sensor elements arranged in a matrix with a color mosaic filter corresponding to at least two color separations, and with an exposure control device that generates partial images in chronological order for recording a color image.

CCD area sensors are usually used in video cameras, but they can also be used to generate individual images. To produce color images that contain the complete color information of an object image, a color image converter with a set of three CCD area sensors with associated color filter masks is used, so that the individual area sensors each detect a color, for example the primary colors red (R), blue (B) and green (G).

For structural reasons, the light-sensitive sensor elements of an Interline Transfer CCD area sensor have an area size that is only half as large as the grid spacing, which means that there are light-insensitive areas between the individual sensor elements. Of course, some of the image information is lost as a result.

In order to lower the price of a color image converter, it has already been proposed to use only one such sensor instead of three CCD area sensors, a color mosaic filter for three color separations being located on the area sensor. Three adjacent sensor elements then form a color resolution element. The local resolution in such color image converters is of course relatively low compared to, for example, a monochrome image converter with a CCD area sensor. In contrast, the inventor has proposed a color image converter which is greatly improved in comparison and which only contains a CCD area sensor with a color mosaic filter with three color separations. The recording takes place in three steps, one step for a partial image, with the CCD area sensor being moved relative to the object image with the aid of piezoelectric actuators in a cycle of three steps such that each pixel is successively placed on one of three image sensor elements for the three Color separations fall. (EP-A-0 396 687).

This principle of staggered recording of partial images with subsequent combination of the partial information into a uniform color image has proven itself. However, this known color image converter is only suitable for recording stationary objects. Due to the design of the CCD area sensors, this has been regarded as inevitable; The step of reading the charges follows the charging time of 20 msec (television standard) or 100 ms (high-resolution sensors) in conventional CCD area sensors, during which photons are captured in each individual image sensor element. The reading of the image information from the sensor takes approximately as long as that which takes place in the preceding time period

Accumulation of the charges and practically cannot be accelerated. If you want to record three partial images with such a CCD area sensor, you need at least a time period that corresponds to approximately four time intervals for accumulating charges or reading out the charge information, namely three charge accumulation periods and - offset by one period - three readouts -Times. In addition, there is the time for (mechanical) displacement of the CCD area sensor relative to the object image. This shifting of the sensor must take place between the individual partial image recordings.

In the case of stationary objects, the relatively long time does not matter for the processes described above. With moving objects, however, essential image information is lost between the recordings of the individual partial images. Looking at a single pixel from which If three shots are taken in succession according to the three color separations, an imaginary point of a moving object is no longer exactly at the point at which it was when the first sub-frame was captured, and when the second sub-frame was captured Finally, in the third sub-picture, the object point is most likely in a completely different place. If the information from the three color separations is then put together, information from different object points is put together, which is noticeable by undesirable color disturbances.

Of course, a color image converter with three separate CCD area sensors can be provided for photographing moving objects ("snapshot"), but such a color image converter would hardly be commercially usable due to its very high price.

The invention is based on the object of specifying a digital color camera which on the one hand only requires a single CCD area sensor and on the other hand also enables the recording of moving objects in good quality.

This object is achieved according to the invention in that two partial images are taken at a time interval which is shorter than the second time period which is required for reading out a partial image from the sensor, and that the spectral characteristic at least between the two images is taken for taking a color image is switchable for some of the pixels of the color image to be generated.

In principle, the latter feature is already known from the aforementioned EP-A-0 396 687: from one and the same pixel, different color separations are obtained in succession by differently colored sensor elements. The spectral characteristic can be changed in the manner explained in more detail below in various ways, for example by moving the CCD area sensor towards the object image in a certain direction and by a certain amount.

The further feature according to the invention relates to the time interval between two successive recordings for the individual partial images. First of all, the time interval between the recordings of the two partial images must be small enough so that a moving object and, accordingly, the associated object image cannot shift significantly with respect to the CCD area sensor. Even with relatively quickly moving objects, a sharp image with essentially correct color information should still be possible.

The time span between two consecutive partial recordings cannot be shortened extremely; because for both partial exposures the individual sensor elements need a certain amount of time to

Being able to capture photons in an amount that represents significant information. Furthermore, a time gap is required between the two partial recordings, in which the spectral characteristic for the individual pixels is changed.

Experiments with embodiments of the invention have shown that sharp color images can be produced with the color image converter according to the invention, which are practically free of color alias disturbances. While it was previously assumed that three color separations are indispensable for such high-quality color images, the inventor has recognized that it is not absolutely necessary to have the information from three extracts available at each image location. If two color separations are available per picture element and the missing third color separation in the immediate vicinity, the value for the third color separation on the respective one can be calculated by simple calculation

Determine image element. For example, consider three adjacent pixels with the indices "1", "2" and "3", on each of which the color information for green (Gl, G2, G3) and partially for blue and Red (Bl, R2, B3) is present, for example the value of blue for the middle point can be calculated using the formula:

B2 = G2 x (Bl + B3) Gl + G3

The color separation that is still missing for each pixel can be obtained by arithmetic or in another way.

In addition to the realization that there are already two color separations for the individual

Sufficient pixels to produce a high-quality color image, it is also important to recognize that even with moving objects, a "simultaneous" recording for all pixels does not necessarily have to take place, but that it is sufficient if the partial images in succession in a relatively short time interval be included.

The taking of pictures in succession at a short time interval is already known per se, namely for displaying a sequence of movements. While it is specifically a question here of determining different movement phases by means of two images taken in quick succession, on the contrary, the object of the invention is to generate an image despite the time interval between two images, which apparently corresponds to only a single point in time.

According to the invention, the mode of operation of the usual CCD area sensors plays a considerable role. As mentioned, two time intervals are necessary to record an image with such CCD area sensors, the first time interval is used for the accumulation of charges in the respective image sensor element, the second interval is used for reading out these charges. However, the charge transfer from the sensor elements to the vertical is carried out prior to the reading

Bucket chains e.g. B. an interline transfer CDD sensor in less than 1 μs, so that the photo element is then ready to collect new charges. While the "second" partial image is then recorded, the layers belonging to the "first" partial image are made from these vertical bucket chains in a manner known per se read out. The complete time for reading out an image in the case of such CCD area sensors specifies the average frame rate in the case of television picture sequences, so that this average frame rate cannot be increased beyond the limit given by the reading out of the picture. According to the invention, however, the time interval between two partial recordings is shortened. This is only possible because there is a restriction to only two fields. The first time span in which charges are accumulated in the individual sensor elements is halved, for example. After the shortened charge accumulation period has ended, the exposure of the sensor elements is prevented, the charge is briefly transferred from the photo element of the respective image sensor element into the associated charge bucket chain, and reading out of the "first" partial image is then started. After the spectral characteristic for the pixel concerned has been changed, the recording of the "second" partial image is started. With the aid of this method of operation, which is completely atypical for a conventional CCD area sensor, the possibility is created according to the invention of generating practically completely free color images of moving objects with only one sensor of color aliases.

The above considerations relate primarily to so-called progressive scan interline transfer sensors, in which there is space in the bucket chains for the complete picture, and which are therefore ideal for a digital camera. However, the invention can also be implemented with so-called television image interline transfer and frame interline sensors. With these, only half of the sensor elements can be evaluated, but they are interesting because of their low price for cameras according to the invention. The frame interline sensor has a storage zone next to the image field and can deliver three independent images in a very short period of time. In addition to the two partial images recorded in the manner explained above, a third image can also be recorded so that all three color separations can be obtained quasi-simultaneously for each image location. Color alias artifacts can thus be completely avoided. In a special embodiment, as already mentioned, the change in the spectral characteristic between the two partial images can take place by means of a relative shift between the object image and the area sensor. If one considers a single pixel of the color image to be generated, then this pixel becomes the first pixel

Example taken with a sensor element sensitive to red, and in the second field with a sensor element sensitive to green, for example. This pixel is thus detected with two different spectral characteristics. This also applies to the other pixels.

The color mosaic filter is designed so that it corresponds to three color separations. In the case of such a color mosaic filter, for example, every second sensor element of the matrix of the area sensor can be sensitive to green, corresponding to the bright fields of a chessboard, the remaining half of the sensor elements being alternately sensitive to red and blue.

The relative shift can correspond to an integer multiple of a pixel spacing, in horizontal or in vertical

Direction. It is most favorable to position the area sensor vertically between the acquisition of the first and the second partial image, i.e. to move in the y-direction, because then the fact is exploited that in this direction the expansion of the light-sensitive surface of the sensor elements is greater than in the x-direction, so positioning errors have less effect.

Alternatively, the spectral characteristic between the two partial images can also be changed by a relative shift in the diagonal direction, v / o each by half a pixel in X-

Direction and is shifted by half a pixel in the Y direction. In this case, a color mosaic filter is preferably used for three color separations. This displacement of the area sensor will second partial image recorded at those image locations that lie in the spaces between the image locations of the first partial image.

In a conventional interline transfer CCD area sensor, in which the dimensions of the effective light entry area of a sensor element correspond to approximately half the grid spacing between adjacent sensor elements, only approximately a quarter of the total image information is recorded when a partial image is recorded. By shifting diagonally, another quarter of the image information is recorded (double resolution). However, only color information corresponding to a single color separation is then available for each recorded pixel. Because of the less information for the color, this embodiment is less favorable than the embodiment described above with a shift by full grid spacings.

This is also a change in the spectral characteristic. The resulting image can be imagined as four times the number of pixels, i.e. twice the number in the x and y directions. Some pixels that were not recorded at all in the first field (spectral zero characteristic) are now recorded in the second field with a colored sensor element, while on the other hand all pixels that were recorded in the first field with a colored sensor element are recorded in the second field at all cannot be recorded (with spectral zero characteristic). Half of all pixels of the imaginary resulting image four times the number of pixels is not captured by the first or the second partial image. The change between non-detection (zero characteristic) and detection (with color filter) of a pixel also means a change in the spectral characteristic in the sense of the invention.

The area sensor can be displaced vertically and / or horizontally in the image plane, for which use is preferably made of piezo actuators. However, magnetostrictive, electrodynamic or other actuators can also be used. The relative displacement between the area sensor and the object image can also be carried out by optical measures, for example by pivoting a prism located in the beam path, with the aid of a Kerr cell or by changing the polarization in cooperation with birefringent materials.

Changing the spectral characteristic between the recordings of the two partial images can also be done by changing the effective spectral sensitivity profiles of the light-sensitive sensor elements

With the help of a color filter, which can be selectively inserted into the beam path or which is permanently in the beam path and offers the possibility of changing its permeability. For example, a liquid crystal component can be provided as a color filter permanently in the beam path. Another possibility for changing the spectral characteristic is to vary the spectral sensitivity curve of the photodiodes of the sensor by applying variable electrical voltages to the CCD area sensor. There is then no need for a relative shift between the object image and the area sensor.

The operation of the sensor elements, that is, the accumulation of charges in the individual sensor elements, is prevented between the recordings of the two partial images. This can be done in a number of ways, as explained below. One way to achieve this is to keep the object dark in the transition phase when changing the spectral characteristic, thus preventing light from falling on the area sensor. For this purpose, for example, the object can be illuminated with a flash for each partial image recording. In order to achieve a change in the spectral characteristic at the same time, it is possible to photograph with a colorless flash when recording the first field and, for example, with a green flash when recording the second field. However, two differently colored flashes only make sense if the color sensor has a color mosaic filter with only two different colors and thus the spectral characteristic is to be changed globally in order to obtain the minimum required third color separation.

The two successive flashes mentioned above can be generated by a single flash device or by means of a device with two flash lamps. The purpose of these measures is to prevent blurring of the image (motion blurring, spectral blurring). For this purpose one can also think of the charge generated during the switching of the spectral characteristic in the

Electrically delete image sensor elements.

An optical / mechanical shutter can be used to shorten the time interval between the recording of the two partial images. It can be used to precisely set the effective exposure times, which by definition are the middle of the respective first time periods in which charges are normally accumulated in the individual sensor elements. These effective exposure times can also be achieved by "active lighting", that is to say the object to be photographed can be photographed with correspondingly timed flashes.

As explained above, partial images are generated for producing high-quality color images in two short successive time periods, with those taking place between the recordings of the partial images

Switching of the spectral characteristics for at least some of the pixels partial color information can be obtained. In order to calculate further color information from this partial information, the analog image data coming from the CCD sensor is digitized with the aid of an analog / digital converter, so that it can be processed in a computing unit. The color information relating to the individual partial images can be processed within the camera receiving the color image converter according to the invention, but the image information can also be processed as it is after the analog / digital conversion. Buffer to process later.

The invention offers a number of special advantages, which are, as it were, by-products of the measures according to the invention: if the CCD sensor is moved between the recordings of the partial images with the aid of a piezoelectric actuator, there is also the possibility, by wobbling the sensor, of information for automatic Calibration and gaining focus. By comparing the two partial images, information can be obtained as to whether the object is possibly moving too quickly in order to obtain a sharp image. If necessary, a warning signal can be generated for the user.

In the following, exemplary embodiments of the invention are described with reference to the

Drawing explained in more detail. Show it:

1 shows a block diagram of an embodiment of a digital color camera according to the invention;

2 shows a schematic illustration of a horizontally and vertically displaceably mounted CCD area sensor;

3 shows a detail of the surface of the sensor according to FIG. 2;

FIG. 4 shows a detail similar to FIG. 3, a direction of displacement of the CCD area sensor being indicated;

5 shows a view similar to FIG. 4, but after the displacement has taken place vertically downward, the one obtained from both partial images

Color information is shown; 6 shows a similar section of the surface of a CCD sensor with a displacement vector according to a second embodiment of the invention;

Fig. 7 shows the detail of the CCD sensor of FIG. 6 after it has taken place

Shift, wherein all the pixels captured by the two partial images are shown;

8 shows a pulse diagram which illustrates the mode of operation of the color camera according to the invention;

9 and 10 are schematic representations of a progressive scan interline transfer CCD area sensor (FIG. 9) and a progressive scan frame interline transfer CCD area sensor (FIG. 10);

11 shows a detail from the surface of a CCD area sensor for a third embodiment of the invention; and

Fig. 12 shows the effective light sensitivity of the sensor elements of a sensor with a color filter pattern according to Fig. 11, if the spectral sensitivity characteristic e.g. by inserting a filter that is only permeable to green.

Fig. 1 shows schematically a digital color camera, which is constructed similarly to a conventional camera with which, for example, 35 mm films are exposed. Instead of the single image plane of a conventional film, a CCD area sensor 1 is located at the relevant point in the color camera according to the invention, which is mounted on a holder 2 with the aid of piezo actuators 3. An object image of an object 78 is projected onto the CCD area sensor with the aid of an objective 16. In the beam path, ie in front of, in or behind the lens 16, there is a shutter 17 and an aperture 76. Depending on the specific embodiment, the CCD area sensor is a progressive scan interline transfer CCD area sensor or, alternatively, a normal one Interline Transfer CCD area sensor for the TV field process, a frame interline transfer CCD area sensor for the TV field process with additional memory zone, a progressive scan frame interline transfer CCD area sensor with additional Storage zone and full screen method or a frame transfer CCD area sensor. A progressive scan interline transfer CCD area sensor for full-screen operation is preferred, since in sensors with field operation the achievable vertical resolution in the camera according to the invention drops to half. Nevertheless, despite the disadvantage mentioned, the use of such sensors is quite interesting, since they are available in large quantities for

Video cameras are produced and are therefore very cheap.

The operation of the camera is controlled by a central processing unit (CPU) 5, the control signals to a driver 7 for the piezo actuators 3, to a CCD driver 8, a shutter control 9 and a flash control

10 there. The analog image signals coming from the CCD area sensor 1 are converted by an analog / digital converter 6 and input into the CPU 5 for storage and / or image processing. The central unit 5 can be connected to an external computer 13 or to an external image memory 15 via interfaces 12 and 14. The

The housing of the camera is indicated by block 77.

A flash unit 18 with a reflector and one or two flash lamps 19 can be attached to the camera housing 77, the method of operation of which is explained below.

As shown in FIG. 2, the CCD area sensor 1 can be moved in the horizontal direction (X direction) and vertical direction (Y direction) with respect to the outer frame of the holder 2 due to the piezoelectric actuators 3. The one to move the

Variable voltages to be applied to the CCD sensor 1 are applied by the central unit 5 via the piezo driver 7. 3 shows a section of the surface of the CCD sensor 1. The individual sensor elements 20 and the color mosaic filter 21 for the individual sensor elements can be seen. G means that the sensor element in question is sensitive to the color green (G). According to FIG. 3, every second sensor element 20 is sensitive to green (G) in the vertical as well as in the horizontal direction. In every second row and every second column, every second sensor element is sensitive to blue (B) and red (R).

FIG. 4 shows the same arrangement as FIG. 3, but with some additional information in FIG. 4. For example, a vector V _v indicates the direction in which the CCD sensor 1 is shifted after the recording of a first partial image (FIG. 3) in order to record a second partial image in the second position.

4, the individual image sensors 20 have a grid spacing d _H and d _v in the horizontal and vertical directions. The size of the individual sensor elements 20 corresponds to approximately half the grid spacing. Between the recordings of two partial images, the CCD sensor is moved in the direction of arrow V _v by a grid spacing, so that an imaginary point A of the sensor is located at point B when the second partial image is recorded.

3, the four sensor elements in the rightmost column receive color information for the colors B, G, B and G of the relevant pixels, the sensor elements in the second column from the right receive information regarding the colors G. , R, G or R for the pixels corresponding to these sensor elements, etc.

On the right in FIG. 4, three superimposed pixels Pyi, Pyi + 1 and Pyi + 2 are shown. The area covered by the section of the CCD sensor 1 in both partial images is indicated by a "object image" I in dashed lines. After recording the first partial image according to FIG. 3, by applying appropriate voltages to the piezoelectric actuators 3, the CCD sensor is displaced by the distance V _v according to a vertical grid spacing d _v before the second partial image is taken. The three pixels Pyi to Pyi + 2 are again shown on the right in FIG. It can be seen that after the acquisition of the second partial image, two color separations are available for these three pixels or image locations Pyi to Pyi +2, namely G _A , B ^, G _A from the first partial image and B _B , G _B , B _B from the second field (for the bottom line of

Pixels are only the color separations obtained in the second field, only the color information obtained in the first field is available for the top line of pixels, which is why these lines are shown in parentheses in FIG. 5).

In the second column from the right in FIG. 5 it can be seen that the color separations G and R are available for each pixel. Using relatively simple calculation rules, the third color separation for each pixel can be calculated from all of this color information, for example using the formula given at the beginning.

After the calculation of the third color separation in each case, color images can be generated, for example printed out, from the total image information then available. For this purpose, the camera shown as a block in FIG. 1 is connected to a memory or an external computer in order to transmit the data of the color images for further processing. The information for the third color separation in each case can be calculated internally in the camera itself, but alternatively also externally in the computer 13 shown at the bottom right in FIG. 1.

6 and 7 show a second embodiment of the invention. In this embodiment, two color separations are not recorded for each pixel, but it is used for an emerging image With double resolution in the x and y directions, only one color separation was recorded, namely for every second pixel of the resulting image. 6 shows the direction of displacement and the amount of displacement V _D for the displacement of the CCD sensor 1 between the recording of the first and the second partial image.

As can be seen in FIG. 7, after the second partial image has been recorded, there is information for a number of pixels which is twice as large as the number of sensor elements of the CCD sensor 1. Due to the diagonal displacement by half a sensor element grid spacing, the effective resolution is doubled in both axis directions. However, only 1/3 of the complete color information, namely only one color separation, is available for each captured pixel. However, since half of the pixels of the resulting double-resolution image are not recorded at all, only 1/6 of the complete color information is available on average

Available. This embodiment is therefore somewhat less favorable in terms of color rendering quality than the embodiment which was explained above in connection with FIGS. 4 to 5, since the embodiment according to FIGS. 6 and 7 is more susceptible to color alias interference. For templates which are known to be achromatic, for example black text on a white background, the second embodiment of the invention can be used very advantageously with this knowledge.

The second can be used for the black and white recording of colorful templates

Embodiment in connection with a black / white sensor, i.e. a sensor without a color mosaic filter can be used for practically artifact-free doubling of the resolution of the sensor, whereby moving objects can also be recorded here.

FIG. 8 shows in the form of a pulse diagram the chronological sequence of the sequence on various parts of the color image converter according to FIG. 1.

8a) shows the horizontal time axis in units of milliseconds. The periodic normal operation of a progressive scan interline transfer CCD sensor 1 is shown schematically in FIG. 8b). The electrons generated by capturing photons are accumulated within an exposure period 35, the duration of which is T. On the

The ordinate in FIG. 8 b) is the charge of a single sensor element of the CCD sensor 1, which increases linearly with time, with constant incidence of light. The center of each exposure time period 35 is designated 41 and is referred to here as the “effective exposure time”. Two adjacent "effective exposure times" 41 have a time interval T, from one another.

At the end of the exposure time, the charge packet accumulated up to that point is transferred from the sensor element under consideration into the associated vertical bucket chain with a transmission pulse 43 or 75.

After this transfer, the next exposure process can begin. 8b) shows four successive exposure processes 37, 38, 39 and 40.

After the charge packet has been transferred from the sensor element in question into the associated bucket chain on the basis of pulse 43 or 75, the reading takes place, FIG. 8d) shows the reading intervals 32 for the respective preceding exposure process 36, 37, 38 or 39 Exposure with the duration T is also referred to here as the “first time period”, the time period for the readout corresponding to the readout interval 32 is also referred to here as the “second time period”.

8e) shows the light transmission curve of the closure 17 (see FIG. 1). At a time 46, the shutter is opened and a time interval T ₂ begins, at the end of which the exposure for a first field is ended due to the transmission pulse 75 (FIG. 8c). The middle of the time interval T ₂ is designated 48 and, similarly to the times 41 in FIG. 8b), means the "effective exposure time" for the first partial image. After the end of the first time period T ₂ , that is to say after the exposure process for the first partial image has ended, the CCD area sensor 1 according to FIGS. 4 and 5 is moved vertically downwards by a grid spacing, so that the individual sensor elements are then moved down by one pixel are.

8h) shows the accumulated charge increasing in time in the individual image sensor element considered here. The increase in the charge naturally only begins after light can fall on the sensor when the closure is opened, and ends when the closure is closed. After the charge has reached a certain value at 53, the transfer from the sensor element into the associated bucket chain takes place, and charges 60 that are generated in the sensor element by light incidence are deleted during this pulse time of the pulses 60 by means of erase pulses 60 shown in FIG. 8g).

As can be seen in FIG. 8i), the reading interval 32 (53) begins shortly after the exposure shown at 53 has ended. The duration of this process cannot be shortened and corresponds to the "normal operation" of the CCD sensor.

After the erase pulses, the exposure process begins at 54 for the second field after the shifted CCD sensor has arrived in its second position according to FIG. 8f). This is followed by a time period T ₂ with the "effective exposure time" 48. The time interval 49 between the two effective exposure times 48 is smaller than the normal time interval T, between two effective exposure times in FIG. 8b).

After the second period T _{2 has ended} , the closure according to FIG.

8e) closed, so that no further charges accumulate, which is represented by the horizontal end section of the charge profile 54 in FIG. 8h). After the reading out of the first partial image, the corresponding charge packet 54 is inserted into the associated bucket chain Transfer sensor element, then the information is read out in the readout interval 32 (54) in FIG. 8i). After the interval 32 (54) has ended, the entire information of the two partial images is available. The analog image information coming from the CCD sensor 1 is converted into digital values by the analog / digital converter 6 and sent to the central unit 5. The image information is then further processed in this central unit 5 or in an external computer 13 in order to calculate the information which is still missing for the third color separation.

The procedure described above can take place in the same way if use is made of the second embodiment according to FIGS. 6 to 7, except that in this case a somewhat modified calculation of the missing color separation information is carried out.

8j) schematically shows the light output 61 represented by a flash lamp 19. If the course of the curve in FIG. 8j) is known in principle, the starting time for the flash can be set such that one for each of the first and the second partial image the same amount of light is available.

8k) shows the amount of charge in the image sensor element under consideration when using the flashing light according to FIG. 8j).

Fig. 81) shows the light output for a double flash, which instead of

Flash with the light output according to Fig. 8j) can be used.

8m) shows the charge accumulation (signal curves 53 and 54) for the two partial images corresponding to the double flashes 63 in FIG. 81). Fig. 8n) shows erase pulse bursts 66 with interposed pulse pauses 47, after which a continuous read operation of the CCD sensor 1 takes place with the shutter 17 open, so that the user can adjust the device (focus etc.), the duration of the pulse -Bursts 66 like that is selected so that the effective exposure time for the image sensor element in question corresponds approximately to the time period T ₂ in FIG. 8e).

Fig. 9 shows schematically a progressive scan interline transfer CCD area sensor 1 in black and white version with those already mentioned

Sensor elements 20, the vertical CCD bucket chains 70 in three-phase design, which form a storage zone and allow the progressive scan operating mode (in contrast to the interlacing method which is customary in television technology). A horizontal CCD bucket chain 71 is connected to the vertical bucket chains and to this an output amplifier 72.

The CCD sensor cutouts according to FIGS. 3 to 7 correspond to the illustration according to FIG. 9, but in FIGS. 3 to 7 the CCD bucket chains 70 and 71 and the output amplifier 72 are omitted.

10 schematically shows a progressive scan frame interline transfer CCD area sensor 1 in black and white version with an additional, optically covered memory zone 73. Such a sensor enables the rapid successive recording of three partial images and thus the complete avoidance of color aliasing; because the

Transfer from the vertical bucket chains 70 into the storage zone 73 can also be carried out quickly enough with 1/1000 sec for the recording of moving objects, as can the transfer of charges from the light-sensitive elements 20 into the vertical bucket chains 70.

In the case of a color image converter with the CCD sensor shown in FIG. 10, the spectral characteristic, based on a pixel of the resulting image, would have to be changed twice, that is to say between the first and the second and the third partial image. This could be done by moving the CCD sensor twice. Of course, use is still made of the features according to the invention of switching the spectral characteristic between two successive partial images, and of the feature that two fields are recorded in a shortened time interval with respect to the "second period".

11 schematically shows the detail of a CCD image sensor with a color mosaic filter 74, which makes the individual sensor elements 20 sensitive to yellow (Y) and cyan (C). The CCD sensor 1 remains in its place during the recording of the first and the second partial image. In the present case, a filter is placed mechanically or optically (for example with the aid of a liquid crystal component) between the recordings for the two partial images

Case a green filter. In the first partial image, the sensor elements sensitive to cyan (C) then take up the information corresponding to the sum of the colors blue and green, the other sensor elements for yellow (Y) take up color information for green plus red. When the second partial image is recorded, the green filter in the beam path only allows the respective green component to pass through to the sensor elements, so that both the sensor elements sensitive to cyan (C) and yellow (Y) only have the color information for Received green as shown in Fig. 12. In this embodiment, two color separations are therefore also available for each sensor element, from which the information for the third color separation can be calculated. In this embodiment, too, the spectral characteristic for the pixels is switched between the recordings for two partial images recorded in a short sequence.

Claims

claims

1. digital color camera with an area sensor (1), in particular a CCD area sensor, consisting of light-sensitive sensor elements (20) arranged in a matrix with a color mosaic filter (21) corresponding to at least two color separations, and with an exposure control device (5, 8, 9, 10), which generates partial images for recording a color image in chronological order, characterized in that two partial images are recorded at a time interval which is shorter than the time required for reading out a partial image from the sensor, and in that between the two

Field recordings the spectral characteristic can be switched over at least for some of the pixels of the color image to be generated.

2. Camera according to claim 1, characterized in that the color mosaic filter has at least three color separations and the change in the spectral characteristic between the two partial images by relative displacement between the object image and the area sensor (1).

3. Camera according to claim 1 or 2, characterized in that the

Relative displacement in the horizontal direction by integer multiples of the pixel spacing (d _H ) or in the vertical direction by integer multiples of the pixel spacing (d _v ).

4. Camera according to claim 1 or 2, characterized in that the

Color mosaic filter preferably corresponds to three color separations and the change in the spectral characteristic between the two partial images takes place by a relative shift between the object image and the area sensor (1) by half a pixel in the x direction and by half a pixel in the perpendicular y direction, whereby the x and the y direction can be specified by the matrix of the sensor elements (20).

5. Camera according to claim 1 to 4, characterized in that the area sensor (1) is arranged displaceably in the image plane.

6. Camera according to claim 1, characterized in that the change in the spectral characteristic between the two sub-images takes place by changing the effective spectral sensitivity curves of the light-sensitive sensor elements.

7. Camera according to claim 6, characterized in that the change in the effective spectral sensitivity curves takes place by means of a color filter which is introduced into the beam path or removed from it or replaced.

8. Camera according to claim 6, characterized in that the change in the effective spectral sensitivity curves takes place by changing the spectral transmission curve of a color filter that is permanently in the beam path.

9. Camera according to claim 8, characterized in that it is in the permanent in the beam path, variable color filter is a liquid crystal device.

10. Camera according to claim 6, characterized in that the change in the spectral sensitivity curves of the two partial images takes place by means of variable electrical voltage which are applied to the CCD sensor and change the spectral sensitivity curve of the photodiodes of the CCD sensor.

11. Camera according to claim 6, characterized in that the change in the effective spectral sensitivity curves takes place by changing the illumination of the object, for example by a colored loose flash for the first drawing and a green flash for the second drawing.

12. Camera according to one of claims 1 to 11, characterized in that the CCD area sensor (1) is an interline

Transfer CCD sensor.

13. Camera according to claim 12, characterized in that the interline transfer CCD sensor is a progressive scan sensor, i.e. is a sensor that does not interlace.

14. Camera according to one of claims 1 to 13, characterized in that the light incident during the change in the spectral characteristic does not contribute to either of the two partial images.

15. A camera according to claim 14, characterized by electrically erasing the charge generated in the area sensor during the change time.

16. Camera according to claim 14, characterized by preventing the incidence of light on the area sensor during the change in the spectral characteristic.

17. A camera according to claim 14, characterized by interrupting the beam path between the object and the XCD area sensor (1) while changing the spectral characteristic.

18. Camera according to one of claims 14 to 17, characterized by the active illumination of the object with a double flash, the spectral characteristic changing between the two flashes.

19. Camera according to claim 1, characterized in that the shorter time interval between the effective exposure times, that is, the centers (41) of the respective first time period at which the two partial images are recorded, is generated by using an optical / mechanical shutter.

20. Camera according to claim 1, characterized in that the shorter time interval between the effective exposure times, that is the centers (41) of the respective first time period at which the two partial images are recorded, is generated more actively by using them

Lighting.

21. Camera according to claim 20, characterized in that the active lighting takes place by means of a flash.

22. Camera according to claim 20, characterized in that the active lighting takes place by means of a double flash.

23. Camera according to claim 22, characterized in that the double flash is generated by using two successively fired flash tubes (19).

24. Camera according to claim 22, characterized in that the double flash is generated by firing one and the same flash tube.

25. Camera according to claim 1, characterized in that an analog / digital converter digitizes the analog image data of the CCD sensor.

26. Camera according to claim 25, characterized in that the digital image data are fed to a storage and computing unit (5) which generates the resulting color image.

27. Camera according to one of claims 1 to 26, characterized in that the resulting color images are stored in the color image converter system, e.g. in a replaceable PCMCIA card.

28. Camera according to one of claims 1 to 27, characterized in that the resulting color images are transmitted directly to a computer (13).

29. Camera according to one of claims 1 to 28, characterized in that the image resolution is further increased by the inclusion of more than two partial images which are shifted relative to one another.

30. Camera according to claim 1, characterized in that for moving the surface sensor (1) piezo actuators (3) are provided.