Other Free Encyclopedias » Online Encyclopedia » Encyclopedia - Featured Articles » Contributed Topics from F-J

Imaging Systems - Scanner Systems, Tri-linear arrays, Multi-shot Area Array Systems

image sensor filters color

ROMANO PADESTE
SINAR AG

Scanner Systems

Tri-linear arrays

Scanner systems were the first systems introduced to professional digital photography that were able to achieve the high number of pixels required for high-end applications. Similar to desktop scanners, these camera systems consist of a linear array sensor and a motor drive. During the capturing of an image, a linear array sensor is moved step-by-step across the entire image area of the camera, thus capturing the image line by line. The sensor is equipped with three lines of pixels, each of which is covered with a red, green, or blue filter, allowing the capture of a color image in a single pass.

The undisputed advantage of a scanning system is the ability to cover a very large image area at a low cost when building the imaging device. The image area covered by a typical scanning system would be comparable to the image area of a 4 × 5-inch film. This makes this technology particularly attractive for studios equipped with 4 × 5-inch view cameras. In this application, this type of sensor allows for use of the same camera for both analog and digital applications. As a consequence of the relatively large pixel pitch of these systems, no special lenses are required.

It is obvious that this technique may only be used for the photography of stationary subjects and is not suited for moving objects. Any movement of the subject or the camera during the capturing process has to be avoided or image recording will not be organized properly. Hence, sturdy studio stands are a must for this technology. Additional disadvantages of scanning systems include the relatively long times required for capturing an image. Special lights are required when using this technology—usually HMI lights that maintain a constant level of brightness during the capturing process. This is to avoid capturing different exposure during the scanning as the sensor travels across the scene. Regular, AC-powered tungsten or halogen light sources change their brightness several times per minute and would therefore cause brighter and darker lines across the image. The additional expense for these types of lights might be a factor and minimize the advantage of the lower cost for scanning camera systems when compared to large-format, area-array systems.

In the early years of digital photography in the high-end advertising studios, scanner systems were very popular and successful. But with the introduction of large-format, area-array devices their importance has decreased.

Multi-shot Area Array Systems

Three-shot/one monochrome area array sensor with RGB filter wheel

Early high-end, digital-color imaging systems were based on a monochrome area array sensor and a filter wheel equipped with red, green, and blue filters. The red, green, and blue channels of the image were captured in a sequence of three shots. This technology was used in the early days of analog photography as well. It may be used with any kind of light source, but it is obvious that color images may only be acquired with stationary subjects. Any kind of movement of the subject or the camera would result in a poor registration of the red, green, and blue channels. Consequently, moving objects can only be captured as grayscale images.

Thanks to the physical separation of the color filters and the imaging sensor, the filters in use can be chosen without major restrictions. Hence, an appropriate set of filters can produce excellent color rendition.

The color filter wheel can be arranged both in front and behind the lens. Both arrangements have inherent advantages and disadvantages. If the filters are in front of the lens, flare and non-image-forming light, as well as the physical dimensions of the filter wheel, is a major concern. If the filters are behind the lens, the requirements for the alignment of the filters and the risk that spots, scratches, and dust may be visible in the image are much higher.

Although this technology was very successful in the beginning of high-end digital imaging, it has more or less disappeared from the range of commercially available products.

Four-shot/one-area array sensor with Bayer pattern

Today’s commonly used area array sensors with red, green, and blue filters arranged in a Bayer pattern have opened up

the opportunity to introduce a new multi-shot technology to capture the full color information for every pixel in the imaging area. The basic idea is to capture multiple images by physically moving the imaging sensor by one pixel between the individual shots. Because of the arrangement of the red, green, and blue filters in the Bayer pattern, a total of four exposures is necessary to capture the three color channels for every pixel in the image area. The four exposures are then superimposed and combined into one image. Because there are twice as many green pixels on the sensor as red and blue pixels, the green channel is captured twice for every position in the image. The two green channel images are therefore averaged before they are combined with the red and blue channels.

This technology offers high-resolution, full-color images for still-live applications with any kind of light source. Thanks to the type of area array sensor used, the same camera may be used for moving objects in one-shot mode. Hence, these systems are very flexible in terms of application and have become quite popular in many fields of professional imaging.

The small pixel size of today’s area array sensors places a very high demand on the mechanical design of these devices, particularly on the part that shifts the sensor between the individual exposures. Most commonly, two piezo drives are used for the horizontal and vertical shifts of the sensor. The registration of the four exposures is very critical and any kind of poor registration will cause artifacts at the edges of the image structures. A very rugged camera and studio stand is therefore required for this type of application. Furthermore, the lighting used during the four exposures needs to be very consistent. Local variations in brightness will show up as artifacts in the final image.

Microscan Area Array Systems

Sixteen-shot/one-area array sensor with Bayer pattern

Microscanning technology takes the aforementioned four-shot technology one step further. It is a fact that not the entire surface of an area array sensor is sensitive to light. Photons that fall between the light-sensitive areas of two individual pixels are therefore simply lost. Consequently, the resolving power of a four-shot system is not as high as it is theoretically thought to be. Microscanning overcomes this problem by shifting the sensor in steps of half a pixel and thus also capturing the image information between the individual exposures of a four-shot image. A total of 16 individual exposures is required to capture the full red, green, and blue information for every position in the image area.

Microscanning offers excellent resolving power, provided that appropriate lenses are used. If the resolving power of the lens is not sufficient, microscanning just captures more pixels, not more image information. And it is obvious that movements of the camera or the object are even more critical than with four-shot technology. Furthermore, the time required to capture an image is considerably longer than with four-shot technology.

Nine-shot/one area-array sensor with Bayer pattern

Nine-shot technology is based on the same idea as sixteen-shot microscanning: to capture the image information that would fall between two individual pixels in a one- or four-shot exposure by moving the area array sensor to the positions in between. But instead of capturing all three channels at every position within the image area, the nine-shot captures just one channel per position. The missing two channels are then mathematically reconstructed from the neighboring pixels, just as with a one-shot exposure sensor. But the nine-shot sensor features a higher resolution image than a one- or four-shot and requires fewer exposures than a sixteen-shot. The drawback of this technology is that the time required for processing a nine-shot image is far greater than with any other type of multi-shot exposure, and there is still a risk of producing color artifacts. Hence the technology is not as widely used as four-shot or sixteen-shot microscanning systems.

Macroscan Area Array Systems

Multiple tiles with one-area array sensor

Another approach to increasing the resolving power of area array sensors is to capture multiple image tiles by shifting the entire sensor by a small distance, which can be less than its width and height. This shift is referred to as macroscanning. The multiple tiles, which slightly overlap each other, are then stitched together into one large image.

The major advantage of macroscanning is that higher resolving power is not achieved by decreasing the size of the pixels but by increasing the size of the image area. Furthermore, macroscanning can combine multiple one-shot or four-shot image tiles, a combination of both, or even sixteen-shot image tiles, resulting in very large, high-resolution images.

One of the disadvantages of macroscanning is that the camera becomes rather bulky. Also, it is critical that neither the camera nor the captured subject move during the exposure cycle. Otherwise, stitching the tiles properly together may not be possible or artifacts may occur in the overlapping areas of the image tiles.

Commercially, macroscanning was quite successful in the beginning of high-end digital imaging with area array sensors. But with the introduction of larger sensors with higher pixel counts, the importance of this technology has decreased.

Imaging Systems for Special Application

Multi-channel imaging systems

Besides high pixel count, resolution, and other image quality factors, the quality of color reproduction is an important issue in digital imaging. Since the color filters used are today most commonly applied directly onto the surface of the imaging sensor, the choice of dyes used for the filters is often limited. If the color filters are physically separated from the imaging sensor (such as on a filter wheel in front of the sensor or lens) the choice is considerably greater and color reproduction may therefore be better, if the appropriate set of filters is selected. To improve color reproduction even further, such as for the reproduction of important artwork, several multi-channel imaging systems have been suggested. These may either use a monochrome imaging sensor and a filter wheel with several narrow-band filters, or a conventional imaging sensor with RGB filters and an additional set of filters in front. All of these systems acquire multiple channels consecutively. Custom software is then required to convert the multi-channel image into a format that can be processed by standard imaging software. Although the results achieved with multi-channel imaging are quite promising, the systems are still in development and not yet commercially available.

IR imaging systems

Most of today’s commonly used imaging sensors are not only sensitive to the visible spectrum, but their sensitivity also expands into the near infrared, usually up to a wavelength of 1000 nm. For standard color imaging, the sensors are therefore equipped with an IR blocking filter, which absorbs wavelengths greater than 700nm. But for several special applications, such as forensic imaging, the sensitivity of the sensor in the near infrared can be used for IR imaging. In this case, the IR blocking filter is replaced with a clear glass that transmits near infrared. If only the infrared information is of interest, an IR transmission filter that absorbs the entire visible spectrum is used in front of the lens. If an imaging sensor fitted with red, green, and blue filters in front of the individual pixels is used, a combination of visible spectrum and IR imaging is possible as well. For example, the blue information may be blocked with a yellow filter and the “blue channel” then contains the IR information, whereas the green and red channels record green and red as usual. The channel chosen to record the IR information and consequently which filters are used depends, of course, on the spectral sensitivity of the individual channels, which may vary with the type of imaging sensor.

Imhotep - DIVINITY., OFFICES., CULT., SOURCES [next] [back] Image Watermarking Using Visual Cryptography

User Comments

Your email address will be altered so spam harvesting bots can't read it easily.
Hide my email completely instead?

Cancel or