Television – Camera – system and detail – Optics
Reexamination Certificate
1998-06-23
2004-01-13
Garber, Wendy R. (Department: 2612)
Television
Camera, system and detail
Optics
C348S218100
Reexamination Certificate
active
06678000
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image pickup apparatus which offers high quality of images by using a technique referred to as “pixel shift”.
2. Description of the Related Art
In recent years, there is an increasing demand for image inputting apparatuses, due to rapid spreading of use of personal computers (PCs).
Typically, scanners are used as the image inputting apparatuses for use in combination with PCs. Such scanners are basically intended to pickup image information written or printed on sheets of paper or the like, such as written or printed texts, photographs and so on.
Scanners employ line sensors as imaging means, and cover a wide range of resolution from 200 dpi (dots per inch) up to very high resolution of 1200 dpi.
In general, sufficient legibility and readability of 10-point fonts of a written or printed text is considered as criteria for the image resolution in practical applications. To meet this requirement, the image inputting apparatuses have to pickup and form images at a resolution which is at least 200 dpi. Thus, the scanners satisfy this requirement insofar as the image resolution is concerned.
In the meantime, image inputting apparatuses relying upon video techniques are also available, a typical example of which is a visualizer which also is known as “document camera.” Many visualizers have been proposed and are actually available as commercial products.
In most cases, a visualizer has an illuminated platen for carrying an original document, and a video camera which is supported on the platen and which picks up the image of the original document.
Such a visualizer employs an area sensor as the imaging means. The image resolution obtainable with this type of apparatus is limited by the resolution achievable by the video camera. In general, a video camera can provide only low levels of image resolution. For instance, the image resolution is about 67 dpi at the highest, when the video camera is used for reading a full A-4 size text which is laid with its longer side extending from left to right. More specifically, the pixel pitch along the longer side of the A-4 size text is calculated as 640/9.5=67.4 (dpi) at the aspect ratio of 1:1 when the video camera follows NTSC specifications.
Referring again to scanners, one of the problems encountered with this type of image inputting apparatus is excessively long scanning time. In fact, it takes about 10 to 20 seconds to scan a color image. Another problem is that the whole image, i.e., the overall framing, cannot be grasped until the entire original image is scanned.
Apparatuses having a preview mode has been proposed, in order to obviate the second-mentioned problem, i.e., inability to grasp the overall framing before completion of scanning. Such apparatuses, however, are not easy to operate, due to necessity of certain steps, including previewing and subsequent pointing of the scanning area by a pointing device such as a mouse.
A fundamental problem encountered with the use of scanners is that the scanners cannot pickup images of three-dimensional objects. In other words, the scanners are effectively used only for reading texts and pictures.
In contrast, video visualizers are easy to use and can pickup images of three-dimensional objects while offering easy framing, thus obviating the problems encountered with the scanners, but are still unsatisfactory as they cannot satisfy requirements in regard to image resolution.
It is true that video visualizers can provide only low levels of resolution, e.g., 67 dpi at the highest, failing to meet the demand for higher image quality, as described before. Such video visualizers, on the other hand, offer a remarkable advantage, by virtue of use of a zooming function. More specifically, the user can control and determine the framing in real time by changing the zooming ratio and moving the original document on the platen while monitoring the image through a display.
It is therefore highly desirable that a video visualizer has a high level of resolution that compares well with that achieved by scanners. Such a video visualizer will make it possible to obtain high-quality images not only from texts or pictures but also from three-dimensional objects.
One of the measures for implementing such an image pickup apparatus would be to increase the number of pixels in the area sensor used in a video visualizer. The area sensors used in existing video visualizers have 410,000 pixels in total, among which 380,000 are effective, in order to satisfy the requirement for resolution corresponding to 450 TV lines. In the meantime, high-quality image sensors intended for use in digital still cameras have been proposed in recent years. These sensors have very large numbers of pixels, e.g., 1,000,000 pixels, 1,300,000 pixels or 1,600,000 pixels. It might be possible to substitute such a high-quality image sensor for a conventional area sensor in a video visualizer, in order to achieve the desired high level of resolution.
Such high-quality image sensors, however, have large chip sizes and, hence, are generally expensive. An attempt to enhance the scale of integration through reduction of pixel size is not considered promising, because such higher scale of integration poses other problems such as impairment of S/N ratio due to reduction in the sensitivity.
Moreover, it is to be pointed out that these high-quality image sensors are still unsatisfactory because they cannot provide image resolution equivalent to that offered by scanners. More specifically, the resolution is as low as 100 dpi when the area sensor having 1,000,000 pixels is used for reading a text of A-4 size. Similarly, the resolution is still as low as 135 dpi, when the area sensor having 1,350,000 pixels is used for reading A-4 size text and, even when an area sensor having 1,600,000 pixels is used, the resolution obtainable is still 135 dpi after conversion into image having the aspect ratio of 1:1. Such resolutions are much lower than 200 dpi easily achievable by a scanner having comparatively low resolution.
Furthermore, the video visualizer incorporating such a high-quality image sensor is required to produce analog video output in accordance with, for example, NTSC specifications, in order to attain compatibility with existing video visualizers. The analog video output is preferably of an image quality on the order of 410,000 in terms of pixel number (380,000 effective pixels). However, the above-mentioned high-quality image sensor, due to its specific pixel arrangement, lacks any output mode which would enable conversion of its output to the 410,000-pixel video rate. In order to implement such an output mode, it is necessary to additionally incorporate a scan rate converting circuit, with the result that the production cost is raised undesirably.
In addition, as is well known, the scan rate conversion itself is a factor which degrades the image quality, because it produces unnatural minute steps or jags at the edges of image. Therefore, even when the scan rate for the output of the high-quality image sensor could be converted to that corresponding to 410,000 pixels, the image quality obtained with such a high-quality image sensor is inevitably inferior to the quality of image derived from the conventional sensor having 410,000 pixels, due to the presence of minute steps or jags on the edges of images.
In order to obviate this problem, a method generally referred to as “pixel shift” has been proposed in which the position of the imaging device relative to the incident light is shifted stepwise at a pitch which is a fraction (1/N) of the pixel number, thereby improving the image resolution.
More specifically, the pixel shift can be realized by using a plane-parallel light transmissive glass plate disposed in the optical path at a location between the image pickup lens and the imaging device. The light-transmissive glass plate is rotated by a predetermined amount, so that the light rays carrying optical image information, which otherwise would impinge upon ins
Canon Kabushiki Kaisha
Fitzpatrick ,Cella, Harper & Scinto
Garber Wendy R.
Whipkey Jason
LandOfFree
High resolution still-image capture apparatus that shifts... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with High resolution still-image capture apparatus that shifts..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and High resolution still-image capture apparatus that shifts... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3259308