Television – Camera – system and detail – Unitary image formed by compiling sub-areas of same scene
Reexamination Certificate
1999-02-25
2004-06-15
Christensen, Andrew (Department: 2615)
Television
Camera, system and detail
Unitary image formed by compiling sub-areas of same scene
C348S240200
Reexamination Certificate
active
06750903
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to a digital camera that processes serial images to obtain a wide angle still image with high resolution. In particular, a super high resolution digital camera produces a wide angle still image with high resolution from images input by a zooming operation.
BACKGROUND OF THE INVENTION
For photography and related uses, many types of digital cameras have been sold in recent years. For achieving the objects of portability, functionality, or picture quality, various technical features have been adopted. Since the photographed image is digital, the image can be directly input into a computer, transmitted between computers and published, for example on the homepage of an Internet site or with a printer. Therefore, the demand for digital cameras has been growing quickly. However, the picture quality of present digital cameras is greatly inferior to that of conventional photographic film (silver salt) cameras. The resolution (typically 320 pixels 240 pixels) of the present average digital camera limits the ability of a user to enlarge the angle of view of the captured image.
Therefore, a need exists for a still image digital camera which provides a high resolution image even when a wide angle photograph is sought.
In salient Video Stills: Content and Context Preserved, ACM Multimedia 93 pp.39-46 written by Laure Tedosio et al, a method for obtaining an image with high resolution by using several conventional images is disclosed. This camera uses an affine transformation as a model for the camera's operation, and calculates 6 parameters from a group of local motion vectors between continuous images and synthesizes one wide angle image with high resolution from the plural images.
On the other hand, in the field of coding images, “A Study of alignment for sprite producing” (IE97-81(1997) pp.19-24) written by Hataizumi et al is known which discloses a method for generating a wide angle still image known as “sprite”. This method uses Helmert's transformation as a model of the camera's operation, and calculates a pan, a zoom and a rotation parameter of the camera, and synthesizes one wide angle image from the plural images using these parameters. Likewise with the aforementioned first method, each parameter is calculated from the group of local motion vectors between continuous images.
SUMMARY OF THE INVENTION
Generating a digital still image with high resolution in a prior art digital camera requires the calculation of local motion vectors between (continuous or serial) images. In order to calculate a motion vector, a search process is executed by block matching of blocks of a size of 16×16 pixels. As a result, the block matching needs to be calculated for hundreds of blocks in the image. Therefore, there is a problem with increasing the image resolution because of the corresponding requirement of greatly increasing the quantity of calculation.
Furthermore, since the camera operation related parameters are estimated based on local motion vectors, for the case where a moving object exists in the image, there is a problem that the moving image causes an error in the resolution of the image. Therefore, it is difficult to generate an image with high resolution by applying conventional methods to a digital camera. Accordingly, in the prior art, a practical problem exists with the detection of camera operation related parameters in real time in order to make the detection performance quick and efficient.
It is an object of the present invention, therefore, to provide a digital camera which generates a wide angle still picture with high resolution similar to that obtained by conventional wide angle photography at a low cost, and in real time from a series of input images, preferably input by recording a scene with the camera.
The camera of the invention receives a series of input images, preferably by capturing the input images with an image sensing system for photographing an arbitrary scene, such as a moving or dynamic image. For recording a series of images, the camera has an operation input switch for starting and ending the recording of a scene, during which the camera's zoom function is operated to either zoom in for enlarging the image (correspondingly reducing the angle of view) of the scene being shot or to zoom out for reducing the image (correspondingly widening the angle of view) of the scene.
The camera also preferably has a display for displaying the high resolution image as it is being recorded and after processing. Preferably, the display is made a part of the camera body for a user to view the image during recording. Optionally, the image taken in by the view finder can be included as part of the display. Also, the display can show a graphical user interface that shows image information related to the zooming operation displayed at the same time on the display with the images being recorded.
As part of the processing of the series of images, a zoom rate detector detects a zoom rate between images, preferably images photographed by the image sensing system of the camera or otherwise input to the camera from storage, for example. The zoom rate detected between each of the images is related to a scaling factor that is used to enlarge or reduce sequential images in the series during the processing of the images which are then stored in an image buffer to obtain the super high resolution image.
For example, an initial image of a series of images is written to the image buffer. Then each succeeding next image is multiplied by the appropriate scaling factor determined in relation to the detected zoom rate and overwritten on the image buffer to provide for greater image detail of that part of the scene which is captured by each next image of the series. This process is repeated in real time during the recording operation of the camera, with each next image of the series (for each sequential image of the series or for every other image thereof, etc.), being scaled up or scaled down according to the zoom rate detected by the zoom rate detector and then being overwritten on the image buffer.
During processing, as the series of images is being recorded and overwritten on the image buffer, preferably the resultant image is displayed on the display in real time. After the recording of the series of images is concluded, or otherwise input to the camera, and all of the images to be processed have been appropriately scaled and overwritten on the image buffer, the resultant image in the image buffer is the super high resolution image. The super high resolution image can be displayed and stored as a typical digital image in the camera in an image memory and otherwise output for storage in an external memory or display.
REFERENCES:
patent: 5657402 (1997-08-01), Bender et al.
patent: 5889553 (1999-03-01), Kino et al.
patent: 5907353 (1999-05-01), Okauchi
patent: 06-141228 (1994-05-01), None
patent: 97/25690 (1997-07-01), None
Kim et al; An Efficient Global Motion Characterization Method For Image Processing Applications, IEEE Transactions on Consumer Electronics, Vol 43, No. 4, Nov. 1997, pp. 1010-1017.*
Teodosio et al, “Sallient Video Stills: Content and Context Preserved,” ACM Multimedia 93, pp. 39-46.
Hataizumi et al, “A Study of Alignment for Sprite Producing,” IE97-81 (1997), pp. 19-24. (Abstract).
Fujita Takehiro
Miyatake Takafumi
Nagasaka Akio
Christensen Andrew
Harris Tia M.
Hitachi , Ltd.
Mattingly Stanger & Malur, P.C.
LandOfFree
Super high resolution camera does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Super high resolution camera, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Super high resolution camera will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3362948