Image analysis – Image transformation or preprocessing – Combining image portions
Reexamination Certificate
1999-02-18
2002-10-01
Lee, Thomas D. (Department: 2624)
Image analysis
Image transformation or preprocessing
Combining image portions
C382S276000
Reexamination Certificate
active
06459819
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to an image input system for obtaining image data to be processed in a computer, and particularly to a image input system which enables to obtain image data of a desired scope through synthesizing data of sequences of partial images (hereinafter called divisional images) of an object taken by an image input device into data of a composite image of the desired scope.
Image input devices applied for obtaining image data to be processed in a computer can be grouped into two categories, electric still cameras and image scanners.
The electric still camera, which can take an image with one shot, has superiority in its portability and handiness, without needing any mechanical structure for scanning the object. On the other side, the electric still camera, resolution thereof depending on a limited number of pixels of its area sensor, has a problem that sufficient resolution cannot be obtained when the object size is large, or the object size should be limited for obtaining necessary resolution. Therefore, the electric still camera is regarded as inadequate for obtaining image data of precise pictures such as character documents or drawings.
The image scanners can be further classified according to their scanning mechanism into flatbed type scanners, spatial scanning type scanners and handy type scanners. The flatbed type scanners or the spatial scanning type scanners have problems of portability because of their certainly large scanning mechanism, and their scanning sizes are also limited. The handy type scanners are portable, but unable to acquire wide images because their scanning widths are to be limited within their sizes. Further, there are problems of usability in any type of the image scanners, requiring certainly complicated manipulation, even though they are suitable for inputting document images in view of their excellent resolution.
For resolving these problems of the conventional image input devices, there have been proposed several devices for obtaining a large size image by synthesizing divisional images which are taken from different parts of an object with necessary resolution, making use of an electric still camera (hereinafter simply called a camera) having inexpensive area sensor. In examples of such devices disclosed in “A Hand-Scan Image Digitizer with an Ability for Realtime Assembling of Mosaic Picture” by Okada et al, Technical Journal IE81-17, IEICE (the Institute of Electronics, Information and Communication Engineers), 1981, or in a Japanese patent application entitled “Image Input Device” by Mizoguchi and laid open as a Provisional Publication No. 259557/'94, the divisional images are taken from a fixed object by shifting a position of a camera. By taking the divisional images through shifting the camera position, the user can select only necessary parts of the object to be inputted, which enables to reduce trimming operation after the image acquisition, not only resolves the problem of tradeoff of the resolution and the image size.
However, in the above prior examples, the image synthesis of the divisional images taken by the camera is performed referring only to information included in the divisional images themselves. Therefore, a large amount of calculation and a high processing cost are required, and further, a problem of incorrect synthesis may occur when the object has a monotonous texture having repeated patterns, for example. Still further, when the camera position is displaced manually, there may arise different distortions in different divisional images, such as skews, magnifications or reductions due to distance variation between the camera and the object, making correct image synthesis still more difficult.
Prior arts for dealing with these problems are proposed in a Japanese patent application entitled “Image Input System” and laid open as a Provisional Publication No. 65483/'96, and a Japanese patent application entitled “Image Input Device” and laid open as a Provisional Publication No. 218941/'97.
In the prior art disclosed in the Provisional Publication No. 65483/'96, the image input system comprises an image sensor, a position detector for detecting a position of the image sensor and a synthesizer for synthesizing divisional images taken by the image sensor. For dealing with the above problems, relative displacement of the image sensor is detected each time when a divisional image is taken, and the image synthesis is performed referring to information of the relative displacement thus detected, together with information included in the divisional images.
In the prior art disclosed in the Provisional Publication No. 218941/'97, the image input device has a position detector for detecting a position of a camera from where each of divisional images of an object is taken, a frame for unitizing the position detector and the camera and enabling displacement of the camera parallel to the object, and a means for nominating limited candidates of a position in a synthesized image where each of the divisional images is to be positioned. The above problems are resolved by synthesizing each of the divisional images at a correct position which is selected among the respective candidates.
In the prior arts disclosed in the Provisional Publications No. 65483/'96 and No. 218941/'97, the image input devices are so configured that the user may take the divisional images by manually displacing the camera unit in his desired direction, and in the prior art disclosed in the Provisional Publication No. 218941/'97, errors included in positional information of the divisional images are designed to be removed through image processing. However, user's freedom in displacement of the camera unit also means that the user should displace the camera unit in various directions for obtaining necessary numbers of divisional images, which may force unnatural movement of the user's arm, resulting in increase of the positional errors as well as decrease of usability.
As above described, nature or characteristic of the image input operation performed by the user composes a variation factor of the divisional images. However, the characteristic of the image input operation has not been considered, for improving positioning precision of the divisional images, in any of the prior arts for synthesizing a whole image from divisional images taken from an object with necessary resolution by displacing the camera unit.
SUMMARY OF THE INVENTION
Therefore, a primary object of the present invention is to provide an image input system which enables to obtain a high-precision composite image at high speed with an easy operation from divisional images taken from an object, through calculating positional relations among the divisional images at high speed and with high precision by making use of the characteristic of the image input operation.
Another object of the invention is to provide an image input system wherein a user can perform image input operation without especially considering overlaps between the divisional images.
In order to achieve the objects, an image input system according to the invention comprises:
a camera;
a position detector for detecting a present position of the camera;
a supporting means for supporting the camera to be movable on an object maintaining a fixed distance to the object, and unitizing the position detector and the camera;
an image acquisition indicating means for generating an image acquisition indication referring to the present position of the camera detected by the position detector, each of a plurality of divisional images being acquired according to the image acquisition indication automatically at fixed intervals so as to have an overlap with another of the divisional images;
an acquisition condition collecting means for collecting acquisition condition information of each of the divisional images according to the image acquisition indication generated by the image acquisition indicating means;
a candidate nominating means for nominating a c
Brinich Stephen
Lee Thomas D.
LandOfFree
Image input system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image input system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image input system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2932238