Computer graphics processing and selective visual display system – Computer graphics processing – Graphic manipulation
Reexamination Certificate
2001-05-07
2003-10-14
Luu, Matthew (Department: 2672)
Computer graphics processing and selective visual display system
Computer graphics processing
Graphic manipulation
C345S634000, C345S215000
Reexamination Certificate
active
06633303
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a system and method for generating a wide-area high-resolution image by capturing partial images (sub-images) of parts of a target object such as a document and thereafter connecting or synthesizing the sub-images.
Description of the Related Art
When a document of a large size is captured and inputted as image data by use of a capture device (video camera etc.) in a high resolution that is suitable for character recognition, the document is generally segmented into parts and the segmented parts are captured individually by the capture device to be synthesized or connected later. An example of such a conventional wide-area high-resolution image generation method has been disclosed in Japanese Patent Application Laid-Open No. HEI10-210272 (hereafter, referred to as “document No.
1
”). In the conventional technique of the document No.
1
, a wide-area high-resolution image is obtained as described below.
First, the whole of a document (the target object of capturing) is preliminarily captured by a camera and thereby an image of the whole of the document (hereafter, referred to as a “total image”) is obtained. Subsequently, the number of sub-images (images of parts of the document) necessary for covering the whole of the document with a predetermined resolution is determined and thereby a zoom ratio (power) to be used for capturing the sub-images is determined. Subsequently, each part of the document (including the periphery (margin) of the part) is successively captured by the camera with the zoom ratio determined above, and thereby sub-images of the parts of the document are obtained. Subsequently, adjoining sub-images are successively connected together by use of image information (pixel intensity, structure such as lines, etc.) and thereby a wide-area high-resolution image of the whole of the document is generated.
A method which is similar to the wide-area high-resolution image generation method of the document No.
1
has been disclosed in a document: Noboru Nakajima, Naoya Tanaka and Keiji Yamada “Document Reconstruction and Recognition from an Image Sequence,” Proceedings 14th International Conference on Pattern Recognition (IAPR (International Association for Pattern Recognition)), Vol. 1 pp. 922-925 (1998) (hereafter, referred to as “document No.
2
). In the method of the document No.
2
, the connection of the adjoining high-resolution sub-images is conducted at small calculation cost, by executing structure analysis to each sub-image and using structure information (hierarchical layout structure such as, characters, character strings, text blocks (paragraphs), columns, etc.) of the sub-images.
However, the conventional wide-area high-resolution image generation methods which have been described above involves the following problems or drawbacks. First, if a document having some large blank areas is segmented into parts and sub-images are obtained by capturing parts of the document, some of the sub-images might be totally blank or might be images having blank margins. In such cases, the connection of the adjoining sub-images and the generation of the wide-area high-resolution image become impossible since the connecting margins of such sub-images do not have structure information (lines, characters, etc.) to be used for the registration. Therefore, the conventional methods are not capable of generating a wide-area high-resolution image when a relatively large blank area exists in the document or in the originally captured total image.
Further, the two adjoining sub-images to be connected together are generally required to have the same resolution. Therefore, in the conventional methods (which successively connect adjoining sub-images for obtaining the wide-area high-resolution image), all the sub-images have to be captured in the same resolution. If a document to be captured includes a short text or character string that has to be captured in a high resolution for character recognition etc., the whole of the document has to be segmented into a lot of small parts according to the required high resolution and all the parts have to be captured in the same high resolution, taking an enormous processing time.
SUMMARY OF THE INVENTION
It is therefore the primary object of the present invention to provide a system and method for generating a wide-area high-resolution image, by which the wide-area high-resolution image can successfully be generated even if a relatively large blank area existed in the document to be captured or in the originally captured total image.
Another object of the present invention is to provide a system and method for generating a wide-area high-resolution image, by which the wide-area high-resolution image can be generated by use of images of different resolutions.
Another object of the present invention is to provide a system and method for generating a wide-area high-resolution image, which can decrease the number of necessary sub-images and shorten the processing time when one or more areas (such as blank areas) that do not require high resolution capturing existed in the originally captured total image.
In accordance with a first aspect of the present invention, there is provided a wide-area high-resolution image generation method comprising a total image acquisition step, an image structure analysis step, a sub-image acquisition step, an image connection step, an image extraction step and a wide-area high-resolution image generation step. In the total image acquisition step, a target object of the generation of a wide-area high-resolution image is captured by a capture device and thereby a total image of the target object is acquired. In the image structure analysis step, image structure analysis is conducted to the total image of the target object and thereby structural elements are extracted from the total image and position information of each structural element is obtained. The sub-image acquisition step is conducted for one or more of the structural elements. In the sub-image acquisition step, one or more partial areas and a resolution to be used for capturing the structural element are determined and sub-images of the partial areas of the structural element are acquired by the capture device with the determined resolution. The image connection step is conducted for each of the structural elements to which the sub-image acquisition step has been conducted. In the image connection step, the sub-images of the partial areas of the structural element are connected together by use of image information of the sub-images and thereby an image of the structural element having the determined resolution is obtained as a synthesis target image. The image extraction step is conducted for each of the structural elements to which the sub-image acquisition step has not been conducted. In the image extraction step, part of the total image corresponding to the structural element is extracted from the total image as a synthesis target image. In the wide-area high-resolution image generation step, the synthesis target images of the structural elements obtained in the image connection steps and the image extraction steps are synthesized so that relative position relationship of the synthesis target images will be the same as that of the structural elements in the total image of the target object based on the position information of the structural elements obtained in the image structure analysis step and thereby a wide-area high-resolution image of the target object is obtained.
In accordance with a second aspect of the present invention, in the first aspect, the wide-area high-resolution image generation method further comprises an attribute determination step. In the attribute determination step, the attribute of each structural element is determined based on the image structure analysis. The sub-image acquisition step is conducted for structural elements having attributes that require higher resolution than that of the total image.
In accordance with a third aspect of the present invention, in the
Luu Matthew
NEC Corporation
Sughrue & Mion, PLLC
LandOfFree
Method, system and record medium for generating wide-area... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method, system and record medium for generating wide-area..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method, system and record medium for generating wide-area... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3153179