Image analysis – Image transformation or preprocessing – Selecting a portion of an image
Reexamination Certificate
1999-03-30
2004-01-20
Mehta, Bhavesh M. (Department: 2625)
Image analysis
Image transformation or preprocessing
Selecting a portion of an image
C382S103000, C382S284000, C382S294000, C358S450000, C358S540000
Reexamination Certificate
active
06681056
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention deals with the acquisition, storage and display of image data, and more specifically with the construction of a composite image representation which allows multiple levels of display resolution of selected image locations.
2. Description of Prior Art
To acquire images that allow considerable zooming requires expensive photographic equipment that uses photographic film, followed by scanning of the developed film in to a computer for storage and display. The increasing availability of digital processing has made it feasible to use a method that permits more inexpensive digital capturing devices, such as digital camera. With that method no scanning is necessary, since images may be transferred to the computer via a disk or a direct connection from the digital camera. However, these typically have much lower resolution than photographic film. To make up for this lack of resolution, an author who wishes to convey an impression of a scene would in all likelihood take a picture of the overall scene at low angular resolution, and take additional higher resolution pictures of interesting areas.
The referred to interesting areas do not necessarily imply areas with a large amount of information. For example, a crowd in a stadium when viewed at a large distance while containing much detailed information may still be deemed uninteresting. In business and electronic commerce applications, interesting might mean detail that an advertiser might coax a viewer to uncover after selecting the image region with an interactive input or pointing device such as a PC mouse, e.g., allowing a user to zoom in on an image of cloth.
Furthermore, a common technique of three dimensional computer graphic is texture mapping, where a texture is mapped or pasted onto a three dimensional geometric structure. The texture map is usually a two dimensional image of the three dimensional surface, i.e., image of a surface material, photograph of a face, satellite terrain photos, etc. It is these texture maps that complements the three dimensional models to generate more realistic computer graphic pictures. Hence the higher the spatial resolution of the texture maps, the better the quality of the final rendering. Additionally, most current texture mapping processes are efficiently executed by special graphics display hardware requiring binary levels of resolution of the texture map.
SUMMARY OF THE INVENTION
The invention allows the user to select a collection of resolutionally non-uniform, captured digital image views of a static scene, thereby defining the more interesting sections of the scene. The invention provides a system and a method for continuous zooming and display of composite image representation having different levels of resolution at different locations. The invention comprises techniques for acquisition of digital images, composite representation for multiple resolution and overlapping digital images, and the continuous zooming display of the complete image data. The invention is comprised of the following components:
1. Digital Image Acquisition—a user collects a set of desired digital images or views of the same scene. The collected images may be of different sizes, pixel resolutions, and spatial locations, however they must have some degree of spatial overlap.
First, the capturing device is fixed in the desired position to obtain the best viewpoint of the scene. One image encompassing the whole target scene is obtained. This first image usually results in a fairly low resolution. Then, depending on the areas of interest or importance, higher resolution images of the scene are taken at varying magnification. The camera zoom and rotation may be adjusted to capture the desired images, the camera position however does not change. The resulting collection of images captures the user preferred set of resolutionally non-uniform representations of one scene, with higher resolution images taken of more important sections of the scene, and lower resolution images are taken of less important sections.
2. Image Registration—determines the parameterized transforms from each distortion-corrected digital image to a global coordinate system. Since the camera position of the inventive digital image acquisition system is fixated, the parameterized transforms need only include the spatial offset and resolutional magnification for each image relative to a global coordinate system.
3. Anti-Distortion Image Filter—Image registration does not work well when images are affected by lens distortions, unfortunately nonlinear optical distortions result from most image acquisition devices. Therefore, the anti-distortion image filter is used to calibrate and rectify all the digital input images resulting in a corresponding collection of distortion-corrected digital images.
4. Coding—Each distortion-corrected digital image is compressed for storage.
5. Storage—The collection of compressed distortion-corrected digital images and their corresponding image registration information is stored.
6. Heterogeneous Image Pyramid Representation (HIP) is composed and generated from the collection of distortion-corrected digital images and their corresponding image registration information. The HIP representation maintains and manages the different levels of spatial resolution in different positions, it may be a representation for a data structure or storage.
The HIP representation is a generalization of an image pyramid, it is a continuous volumetric representation where images are sparsely resident anywhere in the pyramid and thus are not confined to quantized levels. Input digital images residing on the heterogeneous image pyramid representation are not confined to specific sizes, pixel resolutions, spatial position, spatial rotation, or spatial transformations.
7. Spatial Transformation performs the desired mapping function between the point data of the HIP representation and the output pixels of the display device.
8. Anti-Alias Image Filter—inserted when spatial image magnification and reduction transforms are performed which makes aliasing artifacts evident.
9. Image Interpolation and Resampling generates any portion of a scene at any level of resolutional magnification, at any spatial location, and at any arbitrary image size from the HIP representation. Since the HIP representation provides only discrete image data at integral positions, an interpolation stage is introduced to construct the continuous surface through the image data. Next, the continuous surface may be resampled at any arbitrary spatial location or rotation.
10. The Display device renders the image output of the image interpolation and resampling process at varying spatial resolution and display size.
Furthermore, it should be noted that the invention is well suited to be used to pre-compute the desired levels of resolutions of texture for texture mapping.
REFERENCES:
patent: 5204944 (1993-04-01), Wolberg et al.
patent: 5568600 (1996-10-01), Kaba
patent: 5649032 (1997-07-01), Burt et al.
patent: 5666475 (1997-09-01), Salesin et al.
patent: 5680486 (1997-10-01), Mita et al.
patent: 5963664 (1999-10-01), Kumar et al.
patent: 6009197 (1999-12-01), Riley
patent: 6173087 (2001-01-01), Kumar et al.
Hendriks Ferdinand
Shae Zon-Yin
Tseng Belle L.
Wang Xiping
Cameron Esq. Douglas W.
Kassa Yosef
Mehta Bhavesh M.
Scully Scott Murphy & Presser
LandOfFree
Method and system for digital image acquisition and... 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 and system for digital image acquisition and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and system for digital image acquisition and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3189026