Image analysis – Image transformation or preprocessing – Changing the image coordinates
Reexamination Certificate
1998-10-15
2003-05-27
Couso, Yon J. (Department: 2625)
Image analysis
Image transformation or preprocessing
Changing the image coordinates
C382S312000
Reexamination Certificate
active
06571023
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a camera device to which multiple differently focused images can be imaged at the same time, and an image processing method in which the desired image is generated from multiple differently focused images.
2. Related Art Statement
As a conventional example of the image for generating desired image or picture from multiple images, for example, there is an image-processing device based on the region division. In this conventional image processing method, for example, multiple different focused images are prepared, the focused region is judged respectively in respective images, the desired image is generated by performing a series of processing by which a fixed visual effect is given to each region by dividing the region of the multiple image based on the judgment result. In that case, it is general to use the image processing program which describes the procedure to process the above region decision, the region division, and the visual effect successively,
Moreover, as a conventional example of the camera device used to image the multiple images used for the above image processing method, there is for example a usual imaging lens control type camera device
51
shown in FIG.
2
. This camera device
51
is constituted as a camera device, which has one optical system. In order to obtain multiple images which are imaged by different focusing for the same scene (for example, scene
52
of
FIG. 2
) by using conventional camera device
51
. It is necessary to perform imaging of multiple times by changing focusing.
That is, when n kinds of images are imaged according to different focusing by using conventional camera device
51
shown in
FIG. 2
, a zoom lens
53
is controlled by the manual operation, or by servo lens controller
54
provided to outside of the camera, first of all, after the focusing of a zoom lens
53
is controlled to focus on the first depth, the first image is acquired, next, after the focusing of the zoom lens
53
is controlled to focus on the second depth, the second image is acquired, and similarly after the focusing of the zoom lens
53
is controlled to focus on n depth, the nth image is acquired. Thus, the focusing and the imaging of n times are needed to acquire the image focused on n kinds of depth. Moreover, the imaged picture is forwarded from CCD not shown in the figure to a workstation
55
, and will be supplied to the image processing.
Since the above conventional image processing method uses the decision condition “Focused region”, when the region which has the even brightness value in the imaging subject scene and the region with the depth change are existent, the judgment precision of the region decision cannot be secured enough about those regions. Therefore, the above application range of the conventional image processing method is limited to a sharpening of the image etc. according to the integrated focused regions, so that the extension into more functional image processing is extremely difficult, such as the focus blur is arbitrarily adjusted every region and a stereoscopic image etc. is given, thereby generating the pseudo disparity.
Moreover, as long as the above conventional camera device in which the system of controlling the zoom lens from the outside is adopted, it is necessary to repeat the imaging by one imaging section every depth to be focused. Multiple images with different focusing cannot fundamentally be imaged to the dynamically changed imaging subject at the same time, so that it is not possible to use for the imaging of moving scene, and thus the application is limited only to the still image.
In addition, the zoom lens generally used as an imaging lens has an extremely complex control structure using the multiple lenses itself, if focusing is adjusted from the outside, it is not avoided that the optical center of an internal lens group which is the effective view position (main point) changes into back and forth direction of an optical axis. Therefore, when focusing is adjusted, three dimensional conditions will also be added, for example, a phenomenon is caused since an object in the back plain covered with a front object in the picture imaged in a certain condition, appears newly in the scene because of moved view point and in another image and thus imaged. As a result, the requirement “The imaged subject (object etc.) is same, except for the magnification change at the entire image among images in which the same scene is imaged, even in the change of focusing” can not be met, so that as for the image generated by the image processing, deterioration in picture quality which originates in the above phenomenon is not avoided by using the imaged picture or image.
SUMMARY OF THE INVENTION
The first object of the present invention is to provide a camera device in which multiple differently focused images can be imaged at the same time.
The second object of the present invention is to provide an image processing method in which the desired image is generated from multiple differently focused images.
To this first end, according to the present invention, there is provided a camera device for multiple pictures imaging types comprising an imaging lens mechanism; a prism means having an incident surface for receiving a light emanated from the imaging lens mechanism, and n output surfaces for emitting distributed light by dividing the incident light into n multiple equal parts (n=2,3 . . . ), respectively; n imaging sections for imaging the distributed light emitted from the n output surfaces of the prism means, respectively; a focusing control mechanism for adjusting a positional relation between the n output surfaces and the corresponding imaging sections, respectively; and a synchronizing means for synchronizing a timing of the imaging by the n imaging sections; whereby multiple images may be imaged by different focussing, simultaneously.
To this second end, according to the present invention, there is provided an image processing method using multiple images comprising steps of: imaging multiple differently focused images I
m
(m=1, . . . , n); deriving a following identical equation based on the superposition models realized between the different multiple focusing images Im and desired images I;
f
(
T
1
, . . . , T
k
, T
11
, . . . , T
1n
, T
21
, . . . , T
2n
, T
m1
, . . . , T
mk
, T
n1
, . . . , T
nn
, I, I
1
, . . . , I
n
)=0 (1)
where, n=2, 3, . . . , k=1, . . . , n and m=1, . . . , n, T
mk
is a function showing convolution according to blur function decided from camera property and camera parameter, and T
k
is a function specified by a user showing a convolution according to blur function in the image and showing an amount of parallel movement corresponding to disparity;
obtaining a solution by substituting multiple images I
m
and the functions T
mk
and T
k
of different focusing into the identical equation;
performing an image processing for generating desired image I based on this solution; and
selectively generating a desired image I by controlling the function T
k
to obtain one of (a) any one of various images which contain a full focused image in which a full region of the image is sharpened, (b) an arbitrary focused image in which sharpness was adjusted every depth, (c) an image with disparity for binocular vision in which disparity is given, (d) an arbitrary focused image with disparity for binocular vision in which disparity is given to the arbitrary focused image, (e) a partially emphasized image in which only a specific region is emphasized, or (f) a partially extracted image in which only a specific region is extracted.
According to the present invention, as a preprocessing of encoding, a blooming control is preformed on depth direction every different region.
According to the present invention, the light emitted from the imaging lens mechanism enters into an incident plane of the prism means, thereby becoming multiple n (n=2, 3, . . . ) distributed lights div
Aizawa Kiyoharu
Kodama Kazuya
Couso Yon J.
Frank Robert J.
The University of Tokyo
Venable
Wells Ashley J.
LandOfFree
Image processing method using multiple images 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 processing method using multiple images, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing method using multiple images will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3084973