Method and device for synthesizing all-objects-in-focus images
AssigneeTexas Instruments Incorporated
InventorHirohisa Yamaguchi
Life of this patent
Estimating the term…
Abstract
A method and apparatus for forming a synthesized optical image with all of the desired objects in focus. A prescribed number of optical images of the same scene are provided by recording an optical image at a respective focal distance and changing the focal distance for each additional optical image as recorded. The optical images are subjected to wavelet transformation to form a multi-resolution representation. The coefficients of the various multi-resolution representations are compared at the same position to detect the maximum spectral amplitudes. Based on the detected maximum spectral amplitudes, another multi-resolution representation is obtained. Then, an inverse wavelet transformation is performed to obtain the synthesized optical image.
Claims3
A method for synthesizing an all-objects-in-focus two-dimensional image of several objects at various corresponding object distances, comprising the steps of:
- for each object capturing a corresponding pixel image focused at its object distance;
- transforming each pixel image into a corresponding multi-resolution representation expressed as a set of coefficients;
- comparing, at each image position, the various coefficients of the multi-resolution representations to detect the maximum spectral amplitude at each image position;
- deriving a synthesized multi-resolution representation by using the coefficient having the maximum spectral amplitude at each image position;
- reverse-transforming the synthesized multi-resolution representation by inverse wavelet transformation into a synthesized optimally focused image.
A method for synthesizing an all-objects-in-focus two-dimensional image of a scene of N objects (O.sub.1, O.sub.2, O.sub.3. . . O.sub.N) positioned at various corresponding object distances (d.sub.1, d.sub.2, d.sub.3. . . d.sub.N), comprising:
- for each object capturing a corresponding two dimensional original pixel image (I.sub.1, I.sub.2, I.sub.3. . . I.sub.N) of the scene focused at its object distance;
- transforming each original pixel image by wavelet transformation to obtain a corresponding multi-resolution representation comprising a set of stored coefficents (A.sub.ij.sup.(1), A.sub.ij.sup.(2), A.sub.ij.sup.(3). . . . A.sub.ij.sup.(N)) corresponding to image positions (i,j) and spacial frequencies;
- for each image position (i,j), selecting from among the N multi-resolution representations the coefficient of maximum spectral amplitude;
- combining the selected multi-resolution coefficients into a synthesized multi-resolution representation A.sub.ij.sup.(MAX) ;
- and
- reverse-transforming the synthesized multi-resolution representation by a reverse multiple-resolution transformation to create a synthesized optimally focused image.
A device for synthesizing an all-objects-in-focus two dimensional image of a scene of N objects (O.sub.1, O.sub.2, O.sub.3. . . O.sub.N) positioned at various corresponding object distances (d.sub.1, d.sub.2, d.sub.3. . . d.sub.N), comprising:
- a camera for capturing for each object a corresponding two dimensional original pixel image (I.sub.1, I.sub.2, I.sub.3. . . I.sub.N) of the scene focused at its object distance;
- a first filter for transforming each original pixel image by wavelet transformation into a corresponding multi-resolution representation comprising a set of stored coefficents (A.sub.ij.sup.(1), A.sub.ij.sup.(2), A.sub.ij.sup.(3), . . . . A.sub.ij.sup.(N)) corresponding to image positions (i,j) and spacial frequencies;
- a comparator for selecting for each image position (i,j) from among the N multi-resolution representations the coefficient of maximum spectral amplitude;
- a synthesizer circuit for combining the selected multi-resolution coefficients into a synthesized multi-resolution representation A.sub.ij.sup.(MAX) ;
- and
- a second filter for reverse-transforming the synthesized multi-resolution representation by a reverse multiple-resolution transformation to create a synthesized optimally focused image.
Dependencies are inferred from the claim text.