`Dagher et al.
`
`(10) Patent No.:
`(45) Date of Patent:
`
`US 8,824,833 B2
`Sep. 2, 2014
`
`USOO8824.833B2
`
`(54) IMAGE DATA FUSION SYSTEMS AND
`METHODS
`
`(56)
`
`(75) Inventors: Joseph C. Dagher, Boulder, CO (US);
`Amit Ashok, Boulder, CO (US); David
`Tremblay, Boulder, CO (US); Kenneth
`S. Kubala, Boulder, CO (US)
`(73) Assignee: Omnivision Technologies, Inc., Santa
`Clara, CA (US)
`Subject to any disclaimer, the term of this
`patent is extended or adjusted under 35
`U.S.C. 154(b) by 380 days.
`12/865,343
`Jan. 30, 2009
`
`(*) Notice:
`
`(21) Appl. No.:
`(22) PCT Filed:
`(86). PCT No.:
`S371 (c)(1),
`(2), (4) Date:
`(87) PCT Pub. No.: WO2009/097552
`PCT Pub. Date: Aug. 6, 2009
`
`PCT/US2O09/032683
`
`Nov. 29, 2010
`
`(65)
`
`Prior Publication Data
`US 2011/0064327 A1
`Mar. 17, 2011
`
`(2006.01)
`(2006.01)
`(2006.01)
`
`(51) Int. Cl.
`G06K 9/32
`G06T5/00
`G06T5/50
`(52) U.S. Cl.
`CPC. G06T5/50 (2013.01); G06T5/004 (2013.01);
`G06T 2207/20221 (2013.01); G06T 2207/10148
`(2013.01)
`USPC ........................................... 382/294; 382/260
`(58) Field of Classification Search
`CPC ............................................. GO6T 2207/2O221
`See application file for complete search history.
`
`
`
`References Cited
`U.S. PATENT DOCUMENTS
`
`7, 1992 Ritchey........................... 348/39
`5,130,794. A *
`12/1992 Takemoto et al.
`5,172,236 A
`1/1994 Mimura et al.
`5,282,045 A
`5,771,416 A * 6/1998 Mukai et al. ...
`6,128.416 A * 10/2000 Oura ..........
`6,201,899 B1
`3/2001 Bergen
`6,654,013 B1 * 1 1/2003 Malzbender et al. ......... 345.426
`6,856,708 B1
`2/2005 Aoki
`7,274,830 B2
`9, 2007 Bacarella et al.
`(Continued)
`OTHER PUBLICATIONS
`
`396,378
`... 382,284
`
`Snavely et al., “Photo Tourism: Exploring photo collections in 3D.”
`ACM Transactions on Graphics, 25(3), Aug. 2006.*
`(Continued)
`Primary Examiner — Bhavesh Mehta
`Assistant Examiner — Andrew Moyer
`(74) Attorney, Agent, or Firm — Lathrop & Gage LLP
`
`ABSTRACT
`(57)
`Systems and methods for image data fusion include providing
`first and second sets of image data corresponding to an
`imaged first and second scene respectively. The scenes at least
`partially overlap in an overlap region, defining a first collec
`tion of overlap image data as part of the first set of image data,
`and a second collection of overlap image data as part of the
`second set of image data. The second collection of overlap
`image data is represented as a plurality of image data Subsets
`Such that each of the Subsets is based on at least one charac
`teristic of the second collection, and each Subset spans the
`overlap region. A fused set of image data is produced by an
`image processor, by modifying the first collection of overlap
`image databased on at least a selected one of, but less than all
`of the image data Subsets.
`39 Claims, 18 Drawing Sheets
`
`APPL-1025 / Page 1 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`US 8,824,833 B2
`Page 2
`
`(56)
`
`References Cited
`
`U.S. PATENT DOCUMENTS
`
`11/2001 Okisu et al.
`2001/0045982 A1
`10, 2002 Sakurai et al.
`2002fO140823 A1
`2/2003 Elder et al. ...................... 386/46
`2003/0026588 A1*
`3, 2004 Tiana
`2004/0047518 A1
`4/2004 Afsenius
`2004/008.0661 A1
`2004/0105569 A1* 6/2004 Sharma et al. ................ 382/100
`2004/0234154 A1* 11, 2004 Hier ............
`382,254
`2005/0248590 A1* 11, 2005 Tian et al.
`345,660
`2006/0050338 A1
`3f2006 Hattori .............................. 359/9
`2006, OO61678 A1
`3, 2006 Yamazaki
`2007/0188601 A1* 8/2007 Rohaly et al. ................... 348/47
`2007/0247517 A1 10/2007 Zhang et al.
`2008.0056612 A1
`3f2008 Park et al. ..................... 382,284
`2008/02186 13 A1* 9, 2008 Janson et al. ................. 348,262
`2011 OO19048 A1
`1/2011 Raynor et al.
`OTHER PUBLICATIONS
`
`
`
`Bao and Xu, "Complex wavelet-based image mosaics using edge
`preserving visual perception modeling. Computers & Graphics 23.3
`(1999): 309-321.*
`Brown and Lowe, "Recognising panoramas.” Proceedings of the
`Ninth IEEE International Conference on Computer Vision, vol. 2,
`No. 1218-1225, 2003.*
`Klarquistand Bovik, “Fovea: A foveated vergent active stereo vision
`system for dynamic three-dimensional scene recovery.” Robotics and
`Automation, IEEE Transactions on 14.5 (1998): 755-770.*
`Kuhnlenz et al., “A multi-focal high-performance vision system.”
`Robotics and Automation, 2006. ICRA 2006, Proceedings 2006
`IEEE International Conference on, IEEE, 2006.*
`Scassellati, "A binocular, foveated active vision system.” No. AI-M-
`1628, Massachusetts Inst of Tech Cambridge Artificial Intelligence
`Lab, 1999.*
`
`Zhao et al., “Broadband and wide field of view foveated imaging
`system in space.” Optical Engineering 47.10 (2008): 103202
`1032O2.
`Wikipedia, http://en.wikipedia.org/wiki/Image Scaling, Jan. 29.
`2007.*
`Drori, Iddo, and Dani Lischinski. “Fast multiresolution image opera
`tions in the wavelet domain.” Visualization and Computer Graphics,
`IEEE Transactions on 9.3 (2003): 395-411.*
`Hill, Paul R., Cedric Nishan Canagarajah, and David R. Bull. “Image
`fusion using complex wavelets.” BMVC. 2002.*
`International Search Report and Written Opinion issued in related
`PCT patent application PCT/US2009/032683, dated Jan. 30, 2009,
`14 pages.
`Kiyoharu, et al., “Producing Object-Based Special Effects by Fusing
`Multiple Differently Focused Images.” IEEE Transactions on Cir
`cuits and Systems for Video Technology, IEEE Service Center, vol.
`10, No. 2, Mar. 1, 2000.
`Kazuya, et al., “All-in-Focus Image Generation by Merging Multiple
`Differently Focused Images in Three-Dimensional Frequency
`Domain' Advances in Multimedia Information Processing PCT
`2005 Lecture Notes in Computer Science, vol. 3767, pp. 303-314.
`Jan. 1, 2005.
`Hong, Sahyun, et al. “Data Fusion of Multiple Polarimetriv SAR
`Images. Using Discrete Wavelet Transform (DWT)” IEEE, 3323
`3325, 2002.
`Office Action issued in related Taiwanese Patent Application
`098103287 dated Jan. 9, 2013, 29 pages.
`U.S. Appl. No. 13/281,674. Office Action issued Sep. 10, 2013, 28
`pageS.
`U.S. Appl. No. 13/281,674 Response to Office Action filed Dec. 10,
`2013, 9 pages.
`
`* cited by examiner
`
`APPL-1025 / Page 2 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 1 of 18
`
`US 8,824,833 B2
`
`
`
`30
`
`FIG. 1
`
`APPL-1025 / Page 3 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 2 of 18
`
`US 8,824,833 B2
`
`
`
`120
`
`100
`
`FIG. 2A
`
`APPL-1025 / Page 4 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 3 of 18
`
`US 8,824,833 B2
`
`
`
`| N N N
`
`FIG. 2B
`
`APPL-1025 / Page 5 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 4 of 18
`
`US 8,824,833 B2
`
`164
`
`
`
`Multi-aperture Camera
`
`Processor
`
`Image Output Device
`
`FIG. 3
`
`APPL-1025 / Page 6 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 5 of 18
`
`US 8,824,833 B2
`
`
`
`N12 Pixels
`Wide
`
`N/2 Pixels
`Tele
`
`Upsample?
`Interpolate
`
`175
`
`Upsample?
`Interpolate
`
`N Pixels
`Wide
`
`N Pixels
`Tele
`
`
`
`
`
`
`
`
`
`17O
`
`FG. 4
`
`APPL-1025 / Page 7 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 6 of 18
`
`US 8,824,833 B2
`
`High and Low Frequency
`Tele Data
`
`Low-Frequency
`Wide Data
`
`18O ,
`
`ESNOCHSENH
`
`2OO
`
`NYOUIST FREOUENCY
`(cycles/degree)
`
`
`
`FIG. 5
`
`APPL-1025 / Page 8 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 7 of 18
`
`US 8,824,833 B2
`
`190
`
`Upsampled
`Wide Signal
`
`High-passed
`Tele Signal
`
`192
`
`194
`
`196
`
`CO
`Z
`O
`n ?
`Y
`
`8.5
`NYOUIST FRECQUENCY
`GO (cycles/degree)
`
`17
`
`193
`
`
`
`a CO
`
`Z
`O
`5
`
`
`
`8.5
`NYOUIST FREOUENCY
`G) (cycles/degree)
`
`17
`
`191
`
`195
`
`193
`
`FIG. 6
`
`
`
`4. 17
`NYOUIST FREQUENCY
`(Cycles/degree)
`
`High-Resolution
`Overlap Region
`
`APPL-1025 / Page 9 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 8 of 18
`
`US 8,824,833 B2
`
`a^a a Aa
`-
`
`a
`.
`^^^^
`
`
`
`
`
`N/2 Pixels
`Wide
`
`Upsample?
`Interpolate
`
`
`
`
`
`
`
`
`
`
`
`FIG. 7
`
`High-pass
`filter
`
`Upsample?
`Interpolate
`
`
`
`W. 3 2.
`N Pixels 3 Tele
`2 a% 2 %
`2 .
`
`
`
`APPL-1025 / Page 10 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 9 of 18
`
`US 8,824,833 B2
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`300
`
`320
`
`*****
`
`N/2 Pixels
`Wide
`RGB
`
`
`
`Upsample/Interpolate
`
`RGB to YUV
`Conversion
`
`Convert RGB to YUV
`
`334
`
`-
`
`
`
`
`
`Upsample/
`interpolate
`
`High-pass
`filter
`
`No.
`
`No. ·
`
`312
`
`FIG. 8
`
`APPL-1025 / Page 11 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 10 of 18
`
`US 8,824,833 B2
`
`300
`
`
`
`N/2 Pixels
`Wide
`RGB
`
`Upsample/Interpolate
`
`
`
`
`
`32O
`
`338
`
`Discard
`
`UV
`
`
`
`
`
`Convert RGB to YUV --
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`334
`
`v
`
`4
`
`a
`
`...
`
`•
`
`
`
`
`
`
`
`Register Overlap
`Region
`
`Upsample?
`Interpolate
`
`Tele - Wide
`
`Tele (hi)
`
`312
`
`N Pixels
`Wide
`
`FIG. 9
`
`Tele
`
`APPL-1025 / Page 12 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 11 of 18
`
`US 8,824,833 B2
`
`
`
`s
`
`
`
`XepuuunOO
`
`S.
`
`3
`
`APPL-1025 / Page 13 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 12 of 18
`
`US 8,824,833 B2
`
`301
`
`365
`r
`
`302
`
`367 u
`
`CONVERT COLOR
`TO
`YUV
`(YC, Uc, Vc)
`
`
`
`REGISTRATION
`(match Yoto YG)
`
`(UC, Vo)
`
`
`
`
`
`371
`
`/ >
`373
`
`IMAGE FUSION
`
`FIG. 11
`
`APPL-1025 / Page 14 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`US 8,824,833 B2
`
`
`
`SN
`c5
`
`(1)699
`
`APPL-1025 / Page 15 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 14 of 18
`
`US 8,824,833 B2
`
`
`
`(Z)699
`
`APPL-1025 / Page 16 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 15 of 18
`
`US 8,824,833 B2
`
`
`
`APPL-1025 / Page 17 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 16 of 18
`
`US 8,824,833 B2
`
`
`
`§ 18
`
`APPL-1025 / Page 18 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 17 of 18
`
`US 8,824,833 B2
`
`206
`
`10
`
`20
`15
`Pixel position
`
`25
`
`30
`
`
`
`FIG. 16
`
`APPL-1025 / Page 19 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`U.S. Patent
`
`Sep. 2, 2014
`
`Sheet 18 of 18
`
`US 8,824,833 B2
`
`180
`
`
`
`170
`
`140
`
`130
`
`120
`
`110
`
`Pixel position
`
`FIG. 17
`
`APPL-1025 / Page 20 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`1.
`IMAGE DATA FUSON SYSTEMIS AND
`METHODS
`
`US 8,824,833 B2
`
`CROSS-REFERENCE TO RELATED
`APPLICATIONS
`
`This application claims priority to U.S. Provisional Patent
`application No. 61/025,533, filed on 1 Feb. 2008 and entitled
`MULTI-FOCAL LENGTH IMAGE FUSION, U.S. Provi
`sional Patent application No. 61/051,338, filed 7 May 2008
`and entitled TRANSFORM DOMAIN REGISTRATION
`FOR IMAGE FUSION, and U.S. Provisional Patent applica
`tion No. 61/059,319, filed 6 Jun. 2008 and entitled TRANS
`FORM DOMAIN REGISTRATION FOR IMAGEFUSION.
`All of the above-identified applications are incorporated
`herein by reference in their entireties.
`
`10
`
`15
`
`BACKGROUND
`
`25
`
`30
`
`35
`
`40
`
`45
`
`Small, digital cameras integrated into mobile electronics
`such as mobile phones, personal digital assistants (“PDAs)
`and music players are becoming ubiquitous. Each year,
`mobile phone manufacturers add more imaging features to
`their handsets, causing these mobile imaging devices to con
`Verge towards feature sets that consumers expect from stand
`alone digital still cameras. At the same time, the size of these
`handsets is shrinking, making it necessary to accordingly
`reduce the total size of the camera modules while still adding
`imaging features. Optical Zoom is a primary feature that many
`digital still cameras have that many mobile phones may not
`have, primarily due to the severe size constraints in mobile
`imaging devices.
`Cameras (including digital cameras) may be arranged to
`receive electromagnetic radiation (such as visible light)
`through an aperture that can be defined by the camera based
`on a number of well known techniques. For example, an
`optical Sub-system, including one or more lenses and/or other
`optical elements, may define the aperture such that the
`received radiation is imaged by the optical Sub-system and a
`resulting image is directed towards a sensor region such as a
`sensor array that includes a plurality of detectors defining a
`sensing Surface. The sensor region may be configured to
`receive the image and to generate a set of image databased on
`the image. In some common applications, such as when using
`conventional digital cameras to capture images, the camera
`may be aligned to receive electromagnetic radiation associ
`ated with scenery having a given set of one or more objects. In
`these applications the set of image data is, for example, rep
`resented as digital image data using an electrical signal con
`veyed by electrical conductors or stored using memory or
`other digital storage techniques. In addition, the set of image
`data can be processed using a number of known image pro
`cessing techniques.
`In the context of the present disclosure, "Zoom” may be
`understood as a capability to provide different magnifications
`of the same scene and/or object by changing the focal length
`of an optical system, with a higher “level of Zoom’ being
`associated herein with greater magnification and a lower level
`of Zoom being associated with lower magnification. In typical
`film-based cameras, as well as in conventional digital cam
`60
`eras, optical Zoom can be accomplished with multiple lens
`groups that are moved along an optical axis of an imaging
`system for defining a range of different lens configurations.
`For any given configuration, the position of the lens groups
`determines a focallength specific to that configuration. Based
`on well known techniques, camera users can adjustably con
`trol the positioning of the lens groups for selecting a specific
`
`50
`
`55
`
`65
`
`2
`level of Zoom. At any specific level of Zoom associated with a
`selected focal length of a camera's optical Sub-assembly, an
`image represents a portion of a given scene based in part on
`the field of view defined by the lens system. For example, an
`image plane can be defined by the camera's sensor region
`(such as a sensor array), and the resulting image represents a
`field of view consistent with (i) a shape and transverse extent
`of the sensor region’s sensing Surface, and (ii) the selected
`focal length. For a given camera, there is a tradeoff between
`Zoom and field of view Such that camera settings exhibiting
`longer focal lengths generally tend to result in a greater level
`of Zoom in conjunction with correspondingly narrower field
`of view. Conversely, camera settings exhibiting compara
`tively shorter focal lengths tend to result in a lower level of
`Zoom in conjunction with a wider field of view.
`Certain film-based cameras and digital cameras utilize a
`fixed focus imaging system, and these cameras generally do
`not feature adjustable optical Zoom. Fixed focus imaging
`systems are especially common in PDAs. The high complex
`ity, cost and decreased durability typically associated with
`moveable lenses (e.g., in cameras having optical Zoom) limit
`their use in inexpensive camera modules Such as mobile
`phone camera modules and other low cost modules. Film
`based cameras with fixed focus imaging systems generally
`offer no means for the user to adjust the degree of magnifi
`cation while preparing to take a picture. On the other hand,
`digital cameras having fixed optical focus can incorporate
`digital Zoom to allow the user to control the level of Zoom
`before and/or after capturing the image by generating a cor
`responding set of image data. For example, digital Zoom can
`utilize computer-processed cropping followed by signal
`upsampling and data interpolation of the cropped image to
`convert the cropped image to the original display size. As a
`result, however, the resolution of the cropped, final image is
`decreased and the image quality Suffers.
`
`SUMMARY
`
`The following embodiments and aspects thereof are
`described and illustrated in conjunction with systems, tools
`and methods, which are meant to be exemplary and illustra
`tive, not limiting in Scope. In various embodiments, one or
`more problems and/or limitations associated with the above
`described systems and methods have been addressed, while
`other embodiments are directed to other improvements.
`In an embodiment, an imaging method utilizes a multi
`aperture imaging system for producing a fused set of image
`data. This method may include providing a multi-aperture
`camera having first and second Sub-cameras including a first
`Sub-camera, having imaging optics defining a first aperture,
`with the first camera configured for imaging a first scene
`through the first aperture and for generating a first set of
`image data corresponding to the imaged first scene. A second
`camera may be provided, having imaging optics defining a
`second aperture, and the second Sub-camera may be config
`ured for imaging a second scene through the second aperture
`and for generating a second set of image data corresponding
`to the imaged second scene. The second sub-camera can be
`aligned Such that the second scene at least partially overlaps
`the first scene in an overlap region that defines (i) a first
`collection of overlap image data as part of the first set of
`image data for the imaged first scene and (ii) an at least
`generally corresponding, second collection of overlap image
`data as part of the second set of image data for the imaged
`second scene. The second collection of overlap image data of
`the second scene may be represented as a plurality of image
`data Subsets based on at least one associated characteristic of
`
`APPL-1025 / Page 21 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`US 8,824,833 B2
`
`10
`
`25
`
`3
`the second collection of overlap image data, such that each
`Subset is Superimposed across the overlap region. A fused set
`of image data can be produced from the first set of image data
`by changing the first collection of overlap image data in the
`overlap region of the first scene based on at least a selected 5
`one of, but less than all of the image data Subsets.
`In one aspect, representing the second collection of overlap
`image data may include configuring the plurality of image
`data subsets such that each subset is based on a different
`characteristic as compared to the characteristic associated
`with any one of the other subsets.
`In another aspect, the first collection of overlap image data
`may include a first collection of luminance data, and the
`selected one of the image data Subsets may be a luminance
`channel (of luminance data) based on luminance as the char- 15
`acteristic of the second collection of overlap image data, and
`changing of the first collection of overlap image data may
`include combining the first and second collections of lumi
`nance data. Arranging of the second sub-camera may include
`Supplying the second Sub-camera as a grayscale camera for
`providing the luminance channel as being composed of gray
`scale scaled image data.
`In yet another aspect, representing the second collection of
`overlap image data may include filtering the second collec
`tion of overlap image data such that the selected image data
`Subset is composed of filtered data, and filtering the second
`collection of overlap image data may include applying con
`Volution filtering to the second collection of overlap image
`data Such that the selected image data Subset is influenced by
`the convolution filtering. Furthermore, representing the sec
`ond collection of overlap image data may include Scaling the
`second collection of overlap image data such that the selected
`image data Subset is composed of scaled data.
`In an additional aspect, the second collection of overlap
`image data may include intensity information, and Scaling the
`second collection of overlap image data may include chang
`ing at least Some of the intensity information. In this case
`Scaling the second collection of overlap image data includes
`applying a gain for causing the changing of the intensity
`information.
`40
`In another aspect, representing the second collection of
`overlap image data may include transforming at least Some of
`the second collection of overlap image data Such that the
`characteristic of the second collection of overlap image data
`is represented in a transform domain, and the selected image 45
`data Subset is composed of transformed data.
`In a particular aspect, configuring the first Sub-camera may
`include establishing a first focal length for the first sub-cam
`era, and configuring the second camera may include estab
`lishing a second focal length for the second camera. The 50
`second focallength may be different than the first focal length
`such that the second camera exhibits a different field of view
`as compared to the first camera. Configuring the first camera
`may include providing a first sensing Surface that has a first
`shape, with the first shape being characterized by a first trans- 55
`verse width. The first sensing surface may be oriented for
`receiving the imaged first scene to cause the generating of the
`first set of image data. In this particular aspect, configuring
`the second camera may include providing a second sensing
`Surface that has a second shape that matches the first shape 60
`and has a transverse width that matches the first transverse
`width, and the second sensing Surface may be oriented for
`receiving the imaged second scene to cause the generating of
`the first set of image data. Establishing the first focal length
`may cause the first set of image data to exhibit a first level of 65
`Zoom with respect to the first scene, and establishing the
`second focal length may cause the second set of data to
`
`30
`
`35
`
`4
`exhibit a second level of Zoom with respect to the second
`scene, and the first level of Zoom may be greater than the first
`level of Zoom. In some instances, imaging of the first scene
`may causes the first set of image data to have a first angular
`frequency based at least in part on the first focal length, and
`the imaging of the second scene may cause the second col
`lection of overlap data to have a second angular frequency
`based at least in part on the second focal length, Such that the
`second angular frequency is higher than the first angular
`frequency. In this particular aspect, generating the first set of
`image data may include initially producing an initial set of
`image data and then producing the first set of image data from
`the initial set of image data by upsampling the initial set of
`image data for increasing the angular frequency of the first set
`of image data, as compared to the initial image data, to a target
`angular frequency Such that the first set of image data is
`upsampled image data. The initial set of image data may
`include a group of initial data points, and the upsampling may
`cause the first set of image data to include (i) the group of
`initial data points and (ii) an additional number of data points.
`The upsampling of the initial set of image data may further
`include interpolating between the initial data points for
`assigning values for each of the additional of data points.
`Furthermore, the upsampling can include matching the
`increased angular frequency to the second angular frequency
`Such that the target angular frequency of the first set of image
`data is at least approximately equal to the second angular
`frequency.
`In one embodiment, the first Sub-camera may be config
`ured with a first sensor region having a first sensing Surface,
`and the first sensor region may be aligned such that the imag
`ing of the first scene includes projecting an image of the first
`scene through the first aperture and onto the first sensing
`Surface Such that the first sensor region causes the generating
`of the first set of image data. In this example, the second
`Sub-camera may be furnished with a second sensor region
`having a second sensing Surface, and the second sensor region
`may be aligned such that the imaging of the second scene
`includes projecting an image of the second scene through the
`second aperture and onto the second sensing Surface Such that
`the second sensor region causes the generating of the second
`set of image data. In one aspect of this embodiment, the first
`sensing Surface may have a first shape defined by a first
`Surface area and the second sensing Surface may have a sec
`ond shape that at least generally matches the first shape, and
`the second Surface may have a second Surface area that is at
`least approximately equal to the first Surface area. It is noted
`that the first sensor region and the second sensor region may
`each be a part of a single image sensor.
`In another aspect of this embodiment, the first collection of
`overlap image data may initially be represented based on first,
`second and third data channels, and changing the first collec
`tion of overlap image data may include converting the first
`collection of overlap image data, as represented by the first
`second and third data channels, to represent the first collec
`tion of overlap image databased on a different set of three
`data channels. For example, the first, second, and third chan
`nels may be R, G and B channels, respectively, and the dif
`ferent set of data channels may be Y. U and V channels.
`In yet another aspect of this embodiment, the second col
`lection of overlap image data may be initially based on first,
`second, and third channels, and representing the fused set of
`overlap image data may further include converting the second
`collection of overlap image data (as represented by the first,
`second, and third channels) to represent the second collection
`of overlap databased on a different set of three channels. Each
`of the different channels may serve as one of the plurality of
`
`APPL-1025 / Page 22 of 36
`APPLE INC. v. COREPHOTONICS LTD.
`
`
`
`5
`image data Subsets. For example, the three data channels may
`be R,G, and B channels, and the different set of data channels
`may be Y. U and V channels, and the Y channel may serve as
`the selected Subset of overlap image data.
`In an aspect, generating the first set of image data may
`include initially producing a set of initial image data and then
`producing the first set of image data from the initial image
`data by applying a first forward transformation to at least a
`portion of the initial image data Such that the first set of image
`data may be transformed data in a transform domain such that
`the first set of image data least generally represents, in the
`transform domain, at least Some of the portion of the initial
`image data, and representing the second collection of overlap
`image data may include applying a second forward transfor
`mation to at least some of the second set of image data Such
`that the characteristic of the second collection of image data
`is represented in the transform domain, and at least the
`selected image data Subset is composed of transformed data.
`Changing the first collection of overlap image data may
`include merging the selected one of the image data Subsets
`with the first collection of overlap image data in the transform
`domainto generate a merged data set in the transform domain,
`and producing the fused set of image data may include con
`Verting the merged data set from the transform domain by
`applying thereto at least one of (i) a reverse transformation
`and (ii) an inverse transformation.
`In an additional aspect, producing the fused set of image
`data further may include identifying at least one spatial fea
`ture that is present at a feature position within the first col
`lection of overlap image data of the first set of image data,
`searching for a related representation of at least one identified
`spatial feature (in the selected image data subset) such that
`each related representation at least approximately corre
`sponds to one of the identified features, and (for at least a
`selected one of the related representations that is located in
`the selected image data Subset based on the searching) regis
`tering the selected related representation as being associated
`with the feature position of the corresponding identified fea
`ture. In this additional aspect, changing the first collection of
`overlap image data may include modifying each identified
`40
`spatial feature based on the corresponding related represen
`tation of that feature. It is noted that the related representation
`may have a related feature position within the selected image
`data Subset, and searching for the related representation can
`include finding a spatial shift between the related feature
`position and the feature position. It is further noted that find
`ing the spatial shift may include determining that the spatial
`shift is non-zero and is caused by parallax between the first
`and second Sub-cameras.
`The additional aspect may include (i) defining a reference
`block overlying the feature position and having a shape that
`overlies a reference portion of the first collection of overlap
`image data Such that the reference portion of image data at
`least represents the spatial feature, (ii) defining a search
`region within the selected image data Subset, and (iii) desig
`nating a plurality of candidate blocks within the search
`region, each of which candidate blocks overlies an associated
`portion of the selected image data Subset at a candidate posi
`tion therein. In some instances the searching may include
`determining a degree of correspondence between (i) the ref
`erence portion of data overlaid by the reference block and (ii)
`the portion of data associated with each of the plurality of
`candidate blocks, and in this instance one candidate block
`may be selected based on the degree of correspondence. Such
`that the selected candidate block exhibits the highest degree
`of correspondence as compared to the other candidate blocks.
`Registering the selected related representation may include
`
`25
`
`30
`
`35
`
`45
`
`50
`
`55
`
`60
`
`65
`
`US 8,824,833 B2
`
`5
`
`10
`
`15
`
`6
`associating the candidate position of the selected candidate
`block with the feature position, and modifying of the spatial
`feature may include changing the reference portion of data
`based on at least some of the portion of data associated with
`the selected candidate block. Designating the plurality of
`candidate blocks may include defining a first candidate block
`as a specific one of the plurality of candidate blocks, and a
`second candidate block as a different one of the plurality of
`candidate blocks, such that the first and second candidate
`blocks partially overlap one another.
`In addition to the exemplary aspects and embodiments
`described above, further aspects and embodiments will
`become apparent by reference to the drawings and by study of
`the following descriptions. In addition to the exemplary
`aspects and embodiments described above, further aspects
`and embodiments will become apparent by reference to the
`drawings and by study of the following descriptions.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`Exemplary embodiments are illustrated in referenced fig
`ures of the drawings. It is intended that the embodiments and
`figures disclosed herein are to be illustrative rather than lim
`iting.
`FIG. 1 is a schematic view illustrating fusion of image data
`from two sources.
`FIG. 2A is a diagrammatic view of one embodiment of an
`optical design for a multi-aperture camera.
`FIG. 2B is a diagrammatic view of another embodiment of
`an optical design for a multi-aperture camera.
`FIG. 3 is a block diagram illustrating a multi-aperture
`imaging System.
`FIG. 4 is a block diagram illustrating one embodiment of a
`process for creating full-size images from a multi-aperture
`camera that shares a single sensor.
`FIG. 5 is an exemplary plot illustrating the differences in
`angular frequency information contained in the images pro
`duced from optical Sub-systems having different focal
`lengths, but the same f-number.
`FIG. 6 is a series of exemplary plots, shown here to illus
`trate how differences in angular frequency information from
`images produced from Sub-cameras having different focal
`lengths can be exploited in fusing the images.
`FIG. 7 is a combination block diagram and flow chart
`illustrating optional embodiments of the methods for process
`ing and fusing images from a multi-aperture camera.
`FIG. 8 is a combination block diagram and flow chart
`illustrating other embodiments of methods for processing and
`fusing images fro