Methods and systems for content processing
Mobile phones and other portable devices are equipped with a variety of technologies by which existing functionality can be improved, and new functionality can be provided. Some aspects relate to visual search capabilities, and determining appropriate actions responsive to different image inputs. Others relate to processing of image data. Still others concern metadata generation, processing, and representation. Yet others concern user interface improvements. Other aspects relate to imaging architectures, in which a mobile phone's image sensor is one in a chain of stages that successively act on packetized instructions/data, to capture and later process imagery. Still other aspects relate to distribution of processing tasks between the mobile device and remote resources (“the cloud”). Elemental image processing (e.g., simple filtering and edge detection) can be performed on the mobile phone, while other operations can be referred out to remote service providers. The remote service providers can be selected using techniques such as reverse auctions, through which they compete for processing tasks. A great number of other features and arrangements are also detailed.
1. An image processing method comprising the acts:
(a) receiving a set of pixel data, each pixel datum representing an image scene sampled at a location;
(b) processing the set of received pixel data in a first manner to produce first results, and presenting the first results on a device display, said first manner of processing including processing the received pixel data by a first operation comprising white balance correction or Bayer interpolation de-mosaicing;
(c) processing the set of received pixel data in a second manner to produce second results, and presenting the second results on the device display, said second manner of processing not including processing the received pixel data by said first operation;
wherein the results presented on the device display are both based on said set of received pixel data, but the first results are produced using said first operation, and the second results are produced without using said first operation.
2. The method of claim 1 that includes:
presenting the pixel data, processed with said first operation, on the display screen as the first result; and
presenting information obtained from a remote source, based on identification of a subject in the image scene, on the display screen as the second result.
3. The method of claim 1 in which act (c) includes performing a recognition operation on the received set of pixel data.
4. The method of claim 3 in which act (c) includes identifying a subject recognized in said image scene, and presenting information about said identified subject on the device display.
5. The method of claim 3 in which the recognition operation is performed on a barcode or a digital watermark and yields payload data, and the second results include information accessed from a remote source using the payload data.
6. The method of claim 1 that includes presenting the first and second results on the device display at different times.
7. The method of claim 1 that includes presenting the first results on the device display, and thereafter presenting the second results on the device display.
8. The method of claim 1 in which act (b) includes processing the received image data by a white balance correction operation.
9. The method of claim 1 in which act (b) includes processing the received image data by a Bayer interpolation de-mosaicing operation.
10. A system including:
an image sensor including an array of photodiodes on which light can impinge from a scene to produce a set of analog electrical signals, and D/A converters operative to convert said set of signals into digital form, yielding pixel data representing the scene;
a display;
a white balance correction module;
a first image processing chain adapted to render a depiction of the scene on the display, the first image processing chain including the white balance correction module;
means for processing pixel data from the sensor to yield processed results for presentation on the display, said means not including the white balance correction module;
wherein the system is adapted to display the depiction of the scene and the processed results on the display, but the depiction of the scene is produced using white balance correction, and the processed results are produced without using white balance correction.
11. A method employing a system operated by a user, said system including a local processor, an RF transceiver, a camera and a display, the method comprising processing imagery captured by the system camera to identify a type of subject depicted in the imagery, routing corresponding 2D keyvector data for processing by a remote processor associated with the identified type of subject, and making associated result information available to the user on the display, the method more particularly comprising the acts:
type-processing first data, corresponding to first imagery captured by the system camera depicting a first subject, said processing including a processor identifying the depicted first subject as being of a first type;
based at least in part on said identification of the depicted first subject as being of the first type, routing first 2D keyvector data corresponding to the captured first imagery for processing by a first remote processor, said routing being performed without then being directed by the user;
receiving first result data back in response to said processing of the first 2D keyvector data by said first remote processor, and making information corresponding to the first result data available to the user on the system display;
type-processing second data, corresponding to second imagery captured by the system camera depicting a second subject, said processing including a processor identifying the depicted second subject as being of a second type, said second type being different than said first type;
based at least in part on said identification of the depicted second subject as being of the second type, routing second 2D keyvector data corresponding to the captured second imagery for processing by a second remote processor, said routing being performed without then being directed by the user, said second remote processor being different than said first remote processor;
receiving second result data back in response to said processing of the second 2D keyvector data by said second remote processor, and making information corresponding to the second result data available to the user on the system display;
wherein 2D keyvector data corresponding to first and second imagery is routed to first and second different processors, respectively, based at least in part on types of subjects in the first and second imagery, and corresponding result information is made available to the user on the system display.
12. A method employing a device equipped with a camera, a display, and an RF transceiver, comprising the acts:
processing a first image captured by said device to detect first feature or attribute information;
selecting a first processor remote from the device, after said processing of the first image, based on said first feature or attribute information detected from the first image;
routing keyvector information corresponding to the first image to said first processor for processing;
making information resulting from said processing by the first processor available on the display of the device;
processing a second image captured by said device to detect second feature or attribute information, without reference to any barcode or watermark information;
selecting a second processor remote from the device, after said processing of the second image, based on said second feature or attribute information detected from the second image said second processor being distinct from the first processor;
routing information corresponding to the second image to said second processor for processing; and
making information resulting from said processing by the second processor available on the display of the device.
13. The method of claim 12 in which said keyvector data comprises information derived from pixel data, rather than pixel data itself.
14. The method of claim 12 in which the first processor is adapted to perform a first service from the list: a barcode reading service, an optical character recognition service, a watermark reading service, or a facial recognition service, and the second processor is adapted to perform an optical character recognition service or a facial recognition service.
15. The method of claim 12 in which:
said processing the first image to detect first feature or attribute information includes applying a transformation operation to the first image, and analyzing information resulting from said transformation operation to detect telltale signs associated with an image of a barcode, text, a digital watermark or a face.
16. The method of claim 12 in which said processing the first image to detect first feature or attribute information includes applying a Fourier, Fourier-Mellin, DCT, wavelet, Gabor, Laplacian, Wiener, Sobel or signum transform operation on information corresponding to the first image.
17. The method of claim 12 that includes determining, from said feature or attribute information detected from the first image, that the first image likely depicts a subject of a first type from the list: barcode, text, watermark or face, and routing keyvector data corresponding to the first image to the first remote processor as a consequence, wherein said first remote processor is adapted to process keyvector data corresponding to subjects of said first type.
18. The method of claim 17 that includes determining, from said feature or attribute information detected from the second image, that the second image likely depicts a subject of a second type, where the second type is different than said first type, and routing information corresponding to the second image to the second remote processor as a consequence, wherein said second remote processor is adapted to process information corresponding to subjects of said second type.
19. The method of claim 17 that includes performing said routing act without user request after the first image has been captured.
20. The method of claim 12 that includes training a system to identify images as likely containing a particular type of image subject, said training including presenting the system with plural training images that are known to include said particular type of subject, and extracting a feature set about shared similarities from said training images, wherein said trained system is employed to determine an attribute of the first image, on which selection of the first remote processor is based.
21. The method of claim 12 that includes training a system to identify images as likely containing a particular type of image subject, said training including presenting the system with plural training images that are known to include said type of subject, processing said images, and examining resulting attribute data to identify bounded attribute ranges that can be used to identify future images that likely contain such type of image subject, wherein said trained system is employed to determine an attribute of the first image, on which selection of the first remote processor is based.
22. The method of claim 12 that includes selecting said first remote processor based on said feature or attribute information of the first image, and based on user demographic or preference information, or user location.
23. The method of claim 12 that includes routing to said first and second remote processors through a common hub, remote from said device, wherein said hub also selects said first and second remote processors.
24. The method of claim 12 that includes determining, by pattern matching, that the second image depicts a subject of a particular type, and selecting the second remote processor based on said determining.
25. A system comprising:
an RF transceiver;
a camera;
a display; and
first means for responding to camera-captured first and second images depicting subjects of first and second different types by automatically communicating with first and second different remote processors, respectively, and making information resulting from such communications available on the system display, wherein said first means includes classifying means for identifying, from among plural possible classes, a class to which said subject of the first type likely belongs, without reference to any express machine-readable information depicted in the first image.
26. The system of claim 25 in which said classifying means further includes means for identifying, from among plural possible classes, a class to which said subject of the second type likely belongs, by reference to depiction of express machine-readable information in the second image.
27. The system of claim 25 in which said first means further comprises hub means for routing 2 D keyvector data derived from the camera-captured image depicting the subject of the first type, to a remote processor that is selected in accordance with said identified class.
28. The system of claim 25 in which said classifying means employs a feature set earlier learned, by a training process including presentation of plural images that are known to depict subjects of a particular class, to identify the class to which said subject of the first type likely belongs.
29. The system of claim 25 in which said classifying means includes transformation means for applying a transformation operation to the camera-captured first image depicting the subject of the first type, and analyzing information resulting from said transformation operation to detect telltale signs associated with one of plural different classes.
30. The system of claim 29 in which said transformation operation comprises a Fourier, Fourier-Mellin, DCT, wavelet, Gabor, Laplacian, Wiener, Sobel or signum transform.
31. The system of claim 25 in which in which said classifying means employs image metrics comprising color, texture, shape, or keypoint descriptors, in said classifying.
32. The system of claim 25 in which said subject of the first type is a landmark, a consumer product, a device equipped to communicate by wireless link, a face, or a vegetable.
33. The method of claim 11 in which said routing of first or second 2D keyvector data is performed by a processor remote from the system operated by the user, and distinct from the first and second processors.
34. The method of claim 11 in which said routing of first or second 2D keyvector data is performed by a processor remote from the system operated by the user, and remote from the first and second processors.
35. The method of claim 11 in which said type-processing the first data is performed by the local processor.
36. The method of claim 11 that includes selecting a remote processor to serve as said first remote processor, from among plural remote processors, based at least in part on said identification of the depicted first subject as being of the first type.
37. The method of claim 11 in which the first 2D keyvector data comprises information derived from pixel data, rather than pixel data itself.
38. The method of claim 37 in which the captured first imagery comprises millions of bytes of information, and the first 2D keyvector data is smaller by a factor of at least a thousand.
39. The method of claim 11 that further includes the act of processing the first 2D keyvector data by the first remote processor, said processing performed by the first remote processor.
40. The method of claim 12 that includes employing a trained system to determine an attribute of the first image, on which selection of the first remote processor is based, said system having been previously trained to identify images as likely containing a particular type of subject by presenting said system with plural training images that are known to include said particular type of subject.
41. The method of claim 12 in which said processing of the first image to detect first feature or attribute information, on which selection of a first remote processor is based, is performed by a processor in said device.
42. The system of claim 25 comprising a device including said transceiver, camera, display, and a processor, wherein said classifying means comprises the processor configured to identify, from among said plural possible classes, the class to which said subject of the first type likely belongs.
43. The method of claim 11 in which said type-processing of the first data includes identifying, from among plural possible classes, a class to which said first subject likely belongs, without reference to machine-readable information depicted in the first imagery.
44. The method of claim 43 in which said identifying employs a feature set earlier learned, by a training process including presentation of plural images that are known to depict subjects of a particular class, to identify the class to which said first subject likely belongs.
45. The method of claim 11 in which said type-processing of the first data includes transforming the first data by a Fourier, Fourier-Mellin, DCT, wavelet, Gabor, Laplacian, Wiener, Sobel or signum transform operation, and analyzing information resulting from said transforming to detect telltale signs associated with one of plural different classes.
46. The method of claim 11 in which said type-processing of the first data employs an image metric comprising color, texture, shape, or keypoint descriptors.
47. The method of claim 11 in which said first type of subject is a landmark, a consumer product, a device equipped to communicate by wireless link, a face, or a vegetable.