Method and system for producing formatted information related to defects of appliances

• US 20050008242A1
• Filed: 06/05/2002
• Published: 01/13/2005
• Est. Priority Date: 07/12/2001
• Status: Active Grant

• Current Assignee: Lens Correction Technologies Corporation
• Original Assignee: DXO Labs
• Inventors: Guichard, Frederic, Liege, Bruno
• Granted Patent: US 7,760,955 B2
• Time in Patent Office: Days
• US Class Current: 382/254
• CPC Class Codes: G06T 1/0007 Image acquisition
• CPC Class Codes: G06T 3/10 Selection of transformation methods according to the characteristics of the input images
• CPC Class Codes: G06T 5/00 Image enhancement or restoration
• CPC Class Codes: G06T 5/70 Denoising; Smoothing
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening
• CPC Class Codes: G06T 5/80 Geometric correction
• CPC Class Codes: H04N 1/00007 relating to particular apparatus or devices
• CPC Class Codes: H04N 1/00045 using a reference pattern designed for the purpose, e.g. a test chart
• CPC Class Codes: H04N 1/00071 characterised by the action taken
• CPC Class Codes: H04N 1/387 Composing, repositioning or otherwise geometrically modifying originals
• CPC Class Codes: H04N 1/40093 Modification of content of picture, e.g. retouching geometric modifications H04N1/387
• CPC Class Codes: H04N 1/58 Edge or detail enhancement; Noise or error suppression, e.g. colour misregistration correction H04N1/62 takes precedence

Abstract

A method and a system for producing formatted information related to defects of appliances of a chain of appliances. To produce the formatted information related to the defects of an image-capture appliance of the chain, the method includes a first calculation algorithm with which there can be chosen, within a set of parameterizable transformation models, within a set of parameterizable reverse transformation models, within a set of synthesis images, within a set of reference scenes, and within a set of transformed images: a reference scene, and/or a transformed image, and/or a parameterizable transformation model, with which a reference image of the reference scene can be transformed to the transformed image, and/or a parameterizable reverse transformation model, with which the transformed image can be transformed to the reference image, and/or a synthesis image obtained from the reference scene and/or obtained from the reference image. The formatted information is at least partly composed of the parameters of the chosen parameterizable transformation model and/or of the parameters of the chosen parameterizable reverse transformation model.

Claims

1-44. -44. (Canceled).

45. A method for producing formatted information related to defects of appliances of an appliance chain, addressed to an image-processor for modifying quality of images processed by the image-processor the appliance chain including at least one image-capture appliance and/or at least one image-restitution means and/or at least one observer, the method comprising:

producing data of the formatted information characterizing the defects of the appliances of the appliance chain.

46. A method according to claim 45, wherein the appliance chain includes at least one image-capture appliance, the method configured to produce the formatted information related to the defects of the image-capture appliance, the method further comprising:

producing data characterizing the defects of the image-capture appliance, the formatted information further including the data characterizing the defects of the image-capture appliance.

47. A method according to claim 46, wherein the appliance chain includes at least one image-capture appliance, to produce the formatted information related to the defects of the image-capture appliance, the method includes, for calculating the formatted information, a first algorithm by which it is possible to choose, within a set of parameterizable transformation models, within a set of parameterizable reverse transformation models, within a set of synthetic images, within a set of reference scenes, and within a set of transformed images:

a reference scene, and/or

a transformed image, and/or

a parameterizable transformation model with which the reference image obtained by capturing the reference scene by the image-capture appliance can be transformed to the transformed image, and/or

a parameterizable reverse transformation model with which the transformed image can be transformed to the reference image, and/or

a synthetic image obtained from the reference scene and/or obtained from the reference image,

such that the transformed image is close to the synthetic image;

the transformed image exhibiting a deviation compared with the synthetic image;

the formatted information being at least partly composed of parameters of the chosen parameterizable transformation model and/or of parameters of the chosen parameterizable reverse transformation model.

48. A method according to claim 47, wherein the first calculation algorithm makes it possible to choose, within a set of mathematical projections, one mathematical projection with which the synthetic image can be constructed from the reference scene.

49. A method according to claim 46, wherein the appliance chain includes at least one image-capture appliance, to produce the formatted information related to the defects of the image-capture appliance, the method includes, for calculating the formatted information, a second algorithm that includes:

choosing at least one reference scene,

capturing at least one reference image of each reference scene by the image-capture appliance,

choosing, within a set of parameterizable transformation models and within a set of synthetic images;

a parameterizable transformation model with which the reference image can be transformed to a transformed image, and/or

a synthetic image obtained from the reference scene and/or obtained from the reference image,

such that the synthetic image is close to the transformed image;

the transformed image exhibiting a deviation compared with the synthetic image;

the formatted information being at least partly composed of parameters of the chosen parameterizable transformation model and/or of parameters of the chosen parameterizable reverse transformation model, the parameterizable reverse transformation model permitting the transformed image to be transformed to the reference image.

50. A method according to claim 49, wherein the second calculation algorithm makes it possible to choose, within a set of mathematical projections, one mathematical projection with which the synthetic image can be constructed from the reference scene.

51. A method according to claim 47, the method including, for calculating the deviations, a third algorithm that includes:

calculating the deviations between the transformed image and the synthetic image, associating the deviations with the formatted information.

52. A method according to claim 46, wherein the appliance chain includes at least one image-capture appliance, the image or images being images of scenes captured by a user by the image-capture appliance;

to produce the formatted information related to the defects of the image-capture appliance, wherein the method further comprises, for calculating the formatted information, a fourth algorithm that includes;

constructing a synthetic-image class by specified mathematical projections of at least one reference scene onto a surface,

capturing at least one reference image of each reference scene by the image-capture appliance,

choosing, within a set of parameterizable transformation models, that with which the reference image can be transformed to a transformed image close to the synthetic-image class of the reference scene;

the transformed image exhibiting a deviation compared with the synthetic-image class, the formatted information being at least partly composed of the parameters of the chosen parameterizable transformation models.

53. A method according to claim 52, the method including, for calculating the deviations, a fifth algorithm that includes:

calculating the deviations between the transformed image and the synthetic-image class,

associating the deviations with the formatted information.

54. A method according to claim 47, wherein in the method the image-capture appliance of the appliance chain is provided with at least one variable characteristic depending on the image, a fraction of the formatted information being related to the defects of the image-capture appliance provided with the variable characteristic or characteristics, each variable characteristic being configured to be associated with a value to form a combination composed of the set of the variable characteristics and of the values;

the method further comprising;

producing a fraction of the formatted information for a predetermined selection of the combinations, the obtained formatted information in the fraction of the formatted information being entered in a field of the formatted information as measured formatted information.

55. A method according to claim 59, the method further comprising:

obtaining the fraction of the formatted information related to an arbitrary combination from measured formatted information, as extended formatted information, the formatted information including the extended formatted information instead of and in place of the measured formatted information.

56. A method according to claim 55, wherein the measured formatted information and the extended formatted information exhibit an interpolation deviation, the method further comprising:

selecting zero or one or more of the variable characteristics, such that the interpolation deviation for the extended formatted information obtained for the selected variable characteristics is below a predetermined interpolation threshold.

57. A method according to claim 45, wherein the appliance chain includes at least one image-restitution means, the method further configured to provide to the image-processor formatted information related to the defects of the image-restitution means;

the method further comprising;

producing data characterizing the defects of the image-restitution means, the formatted information also including the data characterizing the defects of the image-restitution means.

58. A method according to claim 57, wherein the appliance chain includes at least one image-restitution means to produce the formatted information related to the defects of the image-restitution means, the method including, for calculating the formatted information, a sixth algorithm by which it is possible to choose, within a set of parameterizable restitution transformation models, within a set of parameterizable reverse restitution transformation models, within a set of mathematical restitution projections, within a set of restitution references, and within a set of corrected reference restitution images:

a restitution reference, and/or

a corrected reference restitution image, and/or

a parameterizable restitution transformation model with which the restitution reference can be transformed to the corrected reference restitution image, and/or

a parameterizable reverse restitution transformation model with which the corrected reference restitution image can be transformed to the restitution reference, and/or

a mathematical restitution projection with which a synthetic restitution image can be constructed from the corrected reference restitution image,

such that the synthetic restitution image is close to the restituted reference obtained by restitution of the restitution reference by the image-restitution means;

the restituted reference exhibiting a restitution deviation compared with the synthetic restitution image;

the formatted information being at least partly composed of the parameters of the chosen parameterizable restitution transformation model and/or of the parameters of the chosen parameterizable reverse restitution transformation model.

59. A method according to claim 57, wherein the appliance chain includes at least one image-restitution means to produce the formatted information related to the defects of the image-restitution means, the method including, for calculating the formatted information, a seventh algorithm that includes:

choosing at least one restitution reference,

restituting the restitution reference to a restituted reference by the image-restitution means,

choosing, within a set of parameterizable restitution transformation models and within a set of mathematical restitution projections;

a parameterizable restitution transformation model with which the restitution reference can be transformed to a corrected reference restitution image, and

a mathematical restitution projection with which a synthetic restitution image can be constructed from the corrected reference restitution image;

such that the synthetic restitution image is close to the restituted reference;

the restituted reference exhibiting a restitution deviation compared with the synthetic restitution image;

the formatted information being at least partly composed of the parameters of the chosen parameterizable restitution transformation model and/or of the parameters of the parameterizable reverse restitution transformation model, the parameterizable reverse restitution transformation model permitting the corrected reference restitution image to be transformed to the restitution reference.

60. A method according to claim 57, wherein the appliance chain includes at least one image-restitution means, to produce the formatted information related to the defects of the image-restitution means, the method including, for calculating the formatted information, an eighth algorithm that includes:

choosing a corrected reference restitution image,

choosing, within a set of parameterizable restitution transformation models, within a set of mathematical restitution projections, and within a set of restitution references;

a restitution reference, and/or

a parameterizable restitution transformation model with which the restitution reference can be transformed to the corrected reference restitution image, and/or

a parameterizable reverse restitution transformation model with which the corrected reference restitution image can be transformed to the restitution reference, and/or

a mathematical restitution projection with which a synthetic restitution image can be constructed from the corrected reference restitution image,

such that the synthetic restitution image is close to the restituted reference obtained by restitution of the restitution reference by the image-restitution means;

the restituted reference exhibiting a restitution deviation compared with the synthetic restitution image;

the formatted information being at least partly composed of the parameters of the chosen parameterizable restitution transformation model and/or of the parameters of the chosen parameterizable reverse restitution transformation model.

61. A method according to claim 58, the method including, for calculating the restitution deviations, a ninth algorithm that includes:

calculating the restitution deviations between the restituted reference and the synthetic restitution image,

associating the restitution deviations with the formatted information.

62. A method according to claim 58, wherein in the method the image-restitution means of the appliance chain is provided with at least one variable restitution characteristic depending on the image, a fraction of the formatted information being related to the defects of the image-restitution means provided with the variable restitution characteristic or characteristics, each variable restitution characteristic configured to be associated with a value to form a restitution combination composed of the set of the variable restitution characteristics and of the values;

the method further comprising;

producing the fraction of the formatted information for a predetermined selection of the restitution combinations produced entered in the field of the formatted information as measured formatted restitution information.

63. A method according to claim 62, the method further comprising:

obtaining the fraction of the formatted information related to an arbitrary combination from measured formatted restitution information, as extended formatted restitution information, the formatted information including the extended formatted restitution information instead of and in place of the measured formatted restitution information.

64. A method according to claim 63, wherein the measured formatted restitution information and the extended formatted restitution information exhibit a restitution interpolation deviation, the method further comprising:

selecting zero or one or more of variable restitution characteristics, such that the restitution interpolation deviation for the extended formatted restitution information obtained for the selected variable restitution characteristics is below a predetermined restitution interpolation threshold.

65. A method according to claim 58, wherein the image-restitution means is associated with an image-capture appliance to restitute, in digital form, the restituted reference from the restitution reference, wherein in the method to produce the formatted information related to the defects of the said image-restitution means, the formatted information related to the image-capture appliance associated with the image-restitution means is used.

66. A method according to claim 45, wherein the appliance chain includes at least two appliances, to produce formatted information related to the defects of the appliances of the appliance chain, the method further comprising:

producing formatted information related to each of the appliances of the appliance chain,

combining the formatted information related to each of the appliances of the appliance chain, to obtain the formatted information related to the appliance chain.

67. A system for producing formatted information related to defects of appliances of an appliance chain, addressed to an image-processor, for modifying quality of images processed by the image-processor, the appliance chain including at least one image-capture appliance and/or at least one image-restitution means and/or at least one observer, the system comprising:

data-processing means for producing data of formatted information characterizing the defects of the appliances of the appliance chain.

68. A system according to claim 67, wherein the appliance chain includes at least one image-capture appliance, the system configured to produce the formatted information related to defects of the image-capture appliance, the system further comprising:

data-processing means for producing data characterizing the defects of the image-capture appliance, the formatted information also including the data characterizing the defects of the image-capture appliance.

69. A system according to claim 68, wherein the appliance chain includes at least one image-capture appliance, to produce the formatted information related to the defects of the image-capture appliance, the data-processing means configured to employ a first algorithm for calculating the formatted information, the first calculation algorithm making it possible to choose, within a set of parameterizable transformation models, within a set of parameterizable reverse transformation models, within a set of synthetic images, within a set of reference scenes, and within a set of transformed images:

a reference scene, and/or

a transformed image, and/or

a parameterizable transformation model with which the reference image obtained by capturing the reference scene by the image-capture appliance can be transformed to the transformed image, and/or

a parameterizable reverse transformation model with which the transformed image can be transformed to the reference image, and/or

a synthetic image obtained from the reference scene and/or obtained from the reference image,

such that the transformed image is close to the synthetic image;

the transformed image exhibiting a deviation compared with the synthetic image;

the formatted information being at least partly composed of parameters of the chosen parameterizable transformation model and/or of parameters of the chosen parameterizable reverse transformation model.

70. A system according to claim 69, wherein the data-processing means employs the first calculation algorithm including calculating means for choosing, within a set of mathematical projections, one mathematical projection with which the synthetic image can be constructed from the reference scene.

71. A system according to claim 68, wherein the appliance chain includes at least one image-capture appliance, to produce the formatted information related to the defects of the image-capture appliance, the data-processing means employing a second algorithm for calculating the formatted information, the data-processing means employing the second calculation algorithm including calculating means for choosing at least one reference scene, the image-capture appliance making it possible to capture at least one reference image of each reference scene, the calculating means additionally making it possible to make a choice, within a set of parameterizable transformation models and within a set of synthetic images:

of a parameterizable transformation model with which the reference image can be transformed to a transformed image, and/or

of a synthetic image obtained from the reference scene and/or obtained from the reference image,

the choice being made such that the synthetic image is close to the transformed image;

the transformed image exhibiting a deviation compared with the synthetic image;

the formatted information being at least partly composed of the parameters of the chosen parameterizable transformation model and/or of the parameters of the chosen parameterizable reverse transformation model, the parameterizable reverse transformation model permitting the transformed image to be transformed to the said reference image.

72. A system according to claim 71, wherein the calculating means with which the second calculation algorithm can be employed includes calculating means for choosing, within a set of mathematical projections, one mathematical projection with which the synthetic image can be constructed from the reference scene.

73. A system according to claim 69, wherein the data-processing means is configured to employ a third algorithm for calculating the deviations including calculating means for:

calculating the deviations between the transformed image and the synthetic image,

associating the deviations with the formatted information.

74. A system according to claim 68, wherein in the system, to calculate the formatted information related to an image-capture appliance of the appliance chain, the system further comprises at least one reference scene, the image-capture appliance constructing a reference image for each reference scene, to calculate the formatted information related to the image-capture appliance, the system further comprising calculating and processing means for:

constructing a synthetic-image class by specified mathematical projections of at least one reference scene onto a surface,

choosing, within a set of parameterizable transformation models, that with which the reference image can be transformed to a transformed image close to the synthetic-image class of the reference scene;

the transformed image exhibiting a deviation compared with the synthetic-image class, the formatted information being at least partly composed of the parameters of the chosen parameterizable transformation models.

75. A system according to claim 74, wherein the system includes calculating means for:

calculating the deviations between the transformed image and the synthetic-image class,

associating the deviations with the formatted information.

76. A system according to claim 69, wherein in the system the image-capture appliance of the appliance chain is provided with at least one variable characteristic depending on the image, a fraction of the formatted information being related to the defects of the image-capture appliance provided with the variable characteristic or characteristics, each variable characteristic configured to be associated with a value to form a combination composed of the set of the variable characteristics and of the values;

the system further comprising;

data-processing means for producing the fraction of the formatted information for a predetermined selection of the combinations, the produced formatted information in the fraction of the formatted information being entered in a field of the formatted information as measured formatted information.

77. A system according to claim 76, wherein the data-processing means is configured to obtain the fraction of the formatted information related to an arbitrary combination from measured formatted information as extended formatted information, the formatted information including the extended formatted information instead of and in place of the measured formatted information.

78. A system according to claim 77, wherein the measured formatted information and the extended formatted information exhibit an interpolation deviation, the system further comprising:

selection means for selecting zero or one or more of the variable characteristics, such that the interpolation deviation for the extended formatted information obtained for the variable characteristics selected is below a predetermined interpolation threshold.

79. A system according to claim 67, wherein the appliance chain includes at least one image-restitution means, the system further configured to provide, to the image-processor, formatted information related to the defects of the image-restitution means;

the system further comprising;

data-processing means for producing data characterizing the defects of the image-restitution means, the formatted information also including the data characterizing the defects of the image-restitution means.

80. A system according to claim 79, wherein the appliance chain includes at least one image-restitution means to produce the formatted information related to the defects of the image-restitution means, the system further including calculating means for making a choice, within a set of parameterizable restitution transformation models, within a set of parameterizable reverse restitution transformation models, within a set of mathematical restitution projections, within a set of restitution references, and within a set of corrected reference restitution images:

a restitution reference, and/or

a corrected reference restitution image, and/or

a parameterizable restitution transformation model with which the restitution reference can be transformed to the corrected reference restitution image, and/or

a parameterizable reverse restitution transformation model with which the corrected reference restitution image can be transformed to the restitution reference, and/or

a mathematical restitution projection with which a synthetic restitution image can be constructed from the corrected reference restitution image,

the choice being made such that the synthetic restitution image is close to the restituted reference obtained by restitution of the restitution reference by the image-restitution means, the restituted reference exhibiting a restitution deviation compared with the synthetic restitution image, the formatted information being at least partly composed of the parameters of the chosen parameterizable restitution transformation model and/or of the parameters of the chosen parameterizable reverse restitution transformation model.

81. A system according to claim 79, wherein the appliance chain includes at least one image-restitution means to produce the formatted information related to the defects of the image-restitution means, the system further including calculating means for choosing at least one restitution reference, the image-restitution means making it possible to restitute the restitution reference to a restituted reference, the calculating means further making it possible to make a choice within a set of parameterizable restitution transformation models and within a set of mathematical restitution projections:

of a parameterizable restitution transformation model with which the restitution reference can be transformed to a corrected reference restitution image, and

of a mathematical restitution projection with which a synthetic restitution image can be constructed from the corrected reference restitution image;

the choice being made such that the synthetic restitution image is close to the restituted reference, the restituted reference exhibiting a restitution deviation compared with the synthetic restitution image, the formatted information being at least partly composed of the parameters of the chosen parameterizable restitution transformation model and/or of the parameters of the parameterizable reverse restitution transformation model, the parameterizable reverse restitution transformation model permitting the corrected reference restitution image to be transformed to the said restitution reference.

82. A system according to claim 79, wherein the appliance chain includes at least one image-restitution means to produce the formatted information related to the defects of the image-restitution means, the system including calculating means for choosing a corrected reference restitution image, the calculating means further making it possible to make a choice within a set of parameterizable restitution transformation models, within a set of mathematical restitution projections, and within a set of restitution references:

of a parameterizable restitution transformation model with which a specified restitution reference can be transformed to the corrected reference restitution image, and/or

of a parameterizable reverse restitution transformation model with which the corrected reference restitution image can be transformed to the restitution reference, and/or

of a mathematical restitution projection with which a synthetic restitution image can be constructed from the corrected reference restitution image,

the choice being made such that the synthetic restitution image is close to the restituted reference obtained by restitution of the restitution reference by the image-restitution means, the restituted reference exhibiting a restitution deviation compared with the synthetic restitution image, the formatted information being at least partly composed of the parameters of the chosen parameterizable restitution transformation model and/or of the parameters of the chosen parameterizable reverse restitution transformation model.

83. A system according to claim 80, the system further comprising calculating means for:

calculating the restitution deviations between the restituted reference and the synthetic restitution image,

associating the restitution deviations with the formatted information.

84. A system according to claim 80, wherein in the system the image-restitution means of the appliance chain is provided with at least one variable restitution characteristic depending on the image, a fraction of the formatted information being related to the defects of the image-restitution means provided with the variable restitution characteristic or characteristics, each variable restitution characteristic configured to be associated with a value to form a restitution combination composed of the set of the variable restitution characteristics and of the values, the system further comprising:

data-processing means for producing the fraction of the formatted information for a predetermined selection of the restitution combinations, the produced formatted information in the fraction and entered in a field of the formatted information as the measured formatted restitution information.

85. A system according to claim 84, the system further comprising:

data-processing means for obtaining the fraction of the formatted information related to an arbitrary combination from measured formatted restitution information, the fraction of the formatted information related to an arbitrary restitution combination as extended formatted restitution information, the formatted information including the extended formatted restitution information instead of and in place of the measured formatted restitution information.

86. A system according to claim 85, wherein the measured formatted restitution information and the extended formatted restitution information exhibit a restitution interpolation deviation, the system further comprising:

selection means for selecting zero or one or more variable restitution characteristics, such that the restitution interpolation deviation for the extended formatted restitution information obtained for the variable restitution characteristics selected is below a predetermined restitution interpolation threshold.

87. A system according to claim 80, the restitution means being associated with an image-capture appliance to restitute, in digital form, the restituted reference from the restitution reference, the system further comprising:

data-processing means for producing the formatted information related to the defects of the restitution means, by using the formatted information related to the image-capture appliance associated with the restitution means.

88. A system according to claim 67, wherein the appliance chain includes at least two appliances to produce formatted information related to the defects of the appliances of the appliance chain, the system further comprising:

data-processing means for producing the formatted information related to each of the appliances of the appliance chain and for combining the formatted information related to each of the appliances of the appliance chain, to obtain the formatted information related to the appliance chain.