Image processing apparatus and method, data processing apparatus and method, and program and recording medium
First Claim
Patent Images
1. An image processing apparatus, comprising:
- a storing device for storing an eigen projective matrix generated by a projective operation from a learning image group including pairs of first quality images and second quality images different in image quality from each other, a first sub-kernel tensor generated so as to satisfy a condition specified by a first setting from a projective kernel tensor generated from the eigen projective matrix and the learning image group and specifying a relationship between the first quality image and an intermediate eigenspace and a relationship between the second quality image and the intermediate eigenspace, and a second sub-kernel tensor generated so as to satisfy a condition specified by a second setting from the projective kernel tensor;
an image division device for dividing an input image of a transformation source into a plurality of divided image areas;
a first sub-tensor projection device for projecting image data in the divided image area divided by the image division device using a first projective operation utilizing the eigen projective matrix and the first sub-kernel tensor, and calculating a coefficient vector in the intermediate eigenspace;
a coefficient vector correction device for generating a corrected coefficient vector corrected in the intermediate eigenspace with respect to at least one divided image area among the plurality of the divided image areas on the basis of the plurality of coefficient vectors acquired by calculation by the first sub-tensor projection device with respect to a part or all of the plurality of divided image areas; and
a second sub-tensor projection device for generating a modified image having a different image quality from that of the input image by projecting a coefficient vector group in the intermediate eigenspace, which includes the corrected coefficient vector generated by the coefficient vector correction device and corresponds to the divided image area, using a second projective operation utilizing the second sub-kernel tensor and the eigen projective matrix.
1 Assignment
0 Petitions
Accused Products
Abstract
The image processing apparatus and method, and the program and the recording medium according to the present invention can make the coefficient vector into high precision by noise elimination or correction utilizing the mutual correlation of the divided image areas in the intermediate eigenspace, and allows relaxation of the input condition and robustness. The high correlation in the divided image areas in the intermediate eigenspace can reduce the divided image areas to be processed, and actualize reduction in processing load and enhancement of the processing speed.
14 Citations
27 Claims
-
1. An image processing apparatus, comprising:
-
a storing device for storing an eigen projective matrix generated by a projective operation from a learning image group including pairs of first quality images and second quality images different in image quality from each other, a first sub-kernel tensor generated so as to satisfy a condition specified by a first setting from a projective kernel tensor generated from the eigen projective matrix and the learning image group and specifying a relationship between the first quality image and an intermediate eigenspace and a relationship between the second quality image and the intermediate eigenspace, and a second sub-kernel tensor generated so as to satisfy a condition specified by a second setting from the projective kernel tensor; an image division device for dividing an input image of a transformation source into a plurality of divided image areas; a first sub-tensor projection device for projecting image data in the divided image area divided by the image division device using a first projective operation utilizing the eigen projective matrix and the first sub-kernel tensor, and calculating a coefficient vector in the intermediate eigenspace; a coefficient vector correction device for generating a corrected coefficient vector corrected in the intermediate eigenspace with respect to at least one divided image area among the plurality of the divided image areas on the basis of the plurality of coefficient vectors acquired by calculation by the first sub-tensor projection device with respect to a part or all of the plurality of divided image areas; and a second sub-tensor projection device for generating a modified image having a different image quality from that of the input image by projecting a coefficient vector group in the intermediate eigenspace, which includes the corrected coefficient vector generated by the coefficient vector correction device and corresponds to the divided image area, using a second projective operation utilizing the second sub-kernel tensor and the eigen projective matrix. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22)
-
-
23. An image processing method, including:
-
a storing step for storing, in a storing device, an eigen projective matrix generated by a projective operation from a learning image group including pairs of first quality images and second quality images different in image quality from each other, a first sub-kernel tensor generated so as to satisfy a condition specified by a first setting from a projective kernel tensor generated from the eigen projective matrix and the learning image group and specifying a relationship between the first quality image and an intermediate eigenspace and a relationship between the second quality image and the intermediate eigenspace, and a second sub-kernel tensor generated so as to satisfy a condition specified by a second setting from the projective kernel tensor; an image dividing step for dividing an input image of a transformation source into a plurality of divided image areas; a first sub-tensor-projecting step for projecting image data in the divided image area divided by the image-dividing using a first projective operation utilizing the eigen projective matrix and the first sub-kernel tensor, and calculating a coefficient vector in the intermediate eigenspace; a coefficient vector correction step for generating a corrected coefficient vector corrected in the intermediate eigenspace with respect to at least one divided image area among the plurality of the divided image areas on the basis of the plurality of coefficient vectors acquired by calculation by the first-sub-tensor-projecting device with respect to a part or all of the plurality of divided image areas; and a second sub-tensor-projecting step for generating a modified image having a different image quality from that of the input image by projecting a coefficient vector group in the intermediate eigenspace, which includes the corrected coefficient vector generated by the coefficient-vector-correcting and corresponds to the divided image area, using a second projective operation utilizing the second sub-kernel tensor and the eigen projective matrix.
-
-
24. A non-transitory recording medium in which computer readable code of a computer program is stored, wherein the computer program causes a computer to function as:
-
a storing device for storing an eigen projective matrix generated by a projective operation from a learning image group including pairs of first quality images and second quality images different in image quality from each other, a first sub-kernel tensor generated so as to satisfy a condition specified by a first setting from a projective kernel tensor generated from the eigen projective matrix and the learning image group and specifying a relationship between the first quality image and an intermediate eigenspace and a relationship between the second quality image and the intermediate eigenspace, and a second sub-kernel tensor generated so as to satisfy a condition specified by a second setting from the projective kernel tensor; an image division device for dividing an input image of a transformation source into a plurality of divided image areas; a first sub-tensor projection device for projecting image data in the divided image area divided by the image division device using a first projective operation utilizing the eigen projective matrix and the first sub-kernel tensor, and calculating a coefficient vector in the intermediate eigenspace; a coefficient vector correction device for generating a corrected coefficient vector corrected in the intermediate eigenspace with respect to at least one divided image area among the plurality of the divided image areas on the basis of the plurality of coefficient vectors acquired by calculation by the first sub-tensor projection device with respect to a part or all of the plurality of divided image areas; and a second sub-tensor projection device for generating a modified image having a different image quality from that of the input image by projecting a coefficient vector group in the intermediate eigenspace, which includes the corrected coefficient vector generated by the coefficient vector correction device and corresponds to the divided image area, using a second projective operation utilizing the second sub-kernel tensor and the eigen projective matrix.
-
-
25. A data processing apparatus, comprising:
-
a storing device for storing an eigen projective matrix generated by a projective operation from a learning data group including pairs of first condition data and second condition data different in condition from each other, a first sub-kernel tensor generated so as to satisfy a condition specified by a first setting from a projective kernel tensor generated from the eigen projective matrix and the learning data group and specifying a relationship between the first condition data and an intermediate eigenspace and a relationship between the second condition data and the intermediate eigenspace, and a second sub-kernel tensor generated so as to satisfy a condition specified by a second setting from the projective kernel tensor; a data division device for dividing input data to be processed into a plurality of divided data areas; a first sub-tensor projection device for projecting data in the divided data area divided by the data division device using a first projective operation utilizing the eigen projective matrix and the first sub-kernel tensor, and calculating a coefficient vector in the intermediate eigenspace; and a coefficient vector correction device for generating a corrected coefficient vector corrected in the intermediate eigenspace with respect to at least one divided data area among the plurality of the divided data areas on the basis of the plurality of coefficient vectors acquired by calculation by the first sub-tensor projection device with respect to a part or all of the plurality of divided data areas.
-
-
26. A data processing method used in a data processing apparatus, the data processing method including:
-
a storing step for storing, in a storing device, an eigen projective matrix generated by a projective operation from a learning data group including pairs of first condition data and second condition data different in condition from each other, a first sub-kernel tensor generated so as to satisfy a condition specified by a first setting from a projective kernel tensor generated from the eigen projective matrix and the learning data group and specifying a relationship between the first condition data and an intermediate eigenspace and a relationship between the second condition data and the intermediate eigenspace, and a second sub-kernel tensor generated so as to satisfy a condition specified by a second setting from the projective kernel tensor; a data dividing step for dividing input data to be processed into a plurality of divided data areas; a first sub-tensor-projecting step for projecting data in the divided data area divided by the data-dividing of a first projective operation utilizing the eigen projective matrix and the first sub-kernel tensor, and calculating a coefficient vector in the intermediate eigenspace; and a coefficient vector correction step for generating a corrected coefficient vector corrected in the intermediate eigenspace with respect to at least one divided data area among the plurality of the divided data areas on the basis of the plurality of coefficient vectors acquired by calculation by the first sub-tensor-projecting with respect to a part or all of the plurality of divided data areas.
-
-
27. A non-transitory recording medium in which computer readable code of a computer program is stored, wherein the computer program causes a computer to function as:
-
a storing device for storing an eigen projective matrix generated by a projective operation from a learning data group including pairs of first condition data and second condition data different in condition from each other, a first sub-kernel tensor generated so as to satisfy a condition specified by a first setting from a projective kernel tensor generated from the eigen projective matrix and the learning data group and specifying a relationship between the first condition data and an intermediate eigenspace and a relationship between the second condition data and the intermediate eigenspace, and a second sub-kernel tensor generated so as to satisfy a condition specified by a second setting from the projective kernel tensor; a data division device for dividing input data to be processed into a plurality of divided data areas; a first sub-tensor projection device for projecting data in the divided data area divided by the data division device using a first projective operation utilizing the eigen projective matrix and the first sub-kernel tensor, and calculating a coefficient vector in the intermediate eigenspace; and a coefficient vector correction device for generating a corrected coefficient vector corrected in the intermediate eigenspace with respect to at least one divided data area among the plurality of the divided data areas on the basis of the plurality of coefficient vectors acquired by calculation by the first sub-tensor projection device with respect to a part or all of the plurality of divided data areas.
-
Specification