Electrical computers and digital data processing systems: input/ – Input/output data processing
Reexamination Certificate
2001-11-21
2004-11-16
Peyton, Tammara (Department: 2182)
Electrical computers and digital data processing systems: input/
Input/output data processing
C710S002000, C382S203000, C382S227000, C382S274000, C382S276000, C345S419000, C345S698000
Reexamination Certificate
active
06820137
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a computer-readable recording medium storing a resolution converting program, a resolution converting device and a resolution converting method, and more particularly to a computer-readable recording medium storing a resolution converting program that converts resolution based on a model generated through statistical processing of multiple images, as well as a resolution converting device and a resolution converting method.
DESCRIPTION OF THE RELATED ART
In the conventional art, when image data is transmitted from one device to another, if the resolution of images that can be processed by the latter device is different from the resolution of images generated by the former device, it is necessary to convert the resolution. In such a case, the resolution is converted by performing interpolation such as linear or tertiary interpolation regarding the image data.
Japanese Laid-Open Patent Application Hei 7-152907 discloses, as another method of resolution conversion, a method in which the wavelet component of the input image is predicted, and compensation is made for the high frequency component of an enlarged image with a higher resolution.
However, where interpolation such as linear or tertiary interpolation is performed regarding an image, because only the resolution of the image before enlargement or the resolution of the image that has undergone smoothing but not enlargement is available, the problem exists that the contours of the enlarged image are blurry or exhibit noticeable roughness.
According to the method disclosed by Japanese Laid-Open Patent Application Hei 7-152907, it is necessary to predict the wavelet component for a higher resolution image. When predicting the wavelet component, the results of a prescribed learning process must be used, giving rise to the problem that the results of the learning process affect image quality.
SUMMARY OF THE INVENTION
The present invention was created in order to resolve these problems, and an aspect thereof is to provide a computer-readable recording medium storing a resolution converting program, a resolution converting device and a resolution converting method that can improve the quality of the resolution-converted image.
Another aspect of the present invention is to provide a computer-readable recording medium storing a resolution converting program, a resolution converting device and a resolution converting method that can easily convert the resolution of an image.
According to one embodiment of the present invention, the present invention comprises a computer-readable recording medium storing a resolution converting program that converts resolution based on a model generated through statistical processing of multiple images, wherein the program is implemented using a computer and includes: a step for preparing multiple models having different resolutions; a step for receiving an input image; a step for detecting the resolution of the input image; a step for projecting the input image onto a model selected in accordance with the detected resolution in order to obtain characteristic parameters of the input image; and a step for generating an image by applying the obtained characteristic parameters to a model different from the model used to obtain the characteristic parameters.
According to this embodiment, multiple models having different resolutions are prepared. The input image is projected onto a model corresponding to the resolution of the input image, and the characteristic parameters of the input image are obtained. Using the characteristic parameters thereby obtained, an image is generated using a model different from the model used to obtain the characteristic parameters. Consequently, a computer-readable recording medium storing a resolution converting program capable of improving the quality of the resolution-converted image can be provided.
Preferably, the step for preparing multiple models having different resolutions has a step for converting the resolution of a model generated through statistical processing.
According to this embodiment, the multiple models having different resolutions are generated by converting the resolution of a model generated through statistical processing. Among the models having different resolutions, the images generated based on the same characteristic parameters differ from each other only in terms of resolution. Therefore, because multiple resolution-converted models may be obtained from one model, a computer-readable recording medium storing a resolution converting program that can easily change the resolution of an image can be provided.
Preferably, the step for preparing multiple models having different resolutions has a step for converting the resolutions of multiple images and generating a model by performing statistical processing regarding each image having a different resolution.
According to this embodiment, the multiple models having different resolutions are generated by converting the resolution of each of the multiple images and performing statistical processing regarding each image having a different resolution. Therefore, because multiple models having different resolutions and generated from multiple images are used, a computer-readable recording medium storing a resolution converting program that can easily change the resolution of an image can be provided.
Preferably, the image generation step generates an image using a model having a higher resolution than the detected resolution.
According to this invention, because an image is generated based on a model having a higher resolution than the detected resolution, the quality of the image can be improved when the resolution thereof is increased.
Preferably, the multiple images are images of a person's face.
According to this invention, because the multiple images are images of a person's face, the quality of the resolution-converted image of a face can be improved.
Preferably, a step for extracting the face area from the input image, a step for converting the resolution of the input image, and a step for synthesizing the resolution-converted image and the image generated in the image generation step are further included.
According to this invention, the face area is extracted from the input image and the resolution of the input image is converted. The resolution-changed image and the image generated through the image generation step are then synthesized. Therefore, where a person's face is included in the input image, the quality of the image of the person's face may be improved.
According to another aspect of the present invention, the resolution converting device includes means for preparing multiple models having different resolutions; means for receiving an input image; means for detecting the resolution of said input image; means for projecting said input image onto a model selected in accordance with said detected resolution in order to obtain characteristic parameters of said input image; and means for generating an image by applying said obtained characteristic parameters to a model different from said model used to obtain the characteristic parameters.
According to this invention, multiple models having different resolutions are prepared. The input image is projected onto a model corresponding to the resolution of the input image, and the characteristic parameters of the input image are obtained. Using the thus obtained characteristic parameters, an image is generated based on a model different from the model used to obtain the characteristic parameters. Therefore, a resolution converting device that can improve the quality of the resolution-converted image can be provided.
According to yet another aspect of the present invention, the present invention comprises a resolution converting method that converts resolution using a model generated through statistical processing of multiple images and includes a step for preparing multiple models having different resolutions; a step for receiving an input image; a step for detecting the res
Hasegawa Hiromu
Nakamura Kyoko
Minolta Co. , Ltd.
Peyton Tammara
LandOfFree
Computer-readable recording medium storing resolution... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Computer-readable recording medium storing resolution..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Computer-readable recording medium storing resolution... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3338934