Image processing device and method, and transmission medium, transmission method and image format

• US 6,463,178 B1
• Filed: 02/12/1999
• Issued: 10/08/2002
• Est. Priority Date: 06/16/1997
• Status: Expired due to Fees

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Okuwaki, Tomonori, Kondo, Tetsujiro
• Primary Examiner(s): Moore, David
• Assistant Examiner(s): Do, Anh Hong
• Application Number:
  US09/242,348
• Time in Patent Office: 1,334 Days
• Field of Search: 382/232, 382/256, 382/238, 382/199, 382/125, 348/458, 348/556, 386/112
• US Class Current: 382/232
• CPC Class Codes: H04N 19/587 involving temporal sub-sampling or interpolation, e.g. decimation or subsequent interpolation of pictures in a video sequence
• CPC Class Codes: H04N 19/59 involving spatial sub-sampling or interpolation, e.g. alteration of picture size or resolution
• CPC Class Codes: H04N 19/98 Adaptive-dynamic-range coding [ADRC]

Abstract

Disclosed is an image processing apparatus 10D for processing image data that composes a dynamic image, wherein pixels are thinned for each frame with respect to the pixels of the respective frames that compose a dynamic image, in such a manner that they become the chessboard-like lattice format in both directions of the spatial direction and the temporal direction, so that the quantity of information can be reduced, maintaining the resolutions of horizontal, vertical and oblique directions of the dynamic image.

Claims

1. An image processing apparatus for processing a thinned data that has been produced by thinning the pixels of the image data that composes a dynamic image, said image processing apparatus comprises:

receiving means for receiving a thinned image data that has been produced by thinning the pixels of the respective frames that compose a dynamic image, in such a manner that they become the cheeseboard-like lattice format in both directions of the spatial direction and the temporal direction; and

restoring means for producing the negated pixels, with respect to said thinned image data, and restoring the original image data;

said restoring means comprising;

determining means for determining a stated class that represents the property of the attentional negated pixel of said thinned image data; and

generating means for predicting the original pixel in response to said determined class, and for generating the original pixel data.

2. The image processing apparatus as claimed in claim 1, in which

said generating means includes a memory for storing a predictive data, for each class, that has been previously produced for each class on the basis of learning by the use of the image data that composes a dynamic image in a standard manner, reads out a predictive data that corresponds to said determined class, and generates the original pixel data on the basis of the predictive data that has been read out.

3. The image processing apparatus as claimed in claim 2, in which

said memory stores a set of predictive data for each class as said predictive data;

and

said generating means reads out a predictive coefficient set that corresponds to said determined class, and generates the original pixel data on the basis of the read predictive set and the plural bit-thinned pixels that are located at the periphery of said attentional bit-thinned pixel.

4. The image processing apparatus as claimed in claim 2, in which

said memory stores a pixel value for each class as said predictive data;

and

said generating means reads out a pixel value that corresponds to said determined class, and outputs the pixel value as the original pixel data.

5. The image processing apparatus as claimed in claim 1, in which

said determining means determines a class with respect to said attentional bit-thinned pixel, using the bit-thinned pixels that have been located in the spatial and/or temporal directions.

6. An image processing apparatus for processing a thinned data that has been produced by thinning the pixels of the image data that composes a dynamic image, said image processing apparatus comprises:

receiving means for receiving a modified thinned image data that has been obtained by further performing bit-thinning of the direction of the level of the pixel, with respect to each pixel of the thinned image data that has been produced by thinning the pixels of the respective frames that compose a dynamic image, in such a manner that they become the cheeseboard-like lattice format in both directions of the spatial direction and the temporal direction; and

a restoring means for producing the original pixel with respect to each bit-thinned pixel of said modified thinned image data, thereby producing the thinned image data, and for restoring the original image data from the thinned pixels of said thinned image data;

said storing means comprising;

first determining means for determining a stated class that represents the property of an attentional bit-thinned pixel of said bit-thinned image data;

first generating means for predicting the original pixel in response to said determined class, thereby generating the original pixel data;

second determining means for determining a stated class that represents the property of the attentional negated pixel of the thinned image data that is composed of said pixel data which has been generated by said first generating means; and

second generating means for predicting the original pixel in response to said determined class, and for generating the original pixel data.

7. The image processing apparatus as claimed in claim 6, in which

said first generating means includes a memory for storing a predictive data, for each class, that has been previously produced for each class on the basis of learning by the use of the image data that composes a dynamic image in a standard manner, reads out the predictive data that corresponds to the class that has been determined by said first determining means, and generates the original pixel data on the basis of the very predictive data that has been read out; and

each of said second generating means includes a memory for storing a predictive data, for each class, that has been previously produced for each class on the basis of learning by the use of the image data that composes a dynamic image in a standard manner, reads out the predictive data that corresponds to the class that has been determined by said second determining means, and generates the original pixel data on the basis of the very predictive data that has been read out.

8. The image processing apparatus as claimed in claim 7, in which

the memory of said first generating means stores a predictive coefficient set for each class as said predictive data;

said first generating means reads out a predictive coefficient set that corresponds to said determined class, and generates the original pixel data on the basis of the read predictive set and the plural bit-thinned pixels that are located at the periphery of said attentional bit-thinned pixel;

the memory of said second generating means stores a predictive coefficient set for each class as said predictive data; and

said second generating means reads out a predictive coefficient set that corresponds to said determined class, and generates the original pixel data on the basis of the read predictive set and the plural bit-thinned pixels that are located at the periphery of said attentional negated pixel.

9. The image processing apparatus as claimed in claim 7, in which

the memory of said first generating means stores a pixel value for each class as said predictive data;

said first generating means reads out a pixel value that corresponds to said determined class, and outputs the pixel value as the original pixel data;

the memory of said second generating means stores a pixel value for each class as said predictive data; and

said second generating means reads out a pixel value that corresponds to said determined class, and outputs the pixel value as the original pixel data.

10. The image processing apparatus as claimed in claim 6, in which

said first determining means determines a class with respect to said attentional bit-thinned pixel, using the bit-thinned pixels that have been located in the spatial and/or temporal directions;

and

said second determining means determines a class with respect to said attentional negated pixel, using the bit-thinned pixels that have been located in the spatial and/or temporal directions.