Recursive image encoding/decoding using interpolation within variably sub-divided picture areas
AssigneeBritish Telecommunications public limited company
InventorsHugh Spriggs, Charles Nightingale, Roger D. Turkington
Life of this patent
Estimating the term…
Abstract
Actual picture points of a picture area are compared with interpolated values derived from selected points. If the differences are small, data for only the selected points are transmitted, otherwise the area is subdivided and each sub-area processed in the same way, so that the number of points selected for transmission is greatest in detailed areas of the image.
Claims29
A method of image transmission comprising repetitively:
- (a) generating an estimate of a picture area by interpolation from selected actual samples thereof to provide estimated samples;
- (b) comparing the estimated samples with the actual samples;
- and
- (c) if the estimated and actual samples differ by less than a threshold criterion, transmitting picture data representing the said actual samples, and if they do not, notionally dividing the area under consideration into two or more sub-areas and subjecting each sub-area to steps (a), (b) and (c) until a minimum sub-area size is reached.
An apparatus for carrying out a method of image transmission comprising:
- an interpolation means for generating an estimate of a picture area based on interpolation from a subset of its actual samples,
- means for comparing the estimate with actual sample values, and
- means arranged in dependence on whether the comparison indicates a difference of more or less than a threshold (a) to elect the subset of actual samples for transmission or (b) to notionally divide the picture area into two or more sub-areas and to process each sub-area in like manner.
A decoder for receiving an image encoded by repetitively (a) generating an estimate of a picture area by interpolation from selected actual samples thereof to provide estimated samples, (b) comparing the estimated samples with the actual samples, and (c) if the estimated and actual samples differ by less than a threshold criterion, transmitting picture data representing the said actual samples, and if they do not, notionally dividing the area under consideration into two or more sub-areas and subjecting each sub-area to steps (a), (b) and (c) until a minimum sub-area size is reached , said decoder comprising:
- a frame store,
- processing means arranged to compute from received division codes the addresses of image points for which data is transmitted and to interpolate values for the remaining points.
A method for transmitting an image having a plurality of picture elements, the method comprising the steps of repetitively:
- (a) receiving actual sample values representing respective picture elements;
- (b) generating an estimate of a picture area by interpolation from selected ones of said actual sample values for that area to produce estimated sample values for the remainder of the picture elements of that area;
- (c) comparing said estimated sample values with the actual sample values corresponding to the same elements to ascertain whether they meet a threshold criterion requiring that the differences between estimated and actual values be small;
- and
- (d) if the criterion is met, transmitting picture data indicative of the values of said selected ones of said actual sample values, and if the criterion is not met and the picture area under consideration has not reached a minimum sub-area size, notionally dividing said area into two or more sub-areas, information as to such division being transmitted, and subjecting each sub-area to steps (b), (c) and (d).
An apparatus for transmission of an image including a plurality of picture elements, said apparatus:
- (a) interpolation means for receiving actual sample values representing respective picture elements and generating an estimate of a picture area by interpolation from selected ones of said actual sample values for that area to produce estimated sample values for the remainder of the picture elements of that area;
- (b) means for comparing said estimated sample values with the actual sample values corresponding to the same elements to ascertain whether they meet a threshold criterion requiring that the differences between estimated and actual values be small;
- and
- (c) means arranged in dependence on whether the criterion respectively is or is not met (i) to transmit picture data indicative of the values of said selected ones of said actual sample values, or (ii) subject to a minimum sub-area size, to notionally divide said area into two or more sub-areas and process each sub-area in like manner.
An image decoding apparatus comprising:
- input means for receiving division codes indicating whether an image area and sub-areas of that area are divided, or further divided, into sub-areas, and for receiving actual sample values representing selected picture elements of the image;
- frame store means;
- processing means arranged to compute, from the received division codes, the addresses within the frame store corresponding to the received actual sample values and to store those values in the frame store;
- and
- interpolation means arranged to interpolate, from the received values, estimated sample values for the remaining elements of the image and to store them in the frame store.
A method for encoding and transmitting condensed image data representing a full multi-dimensional array of pixels having individual actual values and address-defined relative locations, said method comprising the steps of:
- (a) dividing said array into groups of contiguous pixels and selecting certain pixels of the array, said selected pixels and the sizes of said groups being chosen such that the pixels of each group have actual values which can be derived within a predetermined error threshold by interpolation from the actual values of ones of said selected pixels ;and
- (b) transmitting as said condensed image data information representing the actual values and addresses of said selected pixels.
A method for decoding compressed image data representing a full multi-dimensional array of pixels having individual actual values and address-defined relative locations, said method comprising the steps of:
- (a) identifying from said compressed image data the actual values and address locations of a predetermined first plurality of pixels in said array located to define differently sized contiguous groups of pixels having actual values which can be derived within a predetermined error threshold by interpolation from the actual values of said first plurality of pixels;
- and
- (b) deriving by interpolation from the values and relative locations of said first plurality of pixels approximated values for the remaining plurality of pixels in said array.
Dependencies are inferred from the claim text.