IMAGE ENCODING-DECODING SYSTEM AND RELATED TECHNIQUES

• US 20070223582A1
• Filed: 01/04/2007
• Published: 09/27/2007
• Est. Priority Date: 01/05/2006
• Status: Abandoned Application

• Current Assignee: British Broadcasting Corporation
• Original Assignee: British Broadcasting Corporation
• Inventors: Borer, Timothy
• Application Number:
  US11/619,896
• Publication Number:
  US 20070223582A1
• Time in Patent Office: Days
• US Class Current: 375/240.120
• CPC Class Codes: H04N 19/105 Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction
• CPC Class Codes: H04N 19/119 Adaptive subdivision aspects, e.g. subdivision of a picture into rectangular or non-rectangular coding blocks
• CPC Class Codes: H04N 19/12 Selection from among a plurality of transforms or standards, e.g. selection between discrete cosine transform [DCT] and sub-band transform or selection between H.263 and H.264
• CPC Class Codes: H04N 19/122 Selection of transform size, e.g. 8x8 or 2x4x8 DCT; Selection of sub-band transforms of varying structure or type
• CPC Class Codes: H04N 19/124 Quantisation
• CPC Class Codes: H04N 19/13 Adaptive entropy coding, e.g. adaptive variable length coding [AVLC] or context adaptive binary arithmetic coding [CABAC]
• CPC Class Codes: H04N 19/132 Sampling, masking or truncation of coding units, e.g. adaptive resampling, frame skipping, frame interpolation or high-frequency transform coefficient masking
• CPC Class Codes: H04N 19/147 according to rate distortion criteria rate-distortion as a criterion for motion estimation H04N19/567
• CPC Class Codes: H04N 19/169 characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
• CPC Class Codes: H04N 19/172 the region being a picture, frame or field
• CPC Class Codes: H04N 19/18 the unit being a set of transform coefficients
• CPC Class Codes: H04N 19/187 the unit being a scalable video layer
• CPC Class Codes: H04N 19/19 using optimisation based on Lagrange multipliers
• CPC Class Codes: H04N 19/33 in the spatial domain
• CPC Class Codes: H04N 19/52 by predictive encoding
• CPC Class Codes: H04N 19/521 for estimating the reliability of the determined motion vectors or motion vector field, e.g. for smoothing the motion vector field or for correcting motion vectors
• CPC Class Codes: H04N 19/53 Multi-resolution motion estimation; Hierarchical motion estimation
• CPC Class Codes: H04N 19/583 Motion compensation with overlapping blocks
• CPC Class Codes: H04N 19/61 in combination with predictive coding
• CPC Class Codes: H04N 19/619 the transform being operated outside the prediction loop
• CPC Class Codes: H04N 19/63 : using sub-band based transform, e.g. wavelets

Abstract

An image encoding-decoding system and related methods of operation includes an encoder for encoding a signal carrying a representation of an image and a decoder for decoding a signal carrying a representation of an image. The encoder comprising a first encoder for encoding a signal carrying a representation of an image at a first quality level; and a second encoder for encoding a signal carrying a representation of the image at a second quality level at greater quality than the first quality level. The encoder also comprises a mixer for mixing a factor of a signal carrying a representation of a prediction of the image at the first quality level and a factor of a signal carrying a representation of a prediction of the image at the second quality level to produce a mixed signal. The encoder is arranged such that: the second encoder encodes the signal carrying a representation of the image at a second quality level based on the mixed signal; and the factors are selected based on a measure of noise introduced by the first encoder and the second encoder when producing the predictions. A corresponding decoder and methods of decoding and encoding are also described.

Claims

1-112. -112. (canceled)

113. An image encoding-decoding system comprising:

(a) an encoder for encoding a signal carrying a representation of an image; and

(b) a decoder for decoding signals carrying an encoded representation of an image wherein;

the encoder comprises;

(1) a first encoder for encoding a signal carrying a representation of an image at a first quality level;

(2) a second encoder for encoding a signal carrying a representation of the image at a second quality level at greater quality than the first quality level and wherein the encoder is arranged such that the second encoder encodes the signal carrying a representation of the image at a second quality level based upon at least one of;

(i) a mixed signal provided from mixing a factor of a signal carrying a representation of a prediction of the image at the first quality level and a factor of a signal carrying a representation of a prediction of the image at the second quality level to produce the mixed signal wherein the factors are selected based upon a measure of noise introduced by the first encoder and the second encoder when producing the predictions; and

(ii) a signal carrying a frequency domain representation of a prediction of the image at both the first quality level and at the second quality level; and

the decoder comprises;

(1) a first decoder for decoding a signal carrying an encoded representation of an image at a first quality level; and

(2) a second decoder for decoding a signal carrying an encoded representation of the image at a second quality level at greater quality than the first quality level and wherein the decoder is arranged such that the second decoder decodes the signal carrying an encoded representation of the image at a second quality level based upon at least one of;

(i) a mixed signal provided by a mixer for mixing a factor of a signal carrying a representation of a prediction of the image at the first quality level and a factor of a signal carrying a representation of a prediction of the image at the second quality level to produce a mixed signal wherein the factors are selected based on a measure of noise introduced when producing the predictions during encoding and which are available to the decoder; and

(ii) a signal carrying a frequency domain representation of a prediction of the image at both the first quality level and at the second quality level.

114. The system of claim 113 wherein the first quality level of the first encoder has a first spatial resolution and the second quality level of the second encoder has a second spatial resolution greater than the first spatial resolution.

115. The system of claim 113 wherein the encoder comprises a mixer for providing the mixed signal by mixing a factor of a signal carrying a representation of a prediction of the image at the first quality level and a factor of a signal carrying a representation of a prediction of the image at the second quality level in the frequency domain.

116. The system of claim 113 wherein the first quality level of the first decoder has a first spatial resolution and the second quality level of the second encoder has a second spatial resolution greater than the first spatial resolution.

117. The system of claim 113 wherein the decoder comprises a mixer for providing the mixed signal by mixing a factor of a signal carrying a representation of a prediction of the image at the first quality level and a factor of a signal carrying a representation of a prediction of the image at the second quality level in the frequency domain.

118. An encoder for encoding a signal carrying a representation of an image, the encoder comprising:

a first encoder for encoding a signal carrying a representation of an image at a first quality level;

a second encoder for encoding a signal carrying a representation of the image at a second quality level at greater quality than the first quality level; and

a mixer for mixing a factor of a signal carrying a representation of a prediction of the image at the first quality level and a factor of a signal carrying a representation of a prediction of the image at the second quality level to produce a mixed signal;

wherein the encoder is arranged such that;

the second encoder encodes the signal carrying a representation of the image at a second quality level based on the mixed signal; and

the factors are selected based on a measure of noise introduced by the first encoder and the second encoder when producing the predictions.

119. An encoder according to claim 118, wherein the first encoder comprises a quantiser for quantising the coefficients of each subband of a signal representing the frequency domain of the image at a first quality level, the first quantiser comprising a plurality of bins, arranged such that in use, the numbers of coefficients which lie within a range of coefficient values included in each bin are output.

120. An encoder according to claim 119, wherein the size of the bins is proportional to a first encoder quantisation factor.

121. An encoder according to claim 120, wherein the measure of noise introduced by the first encoder is derived from the first encoder quantisation factor.

122. An encoder according to claim 118, wherein the second encoder comprises a quantiser for quantising the coefficients of each subband of a signal representing the frequency domain of the image at a second quality level, the second quantiser comprising a plurality of bins, arranged such that in use, the numbers of coefficients which lie within a range of coefficient values included in each bin are output.

123. An encoder according to claim 122, wherein the size of the bins is proportional to a second encoder quantisation factor.

124. An encoder according to claim 123, wherein the measure of noise introduced by the second encoder is derived from the second encoder quantisation factor.

125. An encoder according to claim 118, wherein the encoder comprises a transformer arranged to perform a transform on the signal carrying a representation of an image in order to transform the signal into the frequency domain.

126. An encoder according to claim 125, wherein the transform is a wavelet transform.

127. An encoder according to claim 125, wherein the encoder comprises a reducer which reduces the magnitude of a portion of the frequency components to be encoded by the first encoder.

128. An encoder according to claim 127, wherein the reducer reduces the magnitude of at least some of the portion of the frequency components to be encoded by the first encoder to substantially zero.

129. An encoder according to claim 128, wherein the reducer reduces the magnitude of all of the portion of the frequency components to be encoded by the first encoder to substantially zero.

130. An encoder according to claim 118, wherein the second encoder is arranged to encode signals carrying a representation of an image at a second quality level input into the second encoder in the frequency domain.

131. An encoder according to claim 118, wherein the first encoder is arranged to encode signals carrying a representation of an image at a first quality level input into the first encoder in the spatial domain.

132. An encoder according to claim 118, wherein the first encoder is arranged to encode signals carrying a representation of an image at a first quality level input into the first encoder in the frequency domain.

133. A decoder for decoding signals carrying an encoded representation of an image, the decoder comprising:

a first decoder for decoding a signal carrying an encoded representation of an image at a first quality level;

a second decoder for decoding a signal carrying an encoded representation of the image at a second quality level at greater quality than the first quality level; and

a mixer for mixing a factor of a signal carrying a representation of a prediction of the image at the first quality level and a factor of a signal carrying a representation of a prediction of the image at the second quality level to produce a mixed signal;

wherein the decoder is arranged such that;

the second decoder decodes the signal carrying an encoded representation of the image at a second quality level based on the mixed signal; and

the factors are selected based on a measure of noise introduced when producing the predictions during encoding and which are available to the decoder.

134. A decoder according to claim 133, wherein the measure of noise introduced by encoding is derived from encoder quantisation factors.

135. A decoder according to any of claims 133, wherein the mixer produces a weighted sum output α

X+(1−

α

)Y where α

is a weighting factor, X is the prediction of the image at the first quality level, and Y is the prediction of the image at the second quality level.

136. A decoder according to claim 135, wherein α

depends on a first encoder quantisation factor and a second encoder quantisation factor.

137. A decoder according to claim 133, wherein the weighting factor

$\alpha$

= 1 1 + σ

x 2 σ

y 2

where α

_x² is the error between the signal carrying a representation of an image at a first quality level and a signal carrying a spatial domain representation of a prediction of the image at the first quality level and σ

_y² is the error between the signal carrying a representation of an image at second quality level and a signal carrying a spatial domain representation of a prediction of the image at the second quality level.

138. A decoder according to claim 137, wherein σ

_x² depends on the first encoder quantisation factor.

139. A decoder according to claim 137, wherein σ

_y² depends on the second encoder quantisation factor.

140. A decoder according to any of claims 135, wherein α

depends on the difference between the logarithm of the first encoder quantisation factor and the second encoder quantisation factor.

141. A decoder according to claim 140, comprising a look-up table in which a signal representing α

is output from the look-up table when a signal representing the difference between the first encoder quantisation factor and the second encoder quantisation factor is input to the look-up table.

142. A decoder according to claim 141, wherein the weighting factor

$\alpha$

= 1 1 + q base 2 q enhancement 2

where q_base is the first encoder quantisation factor and q_enhancement is the second encoder quantisation factor.

143. A decoder according to any of claim 135, wherein the first encoder quantisation factor is different to the second encoder quantisation factor.

144. A decoder according to claim 143, wherein the first encoder quantisation factor is greater than the second encoder quantisation factor.

145. A decoder according to any of claim 135, further comprising a receiver for receiving an encoded signal from the second encoder at a time before receiving an encoded signal from the first encoder.

146. A decoder according to any of claim 135, further comprising a receiver for receiving an encoded signal output from the first encoder at an information transmission rate greater than that of the second encoder.

147. A decoder according to any of claim 135, wherein the first decoder is an MPEG2 decoder.

148. A decoder according to any of claim 135, further comprising a receiver for receiving the measure of noise introduced by encoding.

149. A decoder according to any of claim 135, further comprising memory for storing the measure of noise introduced by encoding.

150. A method of encoding a signal carrying a representation of an image, the method comprising:

encoding a signal carrying a representation of an image at a first quality level;

mixing a factor of a signal carrying a representation of a prediction of the image at the first quality level and a factor of a signal carrying a representation of a prediction of the image at the second quality level to produce a mixed signal;

encoding the signal carrying a representation of the image at a second quality level based on the mixed signal; and

selecting the factors based on a measure of noise introduced by the first encoder and the second encoder when producing the predictions.

151. A method of encoding according to claim 150, wherein encoding a signal carrying a representation of an image at a first quality level comprises quantising the coefficients of each subband of a signal representing the frequency domain of the image at a first quality level into bins, and outputting the numbers of coefficients which lie within a range of coefficient values included in each bin.

152. A method of encoding according to claim 150, wherein the size of the bins is proportional to a first encoder quantisation factor.

153. A method of encoding according to claim 150, wherein the measure of noise introduced by the first encoder is derived from the first encoder quantisation factor.

154. A method of decoding a signal carrying an encoded representation of an image, the method comprising:

decoding a signal carrying an encoded representation of an image at a first quality level;

mixing a factor of a signal carrying a representation of a prediction of the image at the first quality level and a factor of a signal carrying a representation of a prediction of the image at a second quality level greater quality than the first quality level to produce a mixed signal;

decoding a signal carrying an encoded representation of the image at a second quality level based on the mixed signal; and

selecting the factors based on a measure of noise introduced when producing the predictions during encoding and which are available to the decoder.

155. A method of decoding according to claim 154, wherein the first quality level is a first spatial resolution and the second quality level is a second spatial resolution greater than the first spatial resolution.

156. A method of decoding according to claim 154, comprising mixing a factor of a signal carrying a representation of a prediction of the image at the first quality level and a factor of a signal carrying a representation of a prediction of the image at a second quality level in the frequency domain.

157. A method of decoding according to claim 154, wherein the measure of noise introduced when producing the predictions is derived from encoder quantisation factors.