Title :
Effect of CNN shape segmentation on MPEG-4 shape bit-rate
Author :
Koskinen, Lauri ; Paasio, Ari ; Laiho, Mika ; Halonen, Kari
Author_Institution :
Electron. Circuit Design Lab., Helsinki Univ. of Technol., Finland
Abstract :
The effect of the cellular nonlinear network (CNN) shape segmentation algorithm on the shape bit-rate of a context-based arithmetic encoder has been simulated. Also, the correlation of shape boundary block and opaque block motion vectors and shape motion vectors and texture motion vectors has been simulated. It is found that, along with content-based MPEG-4 encoding, the intermediate results of the CNN algorithm can be used for power consumption reduction in the encoder.
Keywords :
arithmetic codes; cellular neural nets; image segmentation; image sequences; image texture; motion estimation; video coding; CNN shape segmentation effect; MPEG-4 shape bit-rate; cellular nonlinear network shape segmentation algorithm; content-based MPEG-4 encoding; context-based arithmetic encoder; motion estimation; opaque block motion vectors; power consumption reduction; shape boundary block; shape motion vectors; texture motion vectors; video sequence; Algorithm design and analysis; Arithmetic; Cellular networks; Cellular neural networks; Context modeling; Encoding; Energy consumption; MPEG 4 Standard; Motion estimation; Shape;
Conference_Titel :
Circuits and Systems, 2002. ISCAS 2002. IEEE International Symposium on
Print_ISBN :
0-7803-7448-7
DOI :
10.1109/ISCAS.2002.1010515
