论文部分内容阅读
文章给出将神经网络与矢量量化相结合的两种编码方法,介绍了其原理和实现的算法。经计算机模拟表明,在比特率低于0.3bpp时仍可得到高质量的编码图象
This paper presents two coding methods combining neural network and vector quantization, and introduces the principle and algorithm of the two methods. Computer simulations show that high quality encoded pictures are still available at bit rates below 0.3 bpp