【作者】 李国丽； 刘辉； 邓娜；

【Author】 LI GuoLi LIU Hui DENG Na Department of Electrical Engineering, HeFei University Of Technology, HeiFei 230009, China

【机构】 合肥工业大学电气与自动化工程学院；

【摘要】 研究了一种用于图像压缩的小波神经网络模型,给出了相应的算法和计算公式,并进行仿真模拟,该模型克服了传统BP神经网络隐层节点数目难以确定,收敛速度较慢以及易于收敛到局部极小点等缺点,仿真结果表明网络性能和收敛速率都明显优于传统BP网络,具有良好的应用前景。更多还原

【Abstract】 A model of image compressing using wavelet neural network was studied. The algorithm and mathematical formulas were presented, and the simulated experiments were carried out. The results show that this model can overcome the shortcomings of BP networks such as the difficult on deciding the unit number of hidden layers, the slowly learning rate, and the local converging problem..All these features enable this modle to be of good prospects in application更多还原