摘要
介绍了用前向单层神经网络实现离散Hough变换(HT)的方法,并且通过对其权值矩阵的修正以及神经元输出函数的修正,改善了HT的性能,提高了HT的分辨率。
This paper introduces the method that use feedfoward single layer neural network to realize Hough Transform (HT). The property and resolution of HT was improved by the modified HT's weight Matrix and the output function of Neuron.
出处
《中国图象图形学报(A辑)》
CSCD
1998年第10期836-839,共4页
Journal of Image and Graphics
基金
国家自然科学基金