计算机与数字工程2019，Vol.47Issue(4)：924-929,979,7.DOI:10.3969/j.issn.1672-9722.2019.04.036
结合灰度化权值调整的改进PCNN图像分割
Improved PCNN Image Segmentation Combined with Grayscale Weight Adjustment
摘要
Abstract
Aiming at the problem that the incomplete partition due to the intensity overlap between background and human targets,a modified pulse coupled neural networks method with weight adjustment for color-to-gray is proposed. This method uses RGB color information as input information of PCNN and the initial dynamic threshold of PCNN is found by using the model of bina?ry threshold value to reduce the number of iterations. Then,the proportion of the output weight is according to the difference be?tween the segmentation targets and the background to obtain the best segmentation results. Experimental results show that a modified pulse coupled neural networks method based on weight adjustment for color-to-gray can obtain a better segmentation result. More?over,this method can save time and improve segmentation efficiency compared with the traditional PCNN method.关键词
图像分割/脉冲耦合神经网络/灰度化/彩色模型Key words
image segmentation/pulse coupled neural network/color-to-gray/color model分类
信息技术与安全科学引用本文复制引用
马跃辉,辛月兰,杨晨晓..结合灰度化权值调整的改进PCNN图像分割[J].计算机与数字工程,2019,47(4):924-929,979,7.基金项目
国家自然科学基金面上项目(编号:11375055)资助. （编号:11375055）
