摘要
近几年,随着神经网络理论的深入研究,神经网络技术的并行性计算能力、非线性映射和自适应能力等优点得到了充分的认识,各种神经网络模型在图像处理领域中得到了广泛的应用。本文对各种流行的神经网络模型在图像处理领域中的应用进行了汇总,根据图像处理的具体内容对这些应用进行分类叙述,阐明了神经网络技术在图像处理领域中的优点和不足之处,并对将来神经网络技术在图像处理领域中的应用提出了几点期望。
Neural network(NN) research has gone a long way in the past decade. The specialists in the image processing field have fully understood the advantages of neural network techniques such as the abilities of parallel computing, nonlinear mapping and self-adaptiveness, and applied a variety of neural network models into the image processing field. In this paper, we review many applications of neural network in image processing and discuss the present and future of neural network. All these applications are categorized according to the phases of image processing and discussed in detail. In conclusion of this paper, we list several disadvantages of neural network techniques, and bring forward the prospect of neural network techniques in the image processing field.
出处
《信息与控制》
CSCD
北大核心
2003年第4期344-351,共8页
Information and Control
基金
国家自然科学基金(60084001)
中法先进研究计划资助项目PRA S101-04
关键词
图像处理
神经网络
非线性映射
图像重建
图像复原
图像增强
特征提取
图像识别
neural network
image reconstruction
image recovery
image enhancement
image reduction
image segmentation
feature extraction
image recognition
