| 基于灰度-梯度不变矩的水下目标识别系统 |
| |
| 引用本文： | 张铭钧,尚云超,杨杰.基于灰度-梯度不变矩的水下目标识别系统[J].哈尔滨工程大学学报,2009,30(6). |
| |
| 作者姓名： | 张铭钧 尚云超 杨杰 |
| |
| 作者单位： | 哈尔滨工程大学,机电工程学院,黑龙江,哈尔滨,150001 |
| |
| 基金项目： | 黑龙江省杰出青年科学基金，哈尔滨市科技创新人才专项资金资助项目 |
| |
| 摘 要： | 针对水下图像增强问题,提出了一种基于最小交叉熵的模糊增强方法,该方法定义了一种新的S形隶属度函数,可以减少图像在增强之后灰度信息的丢失;结合最小交叉熵准则自适应确定增强阈值,达到了快速有效的增强效果.针对水下图像光照不均匀问题,提出了基于结合色调的灰度动态线性变换不变矩图像分割算法,结合色调的灰度动态线性变换,抑制受光照不均影响的背景,降低了光照不均对图像分割结果的影响.在图像增强和分割基础上,提取目标的灰度一梯度共轭不变矩特征,以像素梯度特征弥补灰度特征空间信息不足,目标识别率达到94.4%,比灰度不变矩识别提高了6.4%.
|
| 关 键 词： | 水下目标识别 模糊增强 动态线性灰度变换 灰度-梯度不变矩 |
Recognition system for underwater objects based on gray and grads invariant moments |
| |
| ZHANG Ming-jun,SHANG Yun-chao,YANG Jie.Recognition system for underwater objects based on gray and grads invariant moments[J].Journal of Harbin Engineering University,2009,30(6). |
| |
| Authors: | ZHANG Ming-jun SHANG Yun-chao YANG Jie |
| |
| Affiliation: | College of Mechanical and Electrical Engineering;Harbin Engineering University;Harbin 150001;China |
| |
| Abstract: | A fuzzy enhancement method based on minimum cross-entropy was developed in response to problems enhancing underwater images.The method defines a new membership function which guarantees the integrality of the information in enhanced images.By combining minimum cross-entropy rules,the enhancement threshold is determined adaptively,providing a fast and effective enhancement effect.In order to deal with the problem of inhomogeneous illumination,an image segmentation algorithm was developed based on dynamic lin... |
| |
| Keywords: | underwater object recognition fuzzy enhancement dynamic linear gray-level transformation gray and grads invariant moment |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
