摘要
针对传统的图像可识别特征分割方法分割不准确,效率低的问题,提出了有向空间关系和局部灰度聚类模型相结合的图像可识别特征分割方法。通过对图像可识别特征点的显著性检测,建立图像可识别特征轮廓模型。采用梯度下降法提取图像可识别特征,对图像可识别特征进行分析。以此为基础,通过有向空间关系可确定图像可识别特征间的关系,通过局部灰度聚类规则建立图像可识别特征的局部灰度聚类模型,实现对图像可识别特征分割方法的改进。实验结果证明,采用改进的分割方法进行图像可识别特征分割时,其分割精度,分割所需时间,分割效率均要优于传统的分割方法,具有一定的优势。
In order to solve the problems of low accuracy and low efficiency in traditional image segmentation methods, a new method based on the combination of directional spatial relation and local gray clustering model is proposed. Based on the detec- tion of the feature points of the image, an image recognition feature contour model is established. Gradient descent method is used to extract the feature of image recognition. On this basis, by the relationship between the characteristics of image recogni- tion and the space using local gray clustering model and image recognition features through local gray clustering, the image rec ognition feature segmentation method is improved. The experimental results show that the segmentation accuracy, the time re- quired for segmentation, and the segmentation efficiency are better than the traditional segmentation method, hence, it has some advantages.
出处
《微型电脑应用》
2017年第4期35-38,共4页
Microcomputer Applications
关键词
PACS系统
图像
特征
分割方法
PACS system
Image
recognition
Feature
Segmentation method
