摘要
目的在综合考虑整个尿沉渣系统各环节处理效率的基础上,研究一种快速的尿沉渣图像分割方法。方法首先通过快速高斯滤波滤除图像中噪声的影响,随后通过改进的Sobel边缘增强算法和OSTU二值化方法实现图像的二值化,最后采用基于快速连通区域标记算法实现各候选有形成分的分割。结果该方法通过各种优选算法提高了尿沉渣图像分割的速度和效率。结论快速尿沉渣图像分割方法对于尿沉渣有形成分分析和识别系统的总体设计以及实时、高效系统的开发具有较为重大的应用价值和借鉴意义。
Considering the efficiency of each key step in a whole urinary sediment image analyzing and processing system, a fast urinary sediment image segmenting approach is introduced. Firstly, the fast Gaussian filter is used to eliminate effect of image noise, then an improved sobel algorithm is adopted to enhance the edge pixels and subsequently the OSTU method is utilized to obtain the binary image, finally the candidate shaped elements in the urinary sediment image are segmented out by connected components labeling method. Carefully selecting the algorithm in each step, the speed and efficiency of the introduced segmentation method can be assured, therefore the introduced method is able to lay a foundation for design and development of an efficient automatic recognizing system of urinary sediment shaped elements.
出处
《微计算机信息》
2011年第12期19-21,共3页
Control & Automation
基金
上海市科委重大项目(10DZ1990200)
基金申请人:王云龙
郑建立
项目名称:数字式一体化手术室的研发与产业化
基金颁发部门:上海市科学技术委员会(10DZ1990200)
上海市教委科研创新项目(11YZ116)
基金申请人:王云龙
郑建立
项目名称:数字式一体化手术室的研发与产业化
基金颁发部门:上海市科学技术委员会(10DZ1990200)
关键词
尿沉渣图像
图像分割
高斯滤波
改进Sobel算法
urinary sediment image
image segmentation
Gaussian filtering
enhanced Sobel algorithm
