摘要
针对打捆钢筋端面识别与统计过程中存在的非均匀光照、噪声干扰等问题,提出将Otsu算法应用于钢筋数目统计的研究,利用基于连通区域标记的中心点提取算法完成计数。通过Matlab软件对图像进行处理,验证了Otsu算法具备优越的抗干扰能力和自适应性,钢筋端面图像识别与统计的正确率高达98.48%。
For current issues about non-uniform illumination and noise disturbance which exists in the image of bundled steel bar under complicated condition, the Otsu algorithm is proposed and applied to the research of steel bar counting. Steel bars number is counted by using the extraction algorithm of center point based on connected regions labeling. Through the progress of the image in Matlab software, Experimental results show that the proposed algorithm has strong ability in anti-disturbance and adaptive force. The counting accuracy is up to 98.48%.
出处
《沈阳理工大学学报》
CAS
2017年第6期47-50,60,共5页
Journal of Shenyang Ligong University
基金
辽宁省教育厅一般项目(L2015468)
沈阳理工大学重点实验室开放基金项目(4771004kfx14)
关键词
图像增强
OTSU
连通区域
钢筋计数
image enhancement
Otsu
connected regions
steel bar counting
