摘要
研究钢丝绳断丝准确性检测系统,提高检测的准确率。钢丝绳的断丝极为细微,由于数量巨大,会存在断丝局部断丝混搭,容易造成光谱图像中断丝像素的虚连接,造成图像模糊。传统的基于图像融合的钢丝绳断丝检测方式,在图像连接位置对断丝图像的连接情况检查,由于像素虚连接的存在,会出现像素重叠或者扭曲的现象,造成断丝特征检测准确率过低。为了避免上述问题,提出了一种基于高光谱图像的钢丝绳断丝检测系统。完成对图像进行初始化处理,通过改进无缝融合算法对图像连接位置差异像素进行细丝像素虚连校验处理,提高连接位置像素的精度,准确对融合像素进行断丝检测。实验证明,改进检测系统能够提高检测的准确率,为货物装卸安全性能提供了保证。
This paper put forward a new detecting system of broken wires based on hyperspectral images, the ini- tial treatment of the image was completed and the precision of pixels was improved at the connection position and the detection of broken wires was accurately carried out for the fusion pixel through seamless integration algorithm. The results show that the accuracy of the detection was enhanced by improving detecting system, which supplies the guar- antee for the security of cargo handling.
出处
《计算机仿真》
CSCD
北大核心
2012年第8期242-245,共4页
Computer Simulation
关键词
断丝检测
钢丝绳
无缝融合
Broken wires of detection
Wire rope
Seamless integration
