摘要
提出一种基于红外热图像的沥青路面温度离析区域检测方法。将红外热图像中沥青路面的温度和改进后的灰度图像边缘分割模型相结合,建立了沥青路面温度离析区域的检测模型,利用对偶法对模型中的参数进行求解,将BP神经网络优化和最小二乘拟合相结合,确立了红外热图像中沥青路面温度和模型中灰度之间的关系,实现了沥青路面温度离析区域的检测。最后通过仿真实验验证了提出方法的可行性,为道路工程建设以及科研等领域的红外热图像的处理提供了理论依据。
A temperature segregation area of asphalt pavement detection method based on infrared thermal images is proposed. The asphalt pavement temperature in the infrared thermal image and the improved gray image edge segmentation model are combined to create a temperature segregation area of asphalt pavement detection model. The model parameters are solved by using duality method. BP neural network optimization and the least squares fitting are combined to establish the relationship between the asphalt pavement temperature of the infrared thermal image and gray in the model, and detect the temperature segregation areas of the asphalt pavement. Finally, the feasibility of the proposed method is verified by simulation experiments and the method lays the theoretical foundation for engineering construction of roads, scientific research and other areas of infrared thermal image processing.
出处
《计量学报》
CSCD
北大核心
2017年第1期23-27,共5页
Acta Metrologica Sinica
基金
河北省交通运输厅科技计划项目(Y-131114)
关键词
计量学
红外热图像
温度离析
边缘分割
最小二乘拟合
BP神经网络优化
对偶法
metrology
infrared thermal image
temperature segregation
edge segmentation
least squares fitting
BP neural network optimization
duality method