【摘要】 提出利用计算机双目立体视觉技术对低压电器开关电弧进行深度信息提取,构建了双目立体视觉电弧图像采集系统物理框架,并对系统硬件实现作了详细论述,采集了开断电弧从燃弧到灭弧整个过程图像。为了把电弧图像有用信息提取出来,使电弧运动图像边缘清晰可辨,在对传统的图像边缘检测技术进行分析比较后,提出了基于支持向量机高斯核的电弧图像边缘检测方法,采用径向基(RBF)核函数,引入拉格郎日(Lagrange)函数分类,构造了一种新的边缘检测模型。实验表明,该方法对电弧弧柱边缘信息检测的同时,降低了噪声的影响,使得电弧图像具有较好的视觉效果。更多还原

【Abstract】 Based on the theory of binocular stereo vision technique,this paper proposes a method,which can acquire deep information from switching arc of low voltage apparatus,build physical frame of binocular stereo vision arc image acquisition system.Moreover it describes the realization of system hardware in detail,and acquires the whole process image from arc starting to arc blowing out.In order to acquire useful information of arc image and make the image edge of arc moving more clear,a new method of arc image edge detection based on support vector machine(SVM) and Gaussian kernel is proposed after analyze and compare the traditional technique.This new method uses radial basis function(also known as the kernel radial basis function) and Lagrange function classification to make a new edge detection model.Experimental results show that this new method not only can detect edge information of arc column,but also decrease the influence of noise,and finally better visual effect of arc image can be gained.更多还原