论文部分内容阅读
电梯曳引机的红外采集图像普遍具有噪声大、图像画质低,针对由于噪声干扰导致图像特征点不明显、故障区域数据获取难等问题,提出基于Contourlet变换去噪算法。该算法与小波变换相比较,对高斯白噪声有明显的抑制作用,较好的保留了图像的轮廓细节和纹理特征。实验结果证明,Contourlet能有效去除图像中的噪声,提高了识别图像中故障区域的准确率,对曳引机故障区点的排除与获取具有重要意义。
The infrared image collected by the elevator traction machine generally has such problems as high noise and low image quality. Aiming at the problem that the image feature points are not obvious due to noise interference and the data acquisition is difficult in the fault area, an image denoising algorithm based on Contourlet transform is proposed. Compared with wavelet transform, the proposed algorithm can significantly reduce the Gaussian white noise, and retains the contour details and texture features well. The experimental results show that Contourlet can effectively remove the noise in the image and improve the accuracy of identifying the fault area in the image, which is of great significance for the exclusion and acquisition of the fault point of the traction machine.