一种轮胎X射线图像缺陷提取和分割方法

A defect extraction and segmentation method for radial tire X-ray image

针对轮胎X射线图像对比度不高、图像噪声多和存在较大背景起伏的特点,提出一种基于平滑滤波和灰度校正的轮胎X射线图像缺陷提取和分割方法。首先采用平滑滤波的方法,消除带束层帘线影像和图像中的高频噪声;然后依据整幅图像的灰度均值和图像中每个像素的邻域内部灰度均值的比例关系,对像素的灰度进行校正;最后对校正后的图像进行二值化处理,完成对轮胎X图像的分割。实验结果表明,该方法能够有效地实现轮胎X射线图像缺陷的提取和分割。

The image segmentation method based on smoothing filter and gray compensation is proposed to eliminate the redundant background of the X-ray image of the radial tire,which is caused by the low contrast,noise and background fluctuation.Firstly,the smoothing filter is introduced to eliminate the belt and noise of tire X-ray image.The second step is to correct the gray of pixels according to the proportion relationship between the mean of whole image and neighborhood mean of each pixel.Finally,the binarization is performed to complete image segmentation.The experiment results show that the proposed methods can achieve a good defect extraction and segmentation on the tire X-ray images.