基于自适应阈值的Canny算法在MRI边缘检测中的应用

Application of canny algorithm based on adaptive threshold in MR Image edge detection

Canny算法作为一种经典的多级优化算法广泛应用于核磁共振成像(MRI)边缘检测中,MRI具有灰度密度不均一且对比度低的局限性,本文以腰椎间盘MRI为例提出一种改进的Canny算法在边缘检测中进行优化。首先增强图像对比度,在此基础上引入中值滤波有效处理脉冲噪声进行预处理;针对边缘提取精度不足的问题,增加梯度方向模板求取梯度幅值和方向;针对假边缘数量过多及边缘间断等问题,采用梯度强度信息计算法实现阈值自适应;最后通过直方图正规化再次进行图像增强。本文采用峰值信噪比、结构相似性、均方误差及算法运行时间四方面评价准则对传统算法、现有算法以及本文所改进的算法进行验证,结果表明:改进后的Canny算法对MRI检测精度明显提高且算法的自适应性更强,伪边缘有效减少,本文结果对MRI在医学图像处理中有一定的借鉴意义。

As a classic multi-level optimization algorithm,the Canny algorithm is widely used in MRI edge detection. MRI has the limitations of uneven gray density and low contrast. This paper uses lumbar disc MRI as an example to propose an improved Canny algorithm for optimization in edge detection. First,the image contrast is enhanced,and on this basis,the median filtering is introduced to effectively process impulse noise for preprocessing. Then,for the problem of insufficient edge extraction accuracy,the gradient direction template is increased to obtain the gradient amplitude and direction. For the excessive number of false edges and edge discontinuities,the gradient intensity information calculation method is used to achieve threshold adaptation. Finally,the image enhancement is performed again through the histogram normalization. This paper uses four evaluation criteria of PSNR,SSIM,MSE and algorithm running time to verify the traditional algorithm,the existing algorithm and the improved algorithm in this paper. The results show that the improved Canny algorithm significantly improves the accuracy of MRI detection,effectively reduces the false edges,and it is adaptive. The results of this article have certain reference significance for MRI in medical image processing.