陈士磊. 基于张量形态的反调和图像边缘检测算法[J]. 微电子学与计算机, 2015, 32(6): 105-108. DOI: 10.19304/j.cnki.issn1000-7180.2015.06.023
引用本文: 陈士磊. 基于张量形态的反调和图像边缘检测算法[J]. 微电子学与计算机, 2015, 32(6): 105-108. DOI: 10.19304/j.cnki.issn1000-7180.2015.06.023
CHEN Shi-lei. Reverse Harmonic Image Edge Detection Algorithm Based on Tensor Form[J]. Microelectronics & Computer, 2015, 32(6): 105-108. DOI: 10.19304/j.cnki.issn1000-7180.2015.06.023
Citation: CHEN Shi-lei. Reverse Harmonic Image Edge Detection Algorithm Based on Tensor Form[J]. Microelectronics & Computer, 2015, 32(6): 105-108. DOI: 10.19304/j.cnki.issn1000-7180.2015.06.023

基于张量形态的反调和图像边缘检测算法

Reverse Harmonic Image Edge Detection Algorithm Based on Tensor Form

  • 摘要: 提出基于彩色张量形态的反调和图像边缘检测算法,首先构建彩色反调和图像的张量模型,借助于张量丰富的运算更加准确地度量像素间的关系,在张量形态空间中找图像一阶导数中的最大和最小值来检测边界,基于参考全序的彩色张量形态学算子进行算法改进.仿真结果表明,采用该算法进行反调和图像边缘检测,具有较高的反调和平均值,采用张量模型刻画彩色信息不仅能够考虑到彩色分量间的相关性,而且考虑了彩色信息变换特征,检测性能优越.

     

    Abstract: Image edge detection is studied for identification in digital image brightness changes obviously, and improving the visual recognition ability of graphics. The pixel color information transform harmonic image uneven, edge detection is difficult. Contrarian and the traditional arithmetic of image edge detection using tensor model algorithm, due to the structure and operation of the tensor vector is relatively simple, edge detection performance is not good. Proposed a color tensor form of the tune and the image edge detection algorithm based on color tensor model, firstly construct the harmonic images, with the help of tensor rich operations to more accurately measure the relationship between pixels, in tensor form space for image derivative of the maximum and minimum value to detect the boundary, the improved algorithm for color tensor morphology operators based on full order reference. The simulation results show that, using this algorithm to tune and the image edge detection, it has a higher harmonic average, using tensor model to depict the color information can not only take into account the correlation between color components, but also consider the color information transformation characteristics of superior performance in detection.

     

/

返回文章
返回