基于拉普拉斯金字塔的图像细节增强

拉普拉斯金字塔被广泛地运用于到多尺度的图像分解和图像分析中,但由于拉普拉斯金字塔的构成是在高斯金字塔的基础上演变而来的,它被认为是不适合于应用到边缘保持平滑的感知操作中。为了解决这些问题,许多方法已被提出。虽然这些方法被证明是成功的,但他们常伴有较高的计算成本或是需要后期处理。我们算法是运用边缘像素值来区分大型边缘上的小规模细节。在这个结果的基础上,我们提出了一套图像过滤器,以实现图像的细节增强,我们做法的优势是算法的简单性和灵活性,并且不会降低边或引进光晕。

Image Details Enhancement Based on Laplacian Pyramid

The Laplacian pyramid is widely used in the multi-scale image decomposition and image analysis,but because the composition of the Laplacian pyramid is evolved on the basis of the Gaussian pyramid,it was considered not suitable for application to the edge preserving smooth-aware operation.To solve these problems,many methods have been proposed.Although these methods proved to be successful,but they are often accompanied by a high computational cost or need post-processing.Our algorithm is to use the edge of the pixel values to distinguish between large-scale on the edge of small-scale details.On the basis of this result,we propose a set of image filters,in order to achieve the image detail enhancement,we would have the advantage is the simplicity and flexibility of the algorithm,and does not reduce the side or the introduction of Halo.