基于链接突触计算网络的遥感图像对比度增强算法

CONTRAST ENHANCEMENT ALGORITHM OF REMOTE SENSING IMAGES BASED ON LINKING SYNAPTIC COMPUTATION NETWORK

针对卫星遥感图像在增强对比度和去除噪声过程中出现纹理细节丢失和图像自然性降低的问题,提出一种基于链接突触计算网络的卫星遥感图像对比度增强算法。在彩色图像转变至HSV颜色空间的前提下,通过引导滤波器去除噪声和边缘增强技术突出细节;将边缘增强的图像输入链接突触计算网络,采取与显著特征图相结合的方式增强图像亮度和对比度;在RGB空间上,利用改进的自然色彩还原技术对增强图像进行还原。实验结果表明,相对于其他对比度增强算法,该方法在多个评价指标上都有很好的性能体现,在增强图像对比度和亮度的同时,保留了丰富的纹理细节和自然色彩度。

In order to solve the problems of texture details loss and image naturalness reduction in the process of contrast enhancement and noise removal of satellite remote sensing images,we propose a contrast enhancement algorithm of satellite remote sensing images based on linking synaptic computation network.It was based on the premise that the color image was transformed into HSV color space.The noise was removed by guiding filter,and the edge enhancement technology was used to highlight the details.Then,the edge-enhanced image was input into the synaptic computation network,and the brightness and contrast of the image were enhanced by combining with the prominent feature map.In the RGB space,enhanced natural color reproduction technology was used to restore the enhanced image.The experimental results show that compared with other contrast enhancement algorithms,the proposed method has good performance in many evaluation indexes.Moreover,this method retains rich texture details and natural color while enhancing the contrast and brightness of the image.