基于视觉传达的舰船遥感图像复原研究

Research on ship remote sensing image restoration based on visual communication

为可靠完成图像退化处理,提升图像视觉质量,提出基于视觉传达的舰船遥感图像复原方法。该方法采用多项式校正方法获取控制点舰船图像坐标和坐标变换结果,依据变换关系校正舰船遥感图像的几何畸变;通过傅里叶变换方法变换校正后图像,将其变换至频率域内,在该域中分析舰船遥感图像退化过程后,采用超复数频域变换模型实现舰船遥感图像复原,获取复原后图像结果。测试结果表明:能够有效完成图像几何畸变校正,校正后图像对应坐标位置精准性显著提升;空间频率结果均在0.92以上;复原后图像的颜色、目标和背景的对比度等均显著提升,满足人眼视觉标准。

In order to reliably complete image degradation processing and improve image visual quality,a ship remote sensing image restoration method based on visual communication is proposed.This method uses polynomial correction method to obtain the coordinates and coordinate transformation results of the control point ship image,and corrects the geometric distortion of the ship remote sensing image based on the transformation relationship;The corrected image is transformed into frequency domain by Fourier transform method.After analyzing the degradation process of ship remote sensing image in this domain,the hypercomplex number frequency domain transform model is used to restore the ship remote sensing image and obtain the restored image results.The test results show that it can effectively correct the geometric distortion of the image,and the accuracy of the corresponding coordinate position of the corrected image is significantly improved.The spatial frequency results are all above 0.92,the color,target,and background contrast of the restored image have significantly improved,meeting human visual standards.