论文部分内容阅读
宽幅CCD相机的输出非均匀性校正对获取高质量的遥感图像具有重要意义。通过分析造成CCD输出非均匀的原因,提出了基于辐射亮度反演的非均匀性线性校正概念。在已知的多种辐亮度照明下,采集相机每个像素的灰度值,利用最小二乘法建立像元灰度值与辐亮度的函数,求出像元的两项非均匀性校正系数——暗信号与响应度。实验结果表明:利用这两项校正系数进行相机非均匀性校正,改善效果明显。校正前图像的灰度值均方根偏差为1.3%,校正后为0.2%;从图像上看,校正前图像上的明暗条纹很明显,校正后得到消除。不仅如此,采用该校正方法还能够有效去除通道间输出非均匀性造成的整幅拼接图像的明暗不一致,得到均匀清晰的图像,结果令人满意。
The output non-uniformity correction of a wide-angle CCD camera is of great importance for obtaining high-quality remote sensing images. By analyzing the causes of nonuniform CCD output, a nonuniformity linear correction concept based on the radiance inversion is proposed. Under the known illumination of a variety of radiance, the gray value of each pixel of the camera is collected, and the function of gray value and radiance of the pixel is established by the least squares method. Two nonuniformity correction coefficients of the pixel are obtained - Dark signal and responsiveness. Experimental results show that using these two correction coefficients to correct camera non-uniformity, the improvement is obvious. The RMS value of the gray value of the image before correction was 1.3% and the corrected value was 0.2%. From the image point of view, the bright and dark stripes on the image before correction were obvious and eliminated after correction. Not only that, the correction method can also effectively remove the inconsistencies in the brightness and darkness of the entire mosaic image caused by the output inhomogeneity between channels, and obtain a uniform and clear image with satisfactory results.