高光谱图像可视化增强matlab和python

Great_lid

已于 2024-01-29 20:45:31 修改

阅读量702

点赞数

文章标签： matlab python

于 2023-01-20 15:12:12 首次发布

本文链接：https://blog.csdn.net/pp2373886592/article/details/128741851

版权

该文提供了用于高光谱图像可视化的MATLAB和Python函数，包括rsshow_bandwise和normalize函数，用于对图像的每个波段进行数据缩放和归一化处理，以增强视觉效果。通过对图像像素值的上下限设定并线性变换，确保了图像数据在显示时具有更好的对比度和可读性。

摘要由CSDN通过智能技术生成

高光谱图像可视化视觉增强matlab函数

function I=rsshow_bandwise(I, scale)
    if nargin==1
        scale=0.005;
    end
    I = double(I);
    for i = 1:size(I,3)
        band = I(:,:,i);
        q = quantile(band(:),[scale, 1-scale]);
        [low, high] = deal(q(1),q(2));
        band(band>high) = high;
        band(band<low) = low;
        band = (band-low)/(high-low);
        I(:,:,i)=band;
    end
end

function I = normalize(I)
    low = min(I(:));
    high = max(I(:));
    I = (I - low) / (high - low);
end

高光谱图像可视化视觉增强python函数

def rsshow_bandwise(I, scale=0.005):
    num_band = I.shape[-1]
    I = I.astype(np.double)
    for i in range(num_band):
        band = I[:,:,i]
        low, high = np.quantile(band, [scale, 1-scale])
        band[band > high] = high
        band[band < low] = low
        band = (band-low)/(high-low)
        I[:, :, i] = band
    return I

def rsshow(I, scale=0.005):
    low, high = np.quantile(I, [scale, 1-scale])
    I[I > high] = high
    I[I < low] = low
    I = (I-low)/(high-low)
    return I

def normalize(I):
    low, high = I.min(), I.max()
    I = (I - low) / (high - low)
    return I