遙感圖像歸一化

python代碼

import os
import numpy as np
from osgeo import gdal
import glob
import datetime


# 讀圖像文件
def read_img(filename):
    dataset = gdal.Open(filename)  # 打開文件
    im_width = dataset.RasterXSize  # 柵格矩陣的列數
    im_height = dataset.RasterYSize  # 柵格矩陣的行數
    im_geotrans = dataset.GetGeoTransform()  # 仿射矩陣
    im_proj = dataset.GetProjection()  # 地圖投影信息
    bands = dataset.RasterCount
    im_data = dataset.ReadAsArray(0, 0, im_width, im_height).astype(np.float32)  # 將數據寫成數組，對應柵格矩陣
    del dataset  # 關閉對象，文件dataset
    return im_proj, im_geotrans, im_data, im_height, im_width,bands


def write_img(filename, im_proj, im_geotrans, im_data):
    # gdal數據類型包括
    # gdal.GDT_Byte,
    # gdal .GDT_UInt16, gdal.GDT_Int16, gdal.GDT_UInt32, gdal.GDT_Int32,
    # gdal.GDT_Float32, gdal.GDT_Float64

    # 判斷柵格數據的數據類型
    if 'int8' in im_data.dtype.name:
        datatype = gdal.GDT_Byte
    elif 'int16' in im_data.dtype.name:
        datatype = gdal.GDT_UInt16
    else:
        datatype = gdal.GDT_Float32

    # 判讀數組維數
    if len(im_data.shape) != 1:
        im_bands, im_height, im_width = im_data.shape
    else:
        im_bands, (im_height, im_width) = 1, im_data.shape

    # 創建文件
    driver = gdal.GetDriverByName("GTiff")  # 數據類型必須有，因為要計算需要多大內存空間
    dataset = driver.Create(filename, im_width, im_height, im_bands, datatype)

    dataset.SetGeoTransform(im_geotrans)  # 寫入仿射變換參數
    dataset.SetProjection(im_proj)  # 寫入投影
    if im_bands == 1:
        dataset.GetRasterBand(1).WriteArray(im_data)  # 寫入數組數據
    else:
        for i in range(im_bands):
            dataset.GetRasterBand(i + 1).WriteArray(im_data[i])
    del dataset


def Nor(path,out):
    starttime = datetime.datetime.now()
    print('Normalization開始>>>')
    for filename in glob.glob(path):
        a1,bandname= os.path.split(filename)

        print(bandname, '開始>>>>')
        substarttime = datetime.datetime.now()
        proj, geotrans, values, row1, column1,bands = read_img(filename)
        for i in range(bands):  # 對每個圖層進行歸一化
            a = np.min(values[i])
            b = np.max(values[i])
            values[i] =(values[i] - a) / (b - a)
        write_img(out+"\\" + bandname + '.tif', proj, geotrans, values)
        subendtime = datetime.datetime.now() - substarttime
        print(bandname, '結束,一副影像耗費時間：', subendtime)
    endtime = datetime.datetime.now() - starttime
    print('Normalization結束,花費時間:', endtime)


if __name__ == '__main__':
    path=r'E:\湖泊測試數據\改進分水嶺\125.tif'
    out=r'E:\湖泊測試數據\成果'
    Nor(path,out)