问题:现在我有一些点坐标数据和一幅带有地理信息的遥感影像,如何提取这些点对应的遥感影像的光谱信息呢?
废话不多说,直接上码:
先导入一些包
import rasterio
import numpy as np
import os
import pandas as pd
from pyproj import Transformer
先定义函数
# 提取单张影像的像素值的函数
def extract_pixel_values_single_image(image_path, points_df):
with rasterio.open(image_path) as src:
transform = src.transform
band1 = src.read(1)
vals = []
for _, row in points_df.iterrows():
col, row = ~transform * (row['Longitude_degree'], row['Latitude_degree'])
col, row = int(col), int(row)
vals.append(band1[row, col])
return vals
单张影像提取
如果是单张影像,直接利用这个函数即可,函数中band1 = src.read(1)是读取的第一个波段,参数image_path是遥感影像的路径,points_df是对应的点数据,这里Longitude_degree是经度所在列名,Latitude_degree是纬度所在列名
如:
多张影像提取
如果是多张影像提取,可以利用循环读取文件夹,结合单张影像提取函数来读取:
for folder_name1 in ['fn1','fn2']:
for folder_name2 in ['fn11', 'fn22', 'fn33', 'fn44', 'fn55', 'fn66']:
for continent in ['c1','c2','c3','c4','c5']:
for year in range(2000, 2020): # 要处理的年份
csv_path =r'H:\1.Datasets\7.LandCover\Global\\'+folder_name1+'\\data_'+folder_name2+'\\'+continent+'\\data_'+folder_name2+'_'+continent+'_All_' + str(year) + '_' + str(year + 1) + '_LCdata.csv'
df = pd.read_csv(csv_path)
# 数据类型及其对应文件夹路径
data_folders = {
'2m_air_temperature_mean':r"H:\1.Datasets\global_temperature_2m_mean\Global_Annual_temperature_2m_mean_" + str(year) + '.tif',
'2m_air_temperature_max':r"H:\1.Datasets\global_temperature_2m_max\Global_Annual_temperature_2m_max_" + str(year) + '.tif',
'2m_air_temperature_min':r"H:\1.Datasets\global_temperature_2m_min\Global_Annual_temperature_2m_min_" + str(year) + '.tif',
'precipitation_sum':r"H:\1.Datasets\global_total_precipitation\Global_Annual_Total_Precipitation_" + str(year) + '.tif',
'precipitation_max':r"H:\1.Datasets\global_max_precipitation\Global_Annual_max_Precipitation_" + str(year) + '.tif',
'precipitation_min':r"H:\1.Datasets\global_min_precipitation\Global_Annual_min_Precipitation_" + str(year) + '.tif'
}
# 遍历每种数据类型及其文件夹
for data_type, image_path in data_folders.items():
if os.path.exists(image_path):
pixel_values = extract_pixel_values_single_image(image_path, df)
df[data_type] = pixel_values
else:
print(f"File not found: {image_path}")
output_fn = str(year) + '_' + str(year + 1) + '_'+ folder_name2 +'.csv'
df.to_csv(r'H:\1.Datasets\7.LandCover\Global合并\\'+folder_name1+'\\data_'+folder_name2+'\\'+continent+'\\' + output_fn, index=False)
print(year,folder_name1,folder_name2,continent, '完成')
前面的4个for循环主要是为了构建文件名,以及构建一个包含不同特征影像的data_folders,便于输入下面的像元提取循环。
因为我有好几层文件夹的数据都要处理,而且每个文件夹内时间范围是2000-2019,20个文件,所以文件名的构建就较为复杂。各位在用的时候可以根据自己情况进行更改。