【问题标题】:How do i extract 3 bands out of a 4 band image using NumpyArrayToRaster()?如何使用 NumpyArrayToRaster() 从 4 波段图像中提取 3 波段?
【发布时间】:2016-01-24 19:02:52
【问题描述】:

我需要从 4 波段图像中提取 3 波段。我正在使用一个名为 NumpyArrayToRaster() 的函数,它最多只能接受 3 个波段图像。我如何使它适用于 4 波段图像?

这是我现在的代码-

import arcpy
arcpy.CheckOutExtension("Spatial")
from PIL import Image

import matplotlib.pyplot as plt
import numpy as np
from skimage import io
from skimage.segmentation import quickshift

arcpy.env.overwriteOutput = True

# The input 4-band NAIP image
img = r'C:\Users\Alekhya\Desktop\Krishna\NRSC\processed image\newclip4\clip4.tif'


# Convert image to numpy array
imgarr = io.imread(img)
print imgarr
print imgarr.shape
print imgarr.dtype

# Run the quick shift segmentation
segments = quickshift(imgarr, kernel_size=3, convert2lab=False, max_dist=6, ratio=0.5)
print("Quickshift number of segments: %d" % len(np.unique(segments)))


# View the segments via Python
plt.imshow(segments)
print segments
print segments.shape
print type(segments)
print segments.dtype


# Get raster metrics for coordinate info
imgRaster = arcpy.sa.Raster(img)

# Lower left coordinate of block (in map units)
mx = imgRaster.extent.XMin
my = imgRaster.extent.YMin
sr = imgRaster.spatialReference

'''
# Note the use of arcpy to convert numpy array to raster
seg = arcpy.NumPyArrayToRaster(segments, arcpy.Point(mx,my), imgRaster.meanCellWidth, imgRaster.meanCellHeight)

outRaster = r'C:\Users\Alekhya\Desktop\Krishna\NRSC\processed image\newclip4\segments_clip4.tif'
seg_temp = seg.save(outRaster)
arcpy.DefineProjection_management(outRaster, sr)
'''

# Calculate NDVI from bands 4 and 3
b4 = arcpy.sa.Raster(r'C:\Users\Alekhya\Desktop\Krishna\NRSC\processed image\newclip4\clip4.tif\Band_4')
b3 = arcpy.sa.Raster(r'C:\Users\Alekhya\Desktop\Krishna\NRSC\processed image\newclip4\clip4.tif\Band_3')
ndvi = arcpy.sa.Float(b4-b3) / arcpy.sa.Float(b4+b3)
print ndvi

# Extract NDVI values based on image object boundaries
zones = arcpy.sa.ZonalStatistics(segments, "VALUE", ndvi, "MEAN")
zones.save(r'C:\Users\Alekhya\Desktop\Krishna\NRSC\processed image\newclip4\zones_clip4.tif')

# Classify the segments based on NDVI values
binary = arcpy.sa.Con(zones < 20, 1, 0)
binary.save(r'C:\Users\Alekhya\Desktop\Krishna\NRSC\processed image\newclip4\classified_clip4.tif')

【问题讨论】:

  • 向我们展示您的尝试
  • 我面临的问题是将 numpy 数组转换回光栅图像。函数 NumpyArrayToRaster() 接受最多 3 个波段的图像。我目前正在使用 4 波段图像,必须将其转换为光栅。因此,我需要有关如何将其转换为 3 波段图像或任何其他没有任何波段限制的功能的建议。
  • pastie.org/10507013# 是我尝试过的代码的链接
  • 目标 3 波段具有不同的波长(范围)然后在 4 波段图像中?如果是,那么您需要重新整合乐队......见Multi-Band Image raster to RGB

标签: python numpy image-processing image-extraction


【解决方案1】:

在阅读您的代码时,我意识到您需要计算 NDVI 值。也许您稍微修改您的方法,而不是使用函数NumpyArrayToRaster(),您可以使用更简单的方法?

在这里我提供我的代码,其中: - 从多个日期读取单个 Sentinel 数据堆栈(多波段复合) - 使用rasterName/Band_X 识别 NDVI 计算所需的波段 - 从波段计算 NDVI - 将每个日期的 NDVI 保存到输出文件夹,保留其名称和日期

在 Landsat 中,NDVI 的波段是 Band_4 和 Band_3

我的代码:

# calculate NDVI from sentinel data
# list raster and calculate NDVI per each raster individually

# import modules
import arcpy, string

# import environmental settings
from arcpy import env
from arcpy.sa import *

# check out spatial extension
arcpy.CheckOutExtension("spatial")
arcpy.env.overwriteOutput = True

# add workspace
env.workspace = "C:/Users/input"

# List rasters
rasters = arcpy.ListRasters("*", "TIF")

# Output directory
outWd = "C:/Users/output/ndvi"

# calculate ndvi for every sentinel raster
for raster in rasters:
    # define ndvi outputs
    outNDVI = outWd + "/"+ raster.replace(".tif", "_ndvi.tif")
    print "outNDVI is " + outNDVI

    # specify inputs for ndvi and final output
    # NDVI takes NIR and Red, which are in Sentinel  Band 4 and Band 8 
    Red = raster + '\Band_4'
    NIR = raster + '\Band_8'

    # Create Numerator and Denominator rasters as variables, and
    # NDVI as output
    Num = arcpy.sa.Float(Raster(NIR) - Raster(Red))
    Denom = arcpy.sa.Float(Raster(NIR) + Raster(Red))
    NDVI = arcpy.sa.Divide(Num, Denom)
    print "NDVI calculating"

    # save results output
    NDVI.save(outNDVI)
    print "NDVI saved"

【讨论】:

    猜你喜欢
    • 2019-12-17
    • 2021-08-26
    • 1970-01-01
    • 2021-02-08
    • 2018-12-19
    • 1970-01-01
    • 1970-01-01
    • 1970-01-01
    • 1970-01-01
    相关资源
    最近更新 更多