def load_csv():
params = ['coords', 'columns', 'name']
parsed_coordinates = ''
# Open GSLIB file as csv object for formatting purposes
with open(in_file, 'rb') as f:
doc_reader = csv.reader(f, delimiter='\t')
parse_header(params, doc_reader)
# parse_coordinates(params[0])
# Reverse the remaining entries
doc_reversed = reversed(list(doc_reader))
# Create output file to be loaded into ArcMap
with open(out_file, 'wb') as new_file:
doc_writer = csv.writer(new_file,
delimiter=' ',
quoting=csv.QUOTE_NONE,
escapechar=' ')
# Re-write header back to the top of reversed list
for line in header:
new_file.write(line + '\n')
# Write remaining entries into newly created file
for row in doc_reversed:
doc_writer.writerow(row)
# Convert ASCII file to raster. Output will be in same directory as input GSLIB file
arcpy.ASCIIToRaster_conversion(out_file, out_raster)
arcpy.Mirror_management(
out_raster, os.path.join(in_workspace, 'outputRaster_mirrored.tif'))
# Clean-up stray files
arcpy.Delete_management(out_raster)
def random_ClipRaster(trainRaster,labelRaster,outDir,imgNums,imgHeight,imgWidth): #定义随机裁剪函数（训练原图，标签原图，输出路径，裁剪数量，图像行数，图像列数）
Min_X = arcpy.GetRasterProperties_management(trainRaster,"LEFT") #训练原图的X的最小值
Min_Y = arcpy.GetRasterProperties_management(trainRaster,"BOTTOM") #训练原图的Y的最小值
Max_X = arcpy.GetRasterProperties_management(trainRaster,"RIGHT") #训练原图的X的最大值
Max_Y = arcpy.GetRasterProperties_management(trainRaster,"TOP") #训练原图的Y的最大值
CELLSIZEX = arcpy.GetRasterProperties_management(trainRaster,"CELLSIZEX") #获取训练原图单像素的宽度
CELLSIZEY = arcpy.GetRasterProperties_management(trainRaster,"CELLSIZEY") #获取训练原图单像素的高度
COLUMNCOUNT = arcpy.GetRasterProperties_management(trainRaster,"COLUMNCOUNT") #获取训练原图列数
ROWCOUNT =arcpy.GetRasterProperties_management(trainRaster,"ROWCOUNT") #获取训练原图行数
label_Min_X = arcpy.GetRasterProperties_management(labelRaster,"LEFT") #标签原图的X的最小值
label_Min_Y = arcpy.GetRasterProperties_management(labelRaster,"BOTTOM") #标签原图的Y的最小值
label_Max_X = arcpy.GetRasterProperties_management(labelRaster,"RIGHT") #标签原图的X的最大值
label_Max_Y = arcpy.GetRasterProperties_management(labelRaster,"TOP")#标签原图的Y的最大值
label_CELLSIZEX = arcpy.GetRasterProperties_management(labelRaster,"CELLSIZEX") #获取标签原图单像素的宽度
label_CELLSIZEY = arcpy.GetRasterProperties_management(labelRaster,"CELLSIZEY") #获取标签原图单像素的高度
label_COLUMNCOUNT = arcpy.GetRasterProperties_management(labelRaster,"COLUMNCOUNT") #获取标签原图列数
label_ROWCOUNT =arcpy.GetRasterProperties_management(labelRaster,"ROWCOUNT") #获取标签原图行数
rotate_list = ["90","180","270","mirror"]
mkdir_dem = outDir + "train_dem"
mkdir_label = outDir + "train_label"
os.popen("mkdir "+mkdir_dem) #计算机先自动生成文件夹
os.popen("mkdir "+mkdir_label)
if (str(label_Min_X) == str(Min_X)) and (str(label_Min_Y) == str(Min_Y))and ( str(Max_X) == str(label_Max_X))and( str(Max_Y) == str(label_Max_Y))and (str(label_CELLSIZEX) == str(CELLSIZEX))and(str(label_CELLSIZEY)== str(CELLSIZEY))and( str(COLUMNCOUNT) == str(label_COLUMNCOUNT))and(str(ROWCOUNT)== str(label_ROWCOUNT)):
num =0
#print("test")
while num < imgNums:
random_x = random.randint(0,int(str(COLUMNCOUNT)) - imgWidth) #拟随机生成行列号来生成图像，(0,COLUMNCOUNT - imgWidth)为可选列号的范围
random_y = random.randint(0,int(str(ROWCOUNT))- imgHeight) #(0,ROWCOUNT- imgHeight)为可选行号的范围
temp_x = float(str(Min_X)) + random_x * float(str(CELLSIZEX)) #temp_x为所选图像左下角的横坐标
temp_y = float(str(Min_Y)) + random_y * float(str(CELLSIZEY)) #temp_y为所选图像左下角的纵坐标
cood_xy = str(temp_x) + " " +str(temp_y) + " " +str(temp_x + imgWidth*(float(str(CELLSIZEX)))) + " " + str(temp_y + imgHeight*(float(str(CELLSIZEY))))#所选图像的左下坐标和右上坐标
#print(cood_xy)
out_demRaster = outDir+"train_dem\\"+"dem"+"_"+ str(num) + ".tif"#dem_1.tif dem_2.tif
out_labelRaster = outDir+"train_label\\" +"label" +"_" +str(num) +".tif" #label_1.tif label_2.tif
arcpy.Clip_management(trainRaster,cood_xy,out_demRaster,"#","#","NONE","MAINTAIN_EXTENT") #裁剪输出
arcpy.Clip_management(labelRaster,cood_xy,out_labelRaster,"#","#","NONE","MAINTAIN_EXTENT") #裁剪输出
#arcpy.Clip_management(trainRaster,cood_xy,out_demRaster,"#","#","NONE","NO_MAINTAIN_EXTENT") #裁剪输出
#arcpy.Clip_management(labelRaster,cood_xy,out_labelRaster,"#","#","NONE","NO_MAINTAIN_EXTENT") #裁剪输出
dem_numpy = arcpy.RasterToNumPyArray(out_demRaster,nodata_to_value=-9999) #栅格转成numpy,目的是统一设置nodata值为-9999，用于后面的判断
dem_numpy = numpy.hstack(dem_numpy) #把numpy数组元素变成一维，用于方便循环
for value in list(dem_numpy):
if value == -9999:
noDataValue = -9999
break
else :
noDataValue = "None"
if noDataValue == "None":
random_rotate = random.randint(0,3)
out_demRotateRaster = outDir+"train_dem\\"+"dem"+"_"+ str(num) +"_"+rotate_list[random_rotate]+ ".tif" #dem_1.tif dem_2.tif
out_labelRotateRaster = outDir+"train_label\\"+"label"+"_"+ str(num) +"_"+rotate_list[random_rotate]+ ".tif" #dem_1.tif dem_2.tif
if rotate_list[random_rotate] == "mirror":
arcpy.Mirror_management(out_demRaster,out_demRotateRaster) #镜像操作
arcpy.Mirror_management(out_labelRaster,out_labelRotateRaster)
else :
rotate_x = temp_x + (imgWidth*float(str(CELLSIZEX)))/2 #计算旋转中心
rotate_y = temp_y + (imgHeight*float(str(CELLSIZEY)))/2
arcpy.Rotate_management(out_demRaster,out_demRotateRaster,rotate_list[random_rotate],str(rotate_x)+" "+str(rotate_y)) #旋转操作
arcpy.Rotate_management(out_labelRaster,out_labelRotateRaster,rotate_list[random_rotate],str(rotate_x)+" "+str(rotate_y))
print("%dth and conversion %s images are generated"%(num,rotate_list[random_rotate])) #提示图片裁剪好了
num = num+1 #图片保留，继续循环
else: #删除生成的train和label图片
remove_dem = glob.glob(outDir + "train_dem\\" + "dem" + "_" + str(num) + ".*") #每一个图片有好几个文件，都需要删除
remove_label = glob.glob(outDir + "train_label\\" + "label" + "_" + str(num) + ".*")
for tiff in remove_dem:
#print(tiff)
os.remove(tiff)
print("Delete"+ " " + tiff)
for label_tiff in remove_label:
os.remove(label_tiff)
print("Delete"+ " " + label_tiff)
num = num
else :
Error = " Image matching failed " #两张源图片格式如果大小不匹配，无法运用此函数，直接error
print(Error)
return
print("Successfully Done!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!")
arcpy.DefineProjection_management(acoRaster, proj)
#rotates acolite image if applicable
if rotateYN == 'true':
arcpy.Rotate_management(acoRaster, "intermedRot" ,"180")
arcpy.env.snapRaster = pathLandsat
arcpy.CheckOutExtension("Spatial")
#flips acolite image if appropriate
if flipYN == 'true':
if rotateYN == 'true':
arcpy.Mirror_management("intermedRot", "intermedMir")
else:
arcpy.Mirror_management(acoRaster, "intermedMir")
#rescales acolite image
cellSizeXResult = arcpy.GetRasterProperties_management(pathLandsat, "CELLSIZEX")
cellSizeYResult = arcpy.GetRasterProperties_management(pathLandsat, "CELLSIZEY")
cellSizeX = cellSizeXResult.getOutput(0)
cellSizeY = cellSizeYResult.getOutput(0)
#Chooses appropriate raster to apply landsat cell size to
if flipYN == 'true':
arcpy.Rescale_management("intermedMir", "intermedScale", cellSizeX, cellSizeY)
else:
if rotateYN == 'true':
arcpy.Rescale_management("intermedRot", "intermedScale", cellSizeX, cellSizeY)
# parse_coordinates(params[0])
# Reverse the remaining entries
doc_reversed = reversed(list(doc_reader))
# Create output file to be loaded into ArcMap
with open(out_file, 'wb') as new_file:
doc_writer = csv.writer(new_file,
delimiter=' ',
quoting=csv.QUOTE_NONE,
escapechar=' ')
# Re-write header back to the top of reversed list
for line in header:
new_file.write(line + '\n')
# Write remaining entries into newly created file
for row in doc_reversed:
doc_writer.writerow(row)
arcpy.ASCIIToRaster_conversion(out_file, Output_raster, "INTEGER")
# Process: Mirror
arcpy.Mirror_management(Output_raster, Output_Raster_Dataset)
# Process: Raster to Point
arcpy.RasterToPoint_conversion(Output_Raster_Dataset, Output_point_features,
"")
# Process: Delete
arcpy.Delete_management(Output_raster, "")