def pic2data(pathname,region,factor = 'CR'):
try:
'''各区域对应图例坐标、左上角、右下角经纬度'''
'''图片分辨率改变时修改这里的图例位置'''
region_all={'AECN':(523,537,597,793,113.506,38.8444,124.2670,23.1044), #华东
'ACCN':(722,333,798,589,103.626,35.3322,120.005,24.0566), #华中
'ACHN':(948,613,1021,869,65.6571,46.4849,126.862,13.061),#全国
'ANCN':(725,5 ,799,261,104.643,43.0878,122.636,31.1917), #华北
'ANEC':(723,564,797,820,114.539,54.224,131.91,36.3301), #东北
'ABCJ':(884,98 ,955,348,96.4697,34.7539,126.494,25.0003), #长江流域
'ABHH':(692,250,757,499,107.015,40.9176,123.65,30.4626), #黄淮流域
'ACES':(721,537,789,789,103.491,35.7671,128.508,13.2104), #东南沿海
'ASCN':(725,338,791,589,104.654,27.6771,121.318,16.104), #华南
'ANWC':(749,6 ,1017,255,65.4922,43.212,109.31,29.4932),#西北
'ASWC':(727,10 ,791,218,96.1597,34.8143,112.812,20.7235) #西南
}
#'R','CR'的颜色
if factor in ['R','CR']:
'''R & CR'''
colors={(1, 160, 246, 255):12.5,
(0, 216, 0, 255):22.5,
(255, 0, 240, 255):62.5,
(231, 192, 0, 255):37.5,
(214, 0, 0, 255):52.5,
(192, 0, 0, 255):57.5,
(173, 144, 240, 255):72.5,
(150, 0, 180, 255):67.5,
(0, 236, 236, 255):17.5,
(255, 144, 0, 255):42.5,
(255, 0, 0, 255):47.5,
(1, 144, 0, 255):27.5,
(255, 255, 0, 255):32.5}
#'''VIL的颜色'''
elif factor == 'VIL':
colors={(1, 255, 0, 255):12.5,
(0, 200, 0, 255):17.5,
(255, 0, 240, 255):57.5,
(231, 192, 0, 255):32.5,
(214, 0, 0, 255):47.5,
(192, 0, 0, 255):52.5,
(173, 144, 240, 255):72.5,
(0, 236, 236, 255):8.5,
(120, 0, 132, 255):67.5,
(255, 144, 0, 255):37.5,
(255, 0, 0, 255):42.5,
(216, 0, 156, 255):62.5,
(255, 255, 0, 255):27.5,
(1, 144, 0, 255):22.5}
#'''OHP的颜色'''
elif factor == 'OHP':
colors={(255, 0, 240, 255):112.5, #100-125
(214, 0, 0, 255):69, #63-75
(192, 0, 0, 255):87.5,#75-100
(173, 144, 240, 255):130, #130-
(0, 236, 236, 255):17.5, #10-25
(255, 144, 0, 255):44, #38-50
(255, 0, 0, 255):56.5, #50-63
(0, 0, 246, 255):2.5, #0-5
(255, 255, 0, 255):31.5, #25-38
(1, 160, 246, 255):7.5} #5-10
'''河流、行政边界、文字的颜色'''
word_color=[(104, 104, 104, 255),
#(253, 255, 254, 255),
(204, 204, 204, 255),
(190, 232, 255, 255),
(156, 156, 156, 255),
(130, 130, 130, 255)]
'''打开文件'''
img_src = Image.open(pathname)
img_size=img_src.size
region_del=region_all[region]
#定义和图片大小一致的0矩阵
image_value=np.zeros((img_size[0],img_size[1]))
src_strlist = img_src.load()
'''获取格点数据，没有数据为0，河流、行政区域、文字数据为1'''
for i in range(0,img_size[0],1):
for j in range(0,img_size[1],1):
if src_strlist[i,j] in colors: #是否有数据
image_value[i,j]=colors[src_strlist[i,j]]
elif src_strlist[i,j] in word_color: #是否是河流等
image_value[i,j]=1
else: #是不需要的点
image_value[i,j]=0
'''找到图例并填充为0'''
image_value[region_del[0]:region_del[2],region_del[1]:region_del[3]]=0
'''对河流、文字、行政边界的点进行插值'''
image_out=np.copy(image_value)
for i in range(0,img_size[0],1):
for j in range(0,img_size[1],1):
#判断是否为河流等点
if image_value[i,j]==1:
sum_value=0 #周围所有有效点的值的和
count_1=0 #附近河流等点的个数
count_0=0 #附近无效点的个数
count_c=0 #附近数据点的个数
#取该点周围8各点进行判断
for n in [(i-1,j-1),(i-1,j ),(i-1,j+1),
(i ,j-1), (i ,j+1),
(i+1,j-1),(i+1,j ),(i+1,j+1)]:
if image_value[n] > 1:
count_c += 1
sum_value += image_value[n]
elif image_value[n] == 1:
count_1 += 1
else:
count_0 += 1
#周围没有数据点，且有无效点
if count_c == 0 and count_0 > 0:
image_out[i,j]=0
#有数据点
elif count_c >0:
try:
#数据点占多数
if count_c / count_0 > 0.5:
image_out[i,j]=sum_value/count_c
#数据点占少数
elif count_c / count_0 <= 0.5:
image_out[i,j]=0
#无效点个数可能为0
except ZeroDivisionError:
image_out[i,j]=sum_value/count_c
#数据点和无效点都没有
elif count_1 == 8:
#判断再往外的一层
for n in [(i-2,j-2),(i-2,j-1),(i-2,j ),(i-2,j+1),(i-2,j+2),
(i-1,j-2) ,(i-1,j+2),
(i ,j-2) ,(i ,j+2),
(i+1,j-2) ,(i+1,j+2),
(i+2,j-2),(i+2,j-1),(i+2,j ),(i+2,j+1),(i+2,j+2)]:
if image_value[n] > 1:
count_c += 1
sum_value += image_value[n]
elif image_value[n] == 1:
count_1 += 1
else:
count_0 += 1
#没有数据点
if count_c == 0:
image_out[i,j]=0
#有数据点
elif count_c >0:
try:
#数据点占多数
if count_c / count_0 > 0.5:
image_out[i,j]=sum_value/count_c
#数据点占少数
elif count_c / count_0 <= 0.5:
image_out[i,j]=0
#无效点个数可能为0
except ZeroDivisionError:
image_out[i,j]=sum_value/count_c
'''坐标处理'''
#转置
image_out=np.transpose(image_out)
#颠倒y
image_out=image_out[::-1]
#得到经纬度转换后的原始网格对应的经纬度信息
lon,lat=get_latlon(region)
#生成离散点数据
out_index=np.argwhere(image_out > 0)
out_all=np.zeros((len(out_index),3))
out_all[:,0]=lat[out_index[:,0],out_index[:,1]]
out_all[:,1]=lon[out_index[:,0],out_index[:,1]]
out_all[:,2]=image_out[out_index[:,0],out_index[:,1]]
#返回坐标转换后的经纬度对应的数据
return out_all
#如果有问题则返回空list
except:
return []
from mpl_toolkits.basemap import Basemap
import matplotlib.pyplot as plt
region_all=['AECN', #华东
'ACCN', #华中
# 'ACHN',#全国
'ANCN', #华北
'ANEC', #东北
'ABCJ', #长江流域
'ABHH', #黄淮流域
'ACES', #东南沿海
'ASCN', #华南
'ANWC',#西北
'ASWC' #西南
]
lon,lat=get_latlon('ACHN')
#获取图像中的经纬度极值范围
#lon_min=np.min(lon)
#lon_max=np.max(lon)
#lat_min=np.min(lat)
#lat_max=np.max(lat)
dlon=0.04
dlat=0.04
lon_min = 108
lon_max = 132
lat_min = 20
lat_max = 40
#lon_min=latlon_domain[1]
#lon_max=latlon_domain[3]
#lat_min=latlon_domain[0]
#lat_max=latlon_domain[2]
def write_latlon(picfile,datafile):
#从文件路径中提取关键字
name=picfile.split('/')[-1]
factor=name.split('_')[-2]
region=name.split('_')[-1][0:4]
datatime=name.split('_')[4]
#生成输出文件名
outname='MOP_'+region+'_'+'factor'+'_'+datatime+'.latlon'
#获取全国地图的经纬度网格
lon,lat=get_latlon('ACHN')
#获取图像中的经纬度极值范围,生成0.04间隔的等经纬度网格
lon_min1=int(np.min(lon)*100)//4/25
lon_max1=int(np.max(lon)*100)//4/25 +0.04
lat_min1=int(np.min(lat)*100)//4/25
lat_max1=int(np.max(lat)*100)//4/25 +0.04
#生成0.04格距的网格
grid_x, grid_y = np.mgrid[lon_min1:lon_max1+0.04:0.04, lat_min1:lat_max1+0.04:0.04]
#生成网格对应的空数据
grid_z1=np.zeros([np.shape(grid_x)[0],np.shape(grid_x)[1]])
#获得图像中对应的网格点的数值
out_all=pic2data(picfile)
#如果返回空集合，则返回1退出
if out_all==[]:
return 1
#放大系数
fdxs=10
#乘以放大系数
out_all[:,2]=out_all[:,2]*fdxs
try:
# In[] 打开文件读取latlon文件
with open (datafile,'rb+') as f:
#从256个字节开始往后读数据
f.seek(256)
#获得文件中头文件以外的数据大小
datasize=os.path.getsize(datafile)-256
#读取数据
context1=f.read(datasize)
#获取有数值的点的信息
yxn=struct.unpack('h'*int(datasize/2),context1)
#初始判断符为0
panduan=0
#读取原始latlon文件中数据，挨个读取yxn的数值并写入grid_z1中
for i in yxn:
#panduan为0则该值为y
if panduan==0:
y=i
panduan+=1
continue
#panduan为0则该值为x
elif panduan==1:
x=i
panduan+=1
continue
#panduan为0则该值为n
elif panduan==2:
n0=0
n=i
panduan+=1
continue
#panduan为0则该值为数值
elif panduan==3:
grid_z1[x+n0,y]=i
n0+=1
if n0==n:
panduan=0
continue
#将新的区域的数据写入latlon中
for i in out_all:
grid_z1[int(round((i[1]-lon_min1)/0.04)),
int(round((i[0]-lat_min1)/0.04))]=i[2]
image_out=grid_z1
#生成latlon文件数据
out=[]
for y in range(0,np.shape(grid_z1)[1],1): #y
ifFirst=True
for x in range(0,np.shape(grid_z1)[0],1):
if image_out[x,y] > 0:
if ifFirst:
if x!=np.shape(grid_z1)[0]-1:
ifFirst=False
out_tmp=[]
count=1
out_tmp.extend([image_out[x,y]])
else:
out.extend([y,x,1])
out.extend([image_out[x,y]])
else:
if x!=np.shape(grid_z1)[0]-1:
out_tmp.extend([image_out[x,y]])
count+=1
else:
out_tmp.extend([image_out[x,y]])
count+=1
out.extend([y,x-count+1,count])
out.extend(out_tmp)
elif image_out[x,y]==0 and not ifFirst:
out.extend([y,x-count,count])
out.extend(out_tmp)
ifFirst=True
#加入结束判断符
out.extend([-1,-1,-1])
#将out转成二进制
for i in out:
try:
data_b+=struct.pack('h',int(i))
except NameError:
data_b=struct.pack('h',int(i))
#删除256以后的值
f.truncate(256)
#从256开始写入新的数据
f.seek(256)
f.write(data_b)
return 0
#不存在原始文件
except FileNotFoundError:
# In[]将新的区域的数据写入latlon中
for i in out_all:
grid_z1[int(round((i[1]-lon_min1)/0.04)),
int(round((i[0]-lat_min1)/0.04))]=i[2]
image_out=grid_z1
#生成文件
out=[]
for y in range(0,np.shape(grid_z1)[1],1): #y
ifFirst=True
for x in range(0,np.shape(grid_z1)[0],1):
if image_out[x,y] > 0:
if ifFirst:
if x!=np.shape(grid_z1)[0]-1:
ifFirst=False
out_tmp=[]
count=1
out_tmp.extend([image_out[x,y]])
else:
out.extend([y,x,1])
out.extend([image_out[x,y]])
else:
if x!=np.shape(grid_z1)[0]-1:
out_tmp.extend([image_out[x,y]])
count+=1
else:
out_tmp.extend([image_out[x,y]])
count+=1
out.extend([y,x-count+1,count])
out.extend(out_tmp)
elif image_out[x,y]==0 and not ifFirst:
out.extend([y,x-count,count])
out.extend(out_tmp)
ifFirst=True
#加入结束判断符
out.extend([-1,-1,-1])
#将out转成二进制
for i in out:
try:
data_b+=struct.pack('h',int(i))
except NameError:
data_b=struct.pack('h',int(i))
# In[]'''生成文件头'''
dataname= outname.encode('utf-8')+b'\x00'*(128-len(outname))
varname =factor.encode('utf-8')+ b'\x00'*(32-len(factor))
char ='dBZ'.encode('utf-8')+ b'\x00'*(16-len('dBZ'))
label = 0 #unsigned short
unitlen = 2 #short
slat=lat_min1 #float
wlon=lon_min1#float
nlat=lat_max1#float
elon=lon_max1#float
clat=(slat+nlat)/2#float
clon=(wlon+elon)/2#float
rows=np.shape(grid_z1)[1] # int
cols=np.shape(grid_z1)[0] # int
dlat=0.04 #float
dlon=0.04 #float
nodata=0.0 #float
levelbytes =0 #long
levelnum=0 #short
amp= fdxs #short
compmode =1 #short
dates = 0 #unsigned short
seconds= 0 #int
min_value = 0 #short
max_value = 0 #short
#reserved =
# head =dataname + varname + char +struct.pack('Hh6f2i3fq3hHi2h6h', label,unitlen,slat,wlon,nlat,
# elon,clat,clon,rows,cols,0.04,0.04,nodata,
# levelbytes,levelnum,amp,compmode,
# dates,seconds,min_value,max_value,
# 0,0,0,0,0,0);
head =dataname + varname + char +struct.pack('Hh6f2i3fi3hHi2h6h', label,unitlen,slat,wlon,nlat,
elon,clat,clon,rows,cols,dlat,dlon,nodata,
levelbytes,levelnum,amp,compmode,
dates,seconds,min_value,max_value,
0,0,0,0,0,0);
# In[] '''存储'''
all_out=head+data_b
#将二进制数据写入文件
with open(datafile, 'wb') as f:
f.write(all_out)
return 0
def write_latlon(picfile, datafile):
name = picfile.split('/')[-1]
factor = name.split('_')[-2]
region = name.split('_')[-1][0:4]
datatime = name.split('_')[4]
outname = 'MOP_' + region + '_' + 'factor' + '_' + datatime + '.latlon'
lon, lat = get_latlon('ACHN')
#获取图像中的经纬度极值范围,生成0.04间隔的等经纬度网格
lon_min1 = int(np.min(lon) * 100) // 4 / 25
lon_max1 = int(np.max(lon) * 100) // 4 / 25 + 0.04
lat_min1 = int(np.min(lat) * 100) // 4 / 25
lat_max1 = int(np.max(lat) * 100) // 4 / 25 + 0.04
#生成0.04格距的网格
grid_x, grid_y = np.mgrid[lon_min1:lon_max1 + 0.04:0.04,
lat_min1:lat_max1 + 0.04:0.04]
#处理
grid_z1 = np.zeros([np.shape(grid_x)[0], np.shape(grid_x)[1]])
out_all = pic2data(picfile)
if out_all == []:
return 1
try:
with open(datafile, 'rb+') as f:
f.seek(260)
datasize = os.path.getsize(datafile) - 260
context1 = f.read(datasize)
yxn = struct.unpack('h' * int(datasize / 2), context1)
panduan = 0
for i in yxn:
if panduan == 0:
y = i
panduan += 1
continue
elif panduan == 1:
x = i
panduan += 1
continue
elif panduan == 2:
n0 = 0
n = i
panduan += 1
continue
elif panduan == 3:
grid_z1[x + n0, y] = i
n0 += 1
if n0 == n:
panduan = 0
continue
for i in out_all:
grid_z1[int(round((i[1] - lon_min1) / 0.04)),
int(round((i[0] - lat_min1) / 0.04))] = i[2]
image_out = grid_z1
#生成文件
out = []
for y in range(0, np.shape(grid_z1)[1], 1): #y
ifFirst = True
for x in range(0, np.shape(grid_z1)[0], 1):
if image_out[x, y] > 0:
if ifFirst:
if x != np.shape(grid_z1)[0] - 1:
ifFirst = False
out_tmp = []
count = 1
out_tmp.extend([image_out[x, y]])
else:
out.extend([y, x, 1])
out.extend([image_out[x, y]])
else:
if x != np.shape(grid_z1)[0] - 1:
out_tmp.extend([image_out[x, y]])
count += 1
else:
out_tmp.extend([image_out[x, y]])
count += 1
out.extend([y, x - count + 1, count])
out.extend(out_tmp)
elif image_out[x, y] == 0 and not ifFirst:
out.extend([y, x - count, count])
out.extend(out_tmp)
ifFirst = True
out.extend([-1, -1, -1])
for i in out:
try:
data_b += struct.pack('h', int(i))
except NameError:
data_b = struct.pack('h', int(i))
f.truncate(260)
f.seek(260)
f.write(data_b)
return 0
except FileNotFoundError:
for i in out_all:
grid_z1[int(round((i[1] - lon_min1) / 0.04)),
int(round((i[0] - lat_min1) / 0.04))] = i[2]
image_out = grid_z1
#生成文件
out = []
for y in range(0, np.shape(grid_z1)[1], 1): #y
ifFirst = True
for x in range(0, np.shape(grid_z1)[0], 1):
if image_out[x, y] > 0:
if ifFirst:
if x != np.shape(grid_z1)[0] - 1:
ifFirst = False
out_tmp = []
count = 1
out_tmp.extend([image_out[x, y]])
else:
out.extend([y, x, 1])
out.extend([image_out[x, y]])
else:
if x != np.shape(grid_z1)[0] - 1:
out_tmp.extend([image_out[x, y]])
count += 1
else:
out_tmp.extend([image_out[x, y]])
count += 1
out.extend([y, x - count + 1, count])
out.extend(out_tmp)
elif image_out[x, y] == 0 and not ifFirst:
out.extend([y, x - count, count])
out.extend(out_tmp)
ifFirst = True
out.extend([-1, -1, -1])
for i in out:
try:
data_b += struct.pack('h', int(i))
except NameError:
data_b = struct.pack('h', int(i))
# In[]
'''生成文件头'''
dataname = outname.encode('utf-8') + b'\x00' * (128 - len(outname))
varname = factor.encode('utf-8') + b'\x00' * (32 - len(factor))
char = 'dBZ'.encode('utf-8') + b'\x00' * (16 - len('dBZ'))
label = 0 #unsigned short
unitlen = 2 #short
slat = lat_min1 #float
wlon = lon_min1 #float
nlat = lat_max1 #float
elon = lon_max1 #float
clat = (slat + nlat) / 2 #float
clon = (wlon + elon) / 2 #float
rows = np.shape(grid_z1)[1] # int
cols = np.shape(grid_z1)[0] # int
#dlat float
#dlon float
nodata = 0.0 #float
levelbytes = 0 #long
levelnum = 0 #short
amp = 0 #short
compmode = 1 #short
dates = 0 #unsigned short
seconds = 0 #int
min_value = 0 #short
max_value = 0 #short
#reserved =
# head =dataname + varname + char +struct.pack('Hh6f2i3fq3hHi2h6h', label,unitlen,slat,wlon,nlat,
# elon,clat,clon,rows,cols,0.04,0.04,nodata,
# levelbytes,levelnum,amp,compmode,
# dates,seconds,min_value,max_value,
# 0,0,0,0,0,0);
head = dataname + varname + char + struct.pack(
'Hh6f2i3fq3hHi2h6h', label, unitlen, slat * 1000, wlon * 1000,
nlat * 1000, elon * 1000, clat * 1000, clon * 1000, rows, cols,
0.04 * 1000, 0.04 * 1000, nodata, levelbytes, levelnum, amp,
compmode, dates, seconds, min_value, max_value, 0, 0, 0, 0, 0, 0)
# In[]
'''存储'''
all_out = head + data_b
with open(datafile, 'wb') as f:
f.write(all_out)
return 0