Question

使用GDAL API（v1.10 - C ++）编写GeoTiff时，输出错误。栅格地理转换是正确的，块写在正确的位置，但像素写在块中的随机位置和值。

示例：http://i.imgur.com/mntnAfK.png

方法：打开GDAL光栅 - ＆gt;复制投影信息＆amp;尺寸 - ＆gt;创建输出GeoTiff - ＆gt;在偏移处从数组写一个块。

代码：

//Open the input DEM
const char* demFName = "/Users/mount.dem";
const char* outFName = "/Users/geodata/out_test.tif";
auto poDataset = ioUtils::openDem(demFName);
double adfGeoTransform[6];
poDataset->GetGeoTransform( adfGeoTransform );

//Setup driver
const char *pszFormat = "GTiff";
GDALDriver *poDriver;
poDriver = GetGDALDriverManager()->GetDriverByName(pszFormat);
char *pszSRS_WKT = NULL;
GDALRasterBand *poBand;
//Get size from input Raster
int xSize = poDataset->GetRasterXSize();
int ySize = poDataset->GetRasterYSize();

//Set output Dataset
GDALDataset *poDstDS;
char **papszOptions = NULL;

//Create output geotiff
poDstDS = poDriver->Create( outFName, xSize, ySize, 1, GDT_Byte, papszOptions );

//Get the geotrans from the input geotrans
poDataset->GetGeoTransform( adfGeoTransform );
poDstDS->SetGeoTransform( adfGeoTransform );
poDstDS->SetProjection( poDataset->GetProjectionRef() );

//Create some data to write
unsigned char rData[512*512];
//Assign some values other than 0
for (int col=0; col < 512; col++){
    for (int row=0; row < 512; row++){
        rData[col*row] = 50;
    }
}

//Write some data
poBand = poDstDS->GetRasterBand(1);
poBand->RasterIO( GF_Write, 200, 200, 512, 512,
                  rData, 512, 512, GDT_Byte, 0, 0 );

//Close
GDALClose( (GDALDatasetH) poDstDS );
std::cout << "Done" << std::endl;

我出错的任何想法/指示都非常感激。

Answer 1

总是微不足道的......

rData[row*512+col] = 50

在osgeo上向Even Rouault致敬。

GDAL GeoTiff写入损坏（C ++）

1 个答案: