10 Â 10 km sub-blocks. One full AT block (approx. 14,500 km
2 ) contains around
140–150 DSM/DEM sub-blocks. In terms of sub-blocks, DSMs are generated by the
Match-T software using the multi-ray intersection technique, which gives higher
redundancy and accuracy for surface models. Match-T generates DSM by utilizing
the image pyramid layers using feature-based correlation and the least square
correlation method. This means that points are generated within specific pyramid
layers using feature-based matching; then the points are filtered using least squares
adjustments. This process results in a highly precise DSM.
The Match-T software generates dense point clouds based on advanced image
and feature-matching techniques and data-filtering algorithms, and a 5.0-m grid
spacing consists of hundreds of thousands of DSM points per photogrammetric
stereo model. Different algorithms were used by the Match-T software for image
correlation and automatic point matching depending on the defined flat/undulating/
mountainous/extreme terrain type.
When multiple aerial cameras are used to acquire the images for the same block,
separate camera calibrations are defined in the Match-AT/Match-T project file. Each
input image was assigned the relevant aerial camera definition, and its exterior
orientation parameters were imported from the Match-AT aerial triangulation and
block adjustment project (.prj) file.
The density of DSM points generated by Match-T was defined as 5.0 m based on
the Terms of Reference (ToR) of the contract. The generated DSM point cloud also
contained elevation points on non-terrain objects such as buildings and vegetation.
The final triangulated irregular network (TIN) model for a block was generated
based on the filtered and classified DEM points in combined sub-blocks, and the
final regular DEM with 5.0-m grid spacing for a full block was interpolated based on
the final TIN model.
13.3.1.4 Orthophoto Production, Mosaicking, and Tile Cutting from
the AI and SI
Orthophoto maps are digitally rectified aerial images, in which the central projection
aerial images have been mathematically transformed into the orthogonal projection,
producing a geometrically corrected aerial image with a uniform scale between
objects over the whole image area (Fig. 13.6). Adjacent and overlapping
orthorectified aerial images are merged in the orthoimage mosaicking process to
produce further and cut color-balanced seamless orthoimage tiles (orthophoto map
sheets).
The following procedures were used for orthophoto production:
• Orthorectification to geometrically correct the image. The newly generated accurate ortho DEM area was used for the orthorectification process.
• Use of seam lines to avoid image displacements while mosaicking the images
together.
13 Spatial Data Usage in Turkish Agriculture
243
2 ) contains around
140–150 DSM/DEM sub-blocks. In terms of sub-blocks, DSMs are generated by the
Match-T software using the multi-ray intersection technique, which gives higher
redundancy and accuracy for surface models. Match-T generates DSM by utilizing
the image pyramid layers using feature-based correlation and the least square
correlation method. This means that points are generated within specific pyramid
layers using feature-based matching; then the points are filtered using least squares
adjustments. This process results in a highly precise DSM.
The Match-T software generates dense point clouds based on advanced image
and feature-matching techniques and data-filtering algorithms, and a 5.0-m grid
spacing consists of hundreds of thousands of DSM points per photogrammetric
stereo model. Different algorithms were used by the Match-T software for image
correlation and automatic point matching depending on the defined flat/undulating/
mountainous/extreme terrain type.
When multiple aerial cameras are used to acquire the images for the same block,
separate camera calibrations are defined in the Match-AT/Match-T project file. Each
input image was assigned the relevant aerial camera definition, and its exterior
orientation parameters were imported from the Match-AT aerial triangulation and
block adjustment project (.prj) file.
The density of DSM points generated by Match-T was defined as 5.0 m based on
the Terms of Reference (ToR) of the contract. The generated DSM point cloud also
contained elevation points on non-terrain objects such as buildings and vegetation.
The final triangulated irregular network (TIN) model for a block was generated
based on the filtered and classified DEM points in combined sub-blocks, and the
final regular DEM with 5.0-m grid spacing for a full block was interpolated based on
the final TIN model.
13.3.1.4 Orthophoto Production, Mosaicking, and Tile Cutting from
the AI and SI
Orthophoto maps are digitally rectified aerial images, in which the central projection
aerial images have been mathematically transformed into the orthogonal projection,
producing a geometrically corrected aerial image with a uniform scale between
objects over the whole image area (Fig. 13.6). Adjacent and overlapping
orthorectified aerial images are merged in the orthoimage mosaicking process to
produce further and cut color-balanced seamless orthoimage tiles (orthophoto map
sheets).
The following procedures were used for orthophoto production:
• Orthorectification to geometrically correct the image. The newly generated accurate ortho DEM area was used for the orthorectification process.
• Use of seam lines to avoid image displacements while mosaicking the images
together.
13 Spatial Data Usage in Turkish Agriculture
243