There are several ways to define a coordinate system in an image mosaic, click "Transform photo block":
Transform the photo block into a new local coordinate system
- 3 points that will define the coordinate system: 1 starting point of an axis, 1 point in positive direction on this axis and 1 point that defines a plane with this axis.
- At least 1 known distance, better 2-3 known distances.
The photo block is transformed into a new coordinate system with a similarity transformation. The coordinate system can be easily visualized with the Right Hand Rule:
Only point numbers that already exist as object points in the project can be selected. Impossible or non-existent point combinations are displayed in red.
In the 3D view the selected axes are shown as an orange line, the selected plane as a semi-transparent area.
Set X,Y,Z axis, start point of the X,Y,Z axis
With this field the starting point of an axis can be defined. If an axis is defined, those entries in the "Define planes" field are automatically dimmed which are not possible.
Point in positive direction.
Here a point must be selected which lies in an orthogonal right system in positive direction on the selected axis.
Point in the XXX plane
With this and the 2 previous points the definition plane of the coordinate system is spanned. It does not have to be somehow "perpendicular" to the other points. However, it should form as large a triangle as possible with the axis definition points.
Origin point of the coordinate system.
This point receives the coordinates 0,0,0 in the new local coordinate system.
Shift in the origin
Here you can enter a shift of the origin point. This can be useful for special applications.
Set scale with distances
At least one distance must be entered here as the scale for the coordinate system. Better are two distances, then ELCOVISION 10 can detect a scale error, best are three distances, then it can identify a wrong input. The gaps to the other distances are shown directly below the input fields, if they are too big they are colored red. The values can be accepted by clicking on the [^] button
Unit of the coordinate system
This information is necessary e.g. for printing to scale of rectified images or if you want to export the data from ELCOVISION 10 to other programs.
Transform the photo block to a control point field
- Minimum 4 control points measured in the images, well distributed over the object
- Control point file in ELCOVISION 10 .coo format or Leica GSI format
The photo block is transferred to a new coordinate system with a similarity transformation.
In the left column of the dialog you see the coordinate values of the object points in the current coordinate system, in the middle column the coordinates of the control point file, in the right column are the transformed coordinate values and the remaining gaps at the control points. If the dialog is closed a new global orientation is optionally performed. This allows to find and eliminate the remaining gaps and faulty measurements and to increase the accuracy.
Transform the image viewpoints to their known viewpoints
- Images with GPS information in the Exif data
- or ASCII file with GPS positions of the pictures
The photo block is transformed into a new coordinate system with a similarity transformation.
In the left column of the dialog you can see the coordinate values of the images in the current coordinate system, in the right column are the transformed coordinate values of the images and the residual gaps at the image positions. With "normal" drone GPS, gaps in the range of up to 1-3m are normal, due to the poor accuracy of these GPS coordinates. With so-called RTK drones, the gaps are normally less than 10cm
Set or move the photo block to a point
This function is useful for various special applications, e.g. for accurate transformation of image sets taken at different times. You can either enter a displacement or new coordinates for a freely selectable point, the fields in the dialog are then automatically updated accordingly.
Here, photo blocks can be manipulated, cameras can be recalibrated and faulty image measurements can be found.