Object Points can have a 3D position computed in many different ways. They are probably the most complex and flexible in this regard.
How a point can get a valid 3D location:
• it is a Control Point: the user provides the xyz.
• it appears on two or more oriented photographs: it is computed by intersection math or by global optimization math.
• it is a surface draw point that is drawn on a surface that has known 3D position and the point is marked on an oriented photo: it is computed by the surface draw plane intersect math.
• it is attached to a surface that is constrained by one of the coordinate plane constraints (horizontal, vertical front, vertical side) where one or more other points in that surface have known xyz: it is computed using the known points and the coordinate plane information.
• it is part of a "surface is planar" constraint where three or more other points in that surface already have known xyz position: it is computed using the known points and a computed plane equation (from the plane's known points).
• it is attached to a line that is constrained by one of the coordinate axes constraints (vertical, horizontal front/back, horizontal left/right) where the point at the other end of the line has known xyz position: it is computed using the known point at other end and the coordinate axis information.
• it is PhotoModeler's chosen origin point in a single photo constrained project when there are no control points (this is usually the point that is part of the most coordinate plane or coordinate axes constraints): PhotoModeler sets it to be 0,0,0.
• an offset point can be associated with an object point and, as long as the object point has a valid 3D position and the offset parameters are properly defined, the offset point will appear as a 3D point on photos or in the 3D Viewer.
Note: the situations where a point will not have a 3D position are: a) it is not a control point, nor part of a constraint and it appears on fewer than two oriented photographs, or b) it is not a control point, nor part of a constraint, it appears on two or more oriented photographs but the maximum angle of its light rays is smaller than the Bad Angle Threshold (as shown in the Audit Page of the Preferences dialog – see Preferences - Processing and Cameras).
When you combine these you get a very powerful set of capabilities. Examples:
• One photo, no control, lines and points with coordinate axes and plane constraints: All points attached to the chosen origin point with a constraint and all points attached to those points by a constraint will have computable xyz coordinates
• One photo, one control point, lines and points with coordinate axes and plane constraints: All points attached to the control point with a constraint and all points attached to those points by a constraint will have computable xyz coordinates
• One photo, some coordinate axes constraints plus a surface planar constraint: All points are computed similar to above and then for any coplanar constraint that contains three or more already computed points (i.e.. they define a plane that is not necessarily one of the coordinate planes (i.e. can be at an angle)) the remaining points in that constraint can be computed.
• Two photos, referenced points, and some coordinate axes constraints: All points imaged on two or more photos are computed by orientation and global optimization and then all points on one photo but attached to the other points by constraints are also computed.
• Two or more photos with no control points and no constraints: All points imaged on two or more photos are computed by orientation and global optimization.
For more information on Constraints see the section, Using Constraints.