Field surveying for photogrammetric control is generally a two-step process. The first step consists of establishing a network of basic control in the project area. This basic control may consist of horizontal control monuments and benchmarks of vertical control that will serve as a reference framework for subsequent surveys. The basic control also can be established by GPS techniques. The second step involves establishing photo control by means of surveys originating from the basic control network. Photo control points are the actual points appearing in the photos that are used to control photogrammetric operations. Photo control points are connected to the basic control framework by short spur traverses, intersections, and short level loops. Photo control points are surveyed to the accuracy required to control the photogrammetric solution (for example, Third-Order, Class II type plane surveys are sufficient to establish photo control points). The accuracy of basic control surveys is generally of higher order than subsequent photo control surveys.
Control point distribution
The number of ground control points (GCPs) required and their optimum locations depend upon the use that will be made of them.
If photo control is being established for the purpose of orienting stereo models for topographic map compilation, the minimum amount of photo control is three vertical and two horizontal control points. However, some amount of redundant control should be used as a check. Thus, as a practical minimum, each stereo model oriented in a stereoplotter or a softcopy workstation should have three horizontal and four vertical control points. The horizontal points should be widely spaced. Horizontal points in opposite diagonal corners of the neat model are optimum. Vertical control points should be in the corners of the neat model. Vertical control should never be along a single line in the neat model. A fifth vertical control point in the center of each stereo model is useful as a quality control check for stereo model deformation.
In the traditional distribution, one should collect three GCPs in the first model, then use pass points only for a number of photos, then three GCPs in the fourth or fifth model. For a strong "bridging" between GCPs, pass points should be collected in the area of 3-way overlap between the images. Technically, one can triangulate a block with as little as 3 or 4 control points, but the accuracy would be very low. GCPs should be well distributed within the block. Again, aim for at least a few photos with a minimum of 3 GCPs, as this will provide give good solid leveling. Additional GCPs should be available as independent check points for checking the accuracy of the bundle block adjustment (Figures 2 and 3).
Figure 2. Traditional Control Point Distribution
Figure 3. Control Point Distribution for Blocks With 20% - 40% Side lap
In the case of 60% side lap, the number of vertical control points can be reduced to an equal distance in both directions of approximately 4 to 6 photo base lengths. The number of control points can further be reduced by flying additional cross strips (Figure 4).
Figure 4. Control Point Distribution for Blocks With Cross Strips
Combined Block Adjustment with GPS/IMU
When airborne GPS is used, the most economic possibility for the reduction of the number of control points is to use cross strips in the area of the control points. In this case, the required distance between control points is the same as the case of 60% side lap, but a smaller side lap is allowed.
For a combined bundle block adjustment with projection center coordinates determined by kinematic GPS positioning, control points are required only in the block corners. For sufficient reliability, two control points should be located in each corner. If cycle slips of the GPS-positioning cannot be avoided, two cross flights or a line of vertical control points at both sides of the blocks are required (Figure 5).
With GPS projection center coordinates, the adjustment can be handled with as few as four control points. If additional parameters were used to compensate for systematic errors, additional horizontal control points are required. Additional and well-distributed control points increase the reliability and accuracy of the blocks. Furthermore, the accuracy of combined bundle block adjustment depends on the quality of the GPS data. Therefore, careful planning, multiple and accurate base stations, a high-quality GPS receiver, and good software for post processing are needed.
An inertial measurement unit (IMU) is optional and in many instances does not significantly improve the absolute accuracy of block photography. IMU data is helpful in supporting the bundle block adjustment in the case of long single flight strips. Single flight strips supported only by GPS data of the projection centers can extend the distance between control points only up to 10 photo base lengths. Beyond a ten-model gap in control, the strip can rotate around the line of the projection centers. But this can be avoided by incorporating IMU data.