All recording methods have their advantages and disadvantages.
Working on an archaeological site, different techniques should be
combined to benefit from their individual strengths. The decision is
generally based on the object, required level of accuracy and detail,
the objectives as well as the project's budget. It is often observed,
that technologies are chosen or used in such a manner, that is not
suitable for the specific situation on site. To make an informed
choice, the archaeologist should know some of the premises of each
Data acquisition for photogrammetry is fast and can be easily
integrated into the standard work flow of recording, as shown also
drawing of sections and plans is accelerated on the base of photos
in the field. Some of the pressure on site, can so be taken into the
Often, it is believed that laser scanning is faster as the outcome of a scan is a 3D point cloud. It is true that a laser scanner target="" that for the point clouds gathered by laser scanning spatial registration is required and several point clouds have to be joined to a single point cloud. The cleaning of outliers in unstructured point clouds takes up a considerable amount of time as well as triangulation of the mesh and the texturing with photographs to give the model a photo-realistic look. This qualifies the opinion that laser scanning has advantages in time versus photogrammetry. In both cases the time pressure is taken away from the site into the office.
Another subject to critique is the cost. With the introduction of digital photography and new software, photogrammetry has become to a cost-effective documentation method, including the hardware (a good DSLR-camera with wide angle lens dedicated to photogrammetry and a computer for processing) and professional software ca. 5,000 € have to be factored into a project budget while a ToF laser scanner including software to date still costs around 100,000 €. Compared with other cost arising, when conducting an excavation, this is low and the time-savings compensates the investment soon.
It is wrong that laserscanning is more accurate than photogrammetry. As for many other projects, the quality of the input data (metric survey of control points, quality of the imagery, camera calibration etc.) governs the quality of the output. Garbage in, garbage out.
A point often raised by the crtitiques of image-based recording is that photogrammetry is preventing the archaeologists from interpretation. Certainly, this might be the case, but this is rather based on the attidute of the recording archaeologist than in the method. It is a must that orthophoto production is carried out contemporary to the excavation progress to provide the archaeologist with a product, on which he can map his findings in the field. Sufficient staffing and resources have to be planned in to perform this task. - It is of advantage that below the interpretive drawing still a naturalist and thorough record exists.
It is advantageous, that parts of a section or an elevation, which are difficult to access, can be recorded without being touched. Working on scaffolding and unstable ladders can be avoided or the time spent on them, minimized. At tell sites, high dangerous sections often occur; especially, when old trenches of previous excavators are reopened. Photogrammetry reduces danger for the recorder and enhances health and safety matters on site in such situations.
One main difficulty arises because archaeological objects have hidden parts (self occlusion) or they are obscured by other objects (ambient occlusion). In such cases the standard automatic digital surface model (DSM) production algorithms fail to produce a useful product and if possible additional photographs have to be taken to cover these parts. This problem equally occurs with range-based surveys, e.g. laser scanning.
As for all photographic records, an equal and constant lighting is required. Direct sunlight creating strong shadows leads to bad results. Automatic image matching and surface creation algorithms might misinterpret these strong shadows and produce unexisting peaks and features in the 3D-model.
Flash light cannot be used. As theposition of the light source changes within one stereo-pair, image matching algorithms will fail.
To obtain quality projects, it is inevitable that the archaeologist, carrying out such surveys, is trained in the technology. Miissing knowledge often leads to wrong methods used for certain projects.
Further disadvantages originate in the individual method:
A clear disadvantage of rectified photography is its restriction to planar areas, like sections. All objects on different planes have a perspective error; they are displaced and and are either lengthened or shortened. Features not parallel to the main rectification plane have to be rectified individually. While features parallel to the rectification plane can be scaled with just two target points. The subsequent evaluation drawing have to take this into consideration.
A problem in modelling from stero-photogrammetry, originating in surfaces with poor definition or the same texture. This rarely occurs on archaeological features and is restricted to some kind of finds, such as glazed pottery, glass or metal.
Correctly used, the different photogrammetric techniques are a very powerful tool for recording archaeological features and artefacts. The basic requirements and restrictions of each method, especially photogrammetry, as well as the most favourable work-flow on site are given in the folllowing sections.