School of Surveying and Spatial Information Systems
The University of New South Wales
Evaluation of Leica DISTO:
Accuracy and Precise 3D Indoor Surveys
By Jarrod Hocking
Supervised (and edited) by Assoc. Prof. J. M. Rüeger
My thesis investigated the capabilities of two hand-held laser distance meters. The first model is the Leica DISTO lite(5) and the second model is the Leica DISTO classic(5). The first aim was to establish the precision of the DISTOs. Leica states that the accuracy of both instruments is ±1.5 mm and ±(1.5 mm + 2.5 ppm) for longer measurements, both at one standard deviation. There are plenty of testing facilities available for electronic tacheometers (total stations), but for the DISTOs unique facilities had to be designed. The second aim was to investigate the use of the DISTO classic(5) for 3D precise indoor surveys.
A simple range test was performed using a camera tripod and reflective sheeting for targets. The range tests revealed that both instruments do not measure outside their stated range of 0.2 m to 200 m. It is thought that there is some kind of cut off that prevents measurements longer than 200 m.
Short Periodic Error
The standard procedures of testing for cyclic errors in electronic distance meters were employed for the DISTOs, with a few differences. These included a specially designed mounting block and a wooden target, that could be aligned and levelled. It was found that the short periodic ('cyclic') errors of both instruments were too random to model between a measuring distance of 2 m to 23.5 m. The maximum significant cyclic error correction never exceeded ±0.86 mm. No cyclic error corrections were applied during the other tests. The precision (for a mean of four measurements), obtained from the short periodic measurements, were as follows:
DISTO lite(5) = ± 0.37 mm
DISTO classic(5) = ± 0.46 mm
Figure 1: Layout of the cyclic error and back-to-back tests.
To perform the additive constant tests, a primitive baseline was established down a corridor of 44 m length (see Figure 2). The same mounting block was used, together with a second purpose built target plate. Tests on the baseline established that the additive constant of the DISTO lite(5) was insignificant (+0.07 mm) while the additive constant of the DISTO classic(5) was significant (+1 mm). The precisions (for single distance measurements), obtained from the additive constant measurements, were as follows:
DISTO lite5 = ± (0.47 mm + 1.09 ppm)
DISTO classic5 = ± (0.47 mm + 0.94 ppm)
Figure 2: Baseline design for the additive constant test.
Variation of Additive Constant with Distance
A back-to-back test was conducted using the DISTOs against a NIKON NPL-821 reflectorless electronic tacheometer and a steel tape (see Fig. 1) to see if the additive constant changes with distance. The back-to-back tests were conducted twice because a strange phenomenon occurred. The differences between the DISTOs and either the NIKON or the steel tape seemed to peak to about +3.5 mm between 13 m and 18 m. Re-measurements conducted over this range did not reproduce this peak (see Fig. 4). This phenomenon remained unexplained.
Figure 3: Cyclic and Back-to-Back test line.
Figure 4: Original measurements (diamonds) and re-measurements (squares) of classic(5) reduced minus tape reduced, where the x are in m and the y are in mm. Field temperatures: 18.1 °C on first day and 18.3°C on second. The test line was completely in the shade on both occasions.
3D Precise Indoor Surveys
For the 3D precise indoor survey, a room was measured three times, once with the DISTO classic(5) and, for comparison, twice with a NIKON NPL-350 reflectorless electronic tacheometer. Special attachments were designed to measure the 2D and 3D space diagonals using the DISTO classic(5). Corrections were derived for these attachments so the distance from corner to corner could be established. The three independent indoor surveys were transformed onto the one datum using a transformation program. Differences could then be taken between the X, Y and Z coordinates. These differences were much larger than expected. Even between the two NIKON surveys, the results agreed to about 39 mm. Comparing the DISTO survey with the two NIKON surveys revealed differences of up to 48 mm. The comparisons were then redone, including the corners but excluding the features along the walls (i.e. doorways and windows). This time, the two NIKON surveys revealed differences of up to 15 mm. Comparisons between the DISTO survey and the two NIKON surveys revealed differences up to 14 mm. The poor repeatability of the surveys was thought to be caused by poorly defined corners. As most indoor corners are rounded, the fundamental limitations of both methods is the definition of (and the measurement to) corners.
Conclusion and Recommendations
The DISTOs performed much better than expected. The precision, that was determined, was about three times better than the accuracy stated by Leica. This is very impressive for a little hand-held laser distance meter. However, I do not recommend using a DISTO classic(5) for precise indoor surveys for a number of reasons. A large number of measurements is required to establish 3D coordinates by distances only. This is very time consuming. Also, attachments have to be made, corrections have to be applied, two people are required for the fieldwork and there is the reduction in accucay due to hand-held operation. Using reflectorless electronic tacheometers is be a better option, if an a better way of measuring to corners can be found. Further research into the distance variation of the short periodic errors of the two instruments and into a better definition of (and measurement to) corners is recommended.
For more information, please contact:
Assoc. Prof. J. M. Rüeger
School of Surveying and Spatial Information Systems
University of New South Wales
UNSW SYDNEY NSW 2052