The implementation of post-processing data thinning for multibeam echo sounding data

Abstract

Multibeam echosounder (MBES) is considered the best innovation in depth measuring technique if full seabed coverage is the foremost concern. Multiple beams are generated through MBES transducer in a fan-shaped swathe across vessel track within fraction of second. From few to hundred of beams in every ping can be transmitted into different sections or angles from nadir beam direction then translated into multiple depth measurements. The rate of ping is subjected to survey area where the shallower the area is, the higher ping rates is expected compare to deeper area. Thus higher MBES data density is produced in shallow water area. In this shallow zone, over sampled or redundancy of MBES data are common. These over sampled data are too dense to be displayed and in consequence to represent the survey area on bathymetric plans. Another factor is these over sampled data make Digital Terrain Modeling (DTM) and contouring more intense especially on the computer processing software. Therefore these dataset should be reduced in term of it size. The process to reduce the size is called data thinning. Data thinning algorithms should be capable in handling high volume of MBES data. In the process to reduce the dataset, one must bear in mind that the process would not jeopardize the integrity and accuracy of final product. To adhere on these requirements, the resolution of the MBES used during the survey and the smallest expected detail to be mapped must be taken into considerations. This paper elaborates on the development of data thinning programs using Microsoft Visual Basic Version 6. Various algorithms namely Douglas Peucker, Single Swathe Reducer and Skip N Points are referred. Comparisons on the final results based on these three algorithms will be discussed in order to decide which algorithm is most favorable to be used in post-processing data thinning for cleaned MBES dataset. Finally the paper summarizes some of the distinguishing features of this data thinning approach.