FIGURE. THE THREE COMPONENTS OF ECHOES FROM WATER IN THE GREEN CHANNEL CONSIST OF:  SURFACE RETURN, WATER VOLUME BACKSCATTER AND BOTTOM (LEFT) AND AN EXAMPLE OF ATTRIBUTES EXTRACTED FROM THE WAVEFORM (RIGHT)

The workflow addresses Lidar WaveForm coastal Feature Extraction. The objective is to extract, from the bathymetric Lidar full waveform echo, parameters such as peak height, pulse width, pulse area etc. which are useful for the classification of seabed types (substrate and vegetation). The workflow is focused on two key aspects of functionality that can be used in the marine test beds.
•    The first is a service to perform both sequential and spatial indexing and initial control of the data delivered from an airborne lidar bathymetry survey system. This allows for direct access to specific PDR or WF files given specific geographic coordinates is created based on the index results from the PDR data.
•    The second is to contribute to the development of services for characterizing coastal seabed and detecting vegetation using attributes extracted from bathymetric LiDAR waveforms.

The result of the workflow is a point cloud of all computed attributes (Pulse width, Pulse area, Peak amplitude...) that could be interpolated and classified by end users to help seafloor habitat mapping. The workflow has a major role in production of a seafloor 2D map with labels corresponding to identify sea bottom types using Bathymetric Lidar Full WaveForm Data. It offers a full automatic processing opportunity to enhance bathymetric Lidar data exploitation capabilities. 

A scalability test was performed for five different data sets with varying sizes from 57.69 MB to 27361 MB. The average runtime of three executions shows that performance is stable at the higher tested data sizes. The processing speed of the workflow is fairly constant, ranging from about 0,5 MB/s to 0.7 MB/s for the larger data sizes. Parallelization reduced the execution time by a factor of 5.