Remote Sensing: LIDAR: Automated Classification

Spectrum Mapping, LLC utilizes our proprietary pioneering technology of SILC (Spectral Imagery LIDAR Composite) to integrate digital imagery elements with LIDAR surface data. Associating multispectral pixels with individual X,Y,Z values allows true discrimination and automatic classification of roads, buildings, vegetation, water and other features.

SILC is now possible only with Spectrum Mapping's integrated LIDAR and Color / CIR digital camera systems. The precise calibration and alignment of these two subsystems are vital for the software to calculate, display and manipulate spectrally attributed formatted LIDAR data.


Traditional Feature Extraction

Automated feature extraction has long been a goal for the photogrammetric mapping industry. To date no method has been perfected to automatically extract features such as building footprints, edge of roads and walkways, vegetation, or bodies of water. The process requires very high-resolution imagery, with analysis performed largely in a 2-D environment requiring significant manual editing. Incorporating 3-D data and information into the process would greatly aid in analysis and extraction efforts.

Like the photogrammetry industry, LIDAR providers have faced similar issues. Spatial analysis alone is often inconclusive when attempting to determine whether a laser point has hit a small bush, fire hydrant, boulder or a ground surface anomaly. Conventional surface classification filters can only go so far when removing above-ground phenomena, as natural and artificial objects may be spatially interpreted similarly and objectively removed, inadvertently eliminating valid surface detail without differentiation.


SILC

The SILC process allows multispectral pixels to be photogrammetrically associated with individual X,Y,Z values. The imagery is not ‘draped’ to the surface, rather each laser return is mathematically projected through collinearity equations onto its proper position on the frame array taking into consideration the camera model using photogrammetric techniques.

Because image pixels outside the near-nadir area cannot be accurately correlated with the laser points, traditional film cameras are impractical for the collection of imagery for this application. Spectrum Mapping’s medium format digital camera system is ideal for this innovative treatment of LIDAR data. The longer focal length and smaller field of view of our digital camera virtually eliminates building “lean” and matches the swath width of the laser system allowing the proper alignment of laser data and pixels.

By using imagery acquired simultaneously with the surface data, each surface point possesses an accurate spectral signature assigned to its location, allowing accurate classification of features using conventional remote sensing techniques. Through the element of color, SILC data allows urban terrain, forested terrain, agricultural lands, mountains, cliffs, ravines, and wetlands to be subjectively masked and filtered, providing vastly improved bare-earth surfaces and feature extraction.

LIDAR Returns Colored by Elevation

Partial flight line in LAS format LIDAR points color coded by elevation based on chromadepth. Trees, buildings and roads are fairly discernable Are the little bumps near/on the roads cars?

RGB Enhanced SILC Data

Partial flight line in LAS format LIDAR points are attributed with RGB values and displayed much like a 3-D real-time ortho which can be edited Trees, buildings and roads are obvious Structures, cars, paint stripes, parking lots, walkways and park benches are easily identified

LIDAR Returns Colored by Elevation Perspective View

LIDAR Returns, SILC Enhanced, Perspective View

LIDAR Returns Colored by Elevation	CIR SILC Enhanced Data

LIDAR Returns Colored by Elevation Perspective View

CIR Enhanced SILC Data Perspective View

info@specmap.com