Lidar Data

Introduction

The X, Y, Z coordinate of each return is calculated using the location and orientation of the scanner (from the GPS and IMU), the angle of the scan mirror, and the range distance to the object. This produces a collection of returns which is known as a Point Cloud. A point cloud is merely a collection of points with 3-dimensional coordinates.

Returns

The laser pulses reflects off of objects on and above the ground, which includes buildings, trees and other vegetation. Lidar systems can collect multiple returns from the same laser pulse. Multiple returns allow for lidar data to be used to separate the bare terrain from the surface features. Most modern Lidar systems are capable of recording up to 5 returns from each pulse. For example A single pulse may reflect from upper, middle, and lower tree canopy as well as the ground beneath, resulting in multiple measurable returns from that pulse. With each object the pulse hits a certain amount of the light is reflected back and detected and recorded by the sensor.

The first return is the most significant return, and will be associated with the highest feature in the landscape like a treetop or the top of a building. In a forested environment the first returns represent the canopy. Often the last return is the ground but not always. In heavy tree canopies the pulse may not reach all the way to the ground. For example if you could not see sunlight through the canopy there is a good chance that a lidar pulse won't reach the ground. The first return can also represent the ground in areas which have no vegetation, in this case only one return will be detected by the lidar system.

Each return has an X,Y,Z position as well as other attributes. The returns are flagged in several ways, by the return number (i.e. 1 to 5) and standard classes. These classes include Ground, Above-ground (vegetation, building etc), Water and Noise. Not all lidar data is necessarily classified this way. Software has been developed to aid in classifying lidar points. For example, generally if only one return pulse was received this indicates that the pulse hit a solid feature like the ground or a building. Most lidar data that is contracted or acquired from an online source has already been processed and classified to some degree.

Data Characteristics

The resolution of a lidar dataset is determined by the pulse rates, the aircraft speed, flying altitude and field of view of the system. Each point can also have a variety of attributes.

Dataset Attributes

Pulse Rate
The pulse rate is the rate at which the lidar sensor emits pulses. A higher pulse rates produce denser data with more points per unit area. The pulse rates typically range from 10,000 to 200,000 pulse per second.

Point Spacing and Density
The point spacing is the linear spacing of the points in the dataset. The measurement is a given in linear units, for example a dataset might have an average point spacing of 0.25 meters. The point density is the density of points per area or the number of points in a given area. Point density can be estimated from the point spacing using the following equation: Density = 1/(Point Spacing)² and is represented as points per given area, for example 16 point per m². Both point spacing and density are given as an average for the dataset.

Point Count
The point count is the total number of points in the lidar dataset. The larger the area the greater the point count. A high resolution dataset (smaller point spading, greater density) will have a greater point count compared to a lower resolution dataset. The point count can be found in the metadata or file attributes and most datasets contain millions of points.

Point Attributes

In addition to the position (the X,Y,Z coordinates) of the point, individual points can have additional attributes.

Intensity
Intensity is a measure of the return strength (magnitude) of the laser pulse. The strength of the returns varies depending on the reflectivity of the surface object. For example, in a near infrared lidar system points that represent vegetation would have a greater intensity because plants reflect greatly in the near infrared. Intensity is used to help in feature detection and extraction, and in lidar point classification. Intensity values can be used to create an image that resembles a panchromatic photograph.

Intensity image created from lidar data. Image Credit : ESRI.

Return Number
Pulse have multiple returns and each returns is flagged with the return number (usually 1 through 5). Each point in the lidar dataset will have an associated return number.

Classification
Lidar points that have been processed are also classified by the type of object they represent. Lidar points can be classified into a number of categories including bare earth or ground, top of canopy, and water. Below is the table of full classification options as defined by the American Society for Photogrammetry and Remote Sensing (ASPRS).

← Back

Next →

Module Home