Light detection and ranging (LIDAR)

Light Detection and Ranging (LiDAR) Optical detection and ranging (LiDAR)

What is it ？

Types of aerial sensors ？
There are active and passive sensors

Active sensors for topographic mapping
radar
To emit microwaves in a pulsed manner
Determine the distance from objects and their angular positions
（ From the side ）

LiDAR
Transmit optical laser light in the form of pulses
Determine the distance from the object

LiDAR The history of
1960 Laser ranging was developed in the s
1970 s LiDAR Topographic mapping
The original system was “ Single beam ”, Profile device
Early used for topographic mapping , Limited by the lack of accurate geographical reference
Early systems were used for sounding
GPS and inertial navigation （ measurement ） The development of the system improves the accuracy

LiDAR platform
In the air
For highly detailed local elevation data
satellite
Covering a large area , Less detail

Lidar operation theory
Send out a pulse of light , And record the exact time .
Detect the reflection of the pulse , And record the exact time .
Use a constant speed of light , Delay can be converted into " Oblique distance " distance .
Know the position and direction of the sensor , You can calculate the reflecting surface XYZ coordinate .

LiDAR The components of the system
Laser scanner
High precision clock
GPS
IMU- Inertial navigation measurement unit
Data storage and management system
GPS Ground station

LiDAR The components of the system
frequency ：
Per second 50,000（50k） to 200,000（200k） Pulse （Hz）
wavelength
Infrared （1500-2000 nanometer ） Used in meteorology - Doppler lidar
Near infrared （1040-1060 nanometer ） For land surveying and mapping
Turquoise （500-600 nanometer ） For bathymetry
Ultraviolet (uv) （250 nanometer ） Used in meteorology
Safe for eyes ; Low wattage （<1 watts ）.

Electromagnetic spectrum

How lasers work
High voltage electricity makes the quartz flash lamp emit strong light , Excite some atoms in the cylindrical Ruby crystal to a higher energy level .

At a certain energy level , Some atoms emit light particles called photons . At first , Photons are emitted in all directions . Photons from one atom stimulate the emission of photons from other atoms , The intensity of light is rapidly amplified .

Mirrors at both ends reflect photons back and forth , Continue the process of stimulus emission and amplification .

Photons leave through a partial silver mirror at one end . This is the laser . The light waves emitted are in phase with each other , And almost parallel , So that they can travel long distances without spreading .

LiDAR The components of the system
Scanner
The mirror rotates or scans , Project laser pulses onto the surface
The scanning angle can reach 75 degree ; The scanner measures the angle at which each pulse is emitted
Receive reflected pulses from the surface （“ Echo ”）.


LiDAR The components of the system
Global positioning system （GPS）
Record the scanner x、y、z Location – Investigate the ground base station in the flight area
Inertial measurement unit （IMU）
Measure the angular direction of the scanner relative to the ground （ pitch 、 rolling 、 Yaw ）

LiDAR The components of the system
The clock
Record the time when the laser pulse leaves and returns to the scanner


airborne LiDAR Principle – Flight plan
Can't penetrate the clouds
Usually fly at night
On steep terrain there are 30-50% Overlap
Pass through many times in urban areas from different angles （ To avoid LiDAR “ shadow ”）
The flight altitude is usually 200-300 rice （ Higher in urban areas ）
Each laser pulse emitted by the scanner will receive multiple “ Echo ”
Modern systems can recode up to five returns per pulse


LiDAR Principle – Echo
The distance between the sensor and the surface object is calculated by comparing the time the pulse leaves the scanner with the time it receives each return .

LiDAR Principle – Echo
Of each echo x/y/z Coordinates are the position and direction of the scanner （ come from GPS and IMU）、 The angle of the scanning mirror and the range distance to the object are calculated
The set of echoes is called a point cloud

Laser pulse travel time
for example ：
The flight altitude is 300 rice , The object is directly below the sensor 10 Meter high . How long does the laser pulse travel ？1.93467 X 10-6 second （ or 1,935 nanosecond ）.

Spatial range map of data in lidar file
Shaded 、 The fringes in digital units correspond to a flight data in a single lidar file , Odd and even stripes fly in opposite directions .

Maximum unambiguous range （ Pulsed laser ）
Laser energy
Pulse rate （ Non overlapping pulse ）

SPiA And MPiA（ Air monopulse and air multipulse ）
MPiA It's a technology , It is allowed to emit the next laser pulse before receiving the reflection of the previous pulse .

LiDAR The resolution of the /DEM The resolution of the
Laser field of view （FOV）（ Or footprints ）
Ranging range and ranging resolution
Point density / Point spacing

LiDAR- FOV（ Or footprints ） Big or small ？
FOV Related to beam divergence （0.1 To 1 Milliradian ）
Small FOV For detailed local mapping
Big FOV For more complete ground sampling and more interaction with multiple vertical structures
Big FOV This usually results in a low signal-to-noise ratio

Pulsed laser Ranging distance （R） And ranging resolution （ΔR）
among ：c： The speed of light （~299,792,458 rice / second ）
t： send out / Time interval between receiving pulses （ns）
Δt： Resolution of time measurement （ns）.

LiDAR Principle –“ The resolution of the ”
Number of pulses per unit area
The current system can achieve 20+ pulse / Square meters
The resolution is determined by the aircraft speed 、 Flight altitude field of view （FOV）、 Pulse emissivity determines
The spacing of points is not uniform

LiDAR Principles –“ The resolution of the ”
Higher resolution and narrower FOV To penetrate the dense vegetation
Higher resolution can better distinguish surface and surface features , But the cost is that the data set is large , Processing time is slow

Lidar density and DEM The resolution of the – Every time DEM Pixel average 1 Lidar pulses
Point density （ Such as per square meter 8 Pulse ）
Point spacing （ Such as 50 centimeter ）
PS = SQRT(1/PD)
for example ：8 Pulse / rice 2 = 0.35 rice

LiDAR Principle
precision
The vertical accuracy is usually 15 to 20 centimeter （ about 6 Inch ）
The horizontal accuracy is 1/3 to 1 rice
Through low altitude slow flight and narrow FOV Improve accuracy

LiDAR Principle Strength
The intensity of the echo varies with the composition of the surface object on which the echo is reflected
The percentage of reflection is called LiDAR Strength
Can be used to identify land cover types
The intensity value needs to be normalized in each flight

LiDAR The advantages of
All data is georeferenced from the beginning
High precision
It can quickly cover a large area
Compared with photogrammetry , Fast turnover , Low labor intensity , The cost is low
You can collect data in steep terrain and shadows
Can produce DEM and DSM

LiDAR The shortcomings of
Unable to penetrate very dense canopy , Cause elevation model error
Very large data sets , Difficult to explain and deal with
There is no international agreement
cost
200 dollar -300 dollar / Square miles -3 Meter resolution
350 dollar -450 dollar / Square miles -1 Meter resolution

LiDAR Data preprocessing
Data is collected by the onboard computer in a format proprietary to the system supplier
After treatment , To calibrate multiple flight lines , Filter out incorrect values and noise
Classify the returns , And separate them by category ： First return 、 The last time （ Or naked ） Returns , wait .

LiDAR data format
The point value is usually determined by the supplier in ASCII Click file or LAS Format delivery


LAS Format
LiDAR Data exchange format standard
Common binary file format , Keep the data LIDAR Specific information about the nature , And not too complicated
from ASPRS maintain
http://www.lasformat.org/

LAS File components
Common header
Variable length records （VLR）
Projection 、 Metadata 、 Wave packet . User defined data
Point data record
Format 0（20 byte ）- General attribute
Format 1（28 byte ）- Format 0+GPS Time （8 byte ）
Format 2（26 byte ）- Format 0+RGB（6 byte ）
Format 3（34 byte ）- Format 2+GPS Time
Format 4（57 byte ）- Format 1+ Wave packet （29 byte ）
Format 5（63 byte ）- Format 3+ Wave packet
Format 6- 10- For system >15 return
Extend variable length records （EVLR）


LiDAR Processing software
QT Modeler
TerraScan
ArcGIS (Workstation, LiDAR Analyst, 3D Analyst, LP360)
Leica Photogrammetry Suite
ENVI LiDAR

Typical LiDAR To DEM Processing steps of

1. take " original " Point import GIS Format
2. Convert a point to a surface TIN Model
3. take TIN The model is converted to a raster model of the earth's surface
   
   The first 1 Step – Import point
   Multiple LiDAR x/y/z The return is converted into a separate GIS Data sets
   Typical " Generate " function
   Large data sets make many GIS The application is overwhelmed
   
   
   The first 2 Step – establish TIN Model
   Triangular irregular network
   The Delaunay triangle method is usually used to create , All points are connected to the nearest two points
   Delaunay triangulation is as equiangular as possible
   Make sure that any point on the surface is as close to the triangular node as possible
   
   The first 2 Step – establish TIN Model
   x/y/z The return point becomes TIN Node of model triangle
   So the slopes of the sides and faces of the triangle are known
   TIN The model allows linear interpolation of elevation values between triangular nodes
   Better than converting point returns directly to raster data “ edge ”
   
   

The first 2 Step – establish TIN Model
TIN Models are useful representations of the earth's surface
about LiDAR data , Because of its low complexity , Faster drawing , Generally, the grating model is used as the final product



The first 3 Step – Create a grid model
Raster data is represented by a series of regular uniform data units （ Pixels ） Store elevation values
The surface model based on grid is called digital elevation model （DEM）
The above ground feature model based on grid is usually called digital surface model （DSM）

The first 3 Step – Create a raster model
Raster data can be obtained from TIN Create... In the model , The method is to use the linear interpolation method to interpolate the elevation value of each pixel center point
The minimum resolution depends on LiDAR Resolution returned – Rule of thumb ： There is an average of one pulse per pixel






Other products – Height image
Create by subtracting the bare return from the first return – Create trees 、 Raster images of buildings and other surface features



Other products – Sun visor
Shadow relief image created by considering the illumination angle of the sun and shadow
Used in 2D View in 3D Model
The light source usually comes from the north ; This produces the most visually appealing image



Other products – contour
It can be downloaded from TIN Or grid model
The appropriate interval depends on LiDAR Vertical accuracy of data .( The interval between isolines shall be at least twice the vertical accuracy , namely




The public domain of Oregon LiDAR
http://www.blm.gov/or/gis/lidar.php
OSU ftp: ftp://lidar.engr.oregonstate.edu/
USGS NED (3m DEM) and Earth Explorer (las)
ESRI ArcGIS
ArcGIS Point file information tool
ArcGIS LAS Dataset tools

ArcGIS 10.X LAS Data sets
One LAS Data sets can
stay ArcGIS Use in ArcMap and ArcScene Of 2D and 3D function .
According to some lidar filters applied to point clouds , Use the elevation or point attribute renderer to display it as points .
Surface models rendered as triangles .
According to some lidar filters , Use elevation 、 slope 、 Angle or contour line for visualization .
For source LAS File update

Link to the original text
Extraction code ：zio0