Course Project: AT319 Lab 13 The Final Project

I. Intro

Date: 04/19/2023

This project will simulate real problems from gathering data to outputting results with everything we have learned so far. DJI Mavic 2 Pro, Pix4D Mapping, and ArcGIS Pro were used.

II. Gathering Data

A. Site and Weather Conditions

Site Condition:

Weather Condition:

Cloud Cover: 5%
Wind Direction: SE
Wind Speed: 11 km/h
Temp: 15 deg C

B. Flight and Mission Information

a. Flight Information:3

Location: PWA
Date: 04/19/2023
Vehicle: Mavic 2 Pro AI
Sensor:DJI Mavic 2 Pro Camera
Battery: Mavic 2 Pro
Flight Number: 1
Takeoff Time: 10:57
Landing Time: 1118
Altitude (m): 120
Sensor Angle: 80
Overlap: 80%
Sidelap: 80%

b. Crew

PIC: Dingming Lu
VO: Karl Oversteyns

C. GCP Placement and Collection

Five Aeropoint GCPs were used. They were placed and turned on with an order of 1, 2, 3, 4, 5, then turned off and picked up with 5, 4, 3, 2, 1. Then we open a hot spot at the field and uploaded the data to the server.

Go to Appendix Section A to see actual GCP placement in the field.

D. Flight Mission

E. GCP Data Upload

A laptop is connected to a hotspot of a cell phone and meanwhile creating local hotspot called “propeller” to let the GCP upload the data to the server.

II. Pix4D Process

A. Import Data

Referring to previous labs (eg. week 9 lab 8), 297 pictures were added. Also, GCP points (Northing, Easting, and Orthometric height ) were extracted from the CSV file which was downloaded from the Aeropoint server.

Figure 5: GCP data was imported into Pix4D Mapping

B. Initial Processing

Uncheck 2 and 3 → Start. After Initial processing is finished, check if the image quality is good or not.

C. GCP Calibration

Click each GCP and calibrate them on the right plane.

D. DSM and Orthomosaic

Reoptimize, and run 2 and 3.

III. ArcGIS Pro Analysis

A. General Overview

The folder with Orthomosaic and DSM data were connected to the ArcGIS Pro project folder.

A polygon was created to extract the analysis area. Extract by mask and hillshade tools were used.

Figure 11: Map 1: Flight area and GCP placement of the final project. This was only a part of the flight area.

Figure 12: Map 2: The actual GCP locations in the field.

Figure 13: Map 3: Compare DSM data and the ortho view.

B. Classification

Figure 14: Map 4: Initial Classification.

Figure 15: Map 5: fix the errors by reclassify the area.

Figure 16: Map 6: extract the roads from the classification by reclassify the layer and assign a value.

C. Analysis

We need to find these areas:

Requirement 1:
- Grass that within 20 meters of the road
- And 20 meters away from trees that are more than 10 meters high
Requirement 2:
- Grass that is 20 meters away from the road
- And 20 meters away from trees that are more than 10 meters high

To do this, we need to find trees that are more than 10 meters high by using when the DSM data is larger than the elevation + 10 meters.

Then convert raster road to polygon -> create 20 m buffer -> clip -> convert back to raster -> reclassify to have both 1 and 0.

Then convert tall trees to polygon -> create 20 m buffer -> clip -> convert back to raster -> reclassify to have both 1 and 0.

Extract the field from the classification.

To achieve part 1, we need the logic of (Area 20 m away from trees ) AND (Road buffer) AND (Field).

To achieve part 2, we need the logic of (Area 20 m away from trees ) AND (Area other than Road buffer) AND (Field).