Bible Meadow first analysis

On Saturday 21st January 2015 an aerial drone survey was carried out over Bible Meadow, Holt, using a DJI Phantom 3 standard drone. A flight plan was created on Dronedeploy and 107 geo-referenced images were produced.

Below are the first set of results to show you the type of analysis we can do from the images the drone produces.

Analysis.
I will let you draw your own conclusions but I can see:

1. the area is a bowl shape dipping by 3m from the centre to the edge of the analysis
2. the small field boundary shows well on some analysis
3. there are some lines in the main field that look to be at right angles to each other.  They show on the elevation so must be a physical feature.  Field drains would be an obvious one, but I am not sure

enjoy C

Dronedeploy

Flight plan used to automatically survey Bible Meadow

The images are uploaded by the Dronedeploy app and below is some of the output.

107 photos stitched together to produce a Georeferenced, orthorectified map  - accurate to with a cm. This will automatically drop into place on Google Earth

Elevation model.  think contours, blue is low red is high.  If a feature shows in this analysis then it is a bank or ditch.  Of course the hedge line shows up in yellow/red as high ground

NDVI crop health analysis - looks at vegetation with red/green/blue filters.  may be like crop marks but not visible to the naked eye

A different setting for crop health analysis - looks at vegetation with red/green/blue filters.  may be like crop marks but not visible to the naked eye

3D model of the landscape which can be used as a starting point for further anaysis

Pix4mapper

I used a second software called Pix4mapper which is a free photogrammetry software.  It can be used as many times as we want.  It produces a 3D model and lets you preview the overhead shot and elevation model,

Global Mapper

Very expensive program, but you can do some work with the free version.

Contour map 1m - can set to smaller inervals

Fish Ponds - Elevation model

I have managed to create a .dem (digital elevation model) from the aerial photo taken from the drone video captured last week.

This shows the 'fish pond' dimensions the clearest yet.

The high colours on the left are the bushes.  You can also see contour lines on there, but the field is too flat for these to be much use.

First results from drone flight over Croxtons

On Saturday 21st Jan 17 we flew a drone over Croxton's Field to product aerial photographs and digital elevation models to see the condition of the trenches dug over the past few years.  These are some of the early images that have been produced.

1. Aerial photo

2. Crop health (trenches in red)

3. Elevation model (remember the field is quite flat)

4. Slope direction shader

Aerial photography comes to Waterless Brook Project

Last Saturday I attempted my first archaeological survey by drone.  Lots more to learn, but I have managed to get some decent results from the first flight. I didn't get the drone exactly over the middle of the feature so the smaller cell is not as clear as the larger one.  All part of the learning curve.

Basic analysis - Overhead shot (Fig1)
Hover over a feature as low as a few meters to over 100m.

Extract from Video (Fig2)
Using a 'Point of Interest' feature the drone will circle round filming the ground.  Images are then extracted from the video (every 4 seconds) which will be used to produce a 3D model by photogrammetry.
These give a more angled image.

Create a 3D model (Fig3)
Using photogrammetry software a selection of overlapping images taken from all angles around a feature can be used to produce a 3D model.

This model is about 100 times more detailed than the one I have tried to create from lidar, so very promising.


Next Steps
Now that I can produce a detailed model (point cloud) from the video, I need to find a way to convert it back to raw data, so that we can do the hill shade and elevation analysis to produce drawings, possibly to within a few centimetres accuracy.

Fig1: Overhead shot

Fig2: Frame from video

Fig3: 3D model created by photogrammetry software