Creating True Orthomosaics

264
1
Jump to solution
11-29-2023 01:13 AM
İlaydaUzel
New Contributor

 

Drone2Map, 

 Can we manually adjust the resolution when creating TRUE ORTHOMOSAICS from high-resolution images obtained at low altitudes?

·         Our camera calibration is not possible; how does your application identify the camera?

·         We notice distortions, especially at corner points, when creating orthophotos. Is there an option for manual intervention to resolve this issue?

·          We are combining RGB and thermal orthomosaics from different cameras in a GIS application, but they don’t seem to align. Is there a way to solve this in your application? We use geographic referencing in Quantum GIS software. Is geographic referencing possible in your application?

·         Does your program produce a projection matrix that converts 3D world coordinates into 2D pixel coordinates (this is most important for us)? If not, how can we obtain the intrinsic and extrinsic parameters? The interior and exterior rotation parameters provided by your software cannot enable us to run the projection matrix correctly. What kind of camera model do you use when performing the world to image coordinate transformation?

0 Kudos
1 Solution

Accepted Solutions
MarkBarker1
Esri Contributor

Hi @İlaydaUzel ,

For your first question on adjusting the resolution, I am assuming you are talking about adjusting the output product resolution. If so, that is a setting under the processing options. (Home tab> Options> General> Project Resolution. You can manually define a pixel size to make the output products lower resolution.


The camera model is identified through the image exif. Drone2Map will automatically detect the camera model based off the specific necessary metadata fields. If they are not present, then you should get a warning when you try to process the imagery that no camera model is present. If you are able to get to the point of processing the imagery then something must be getting populated in the Edit Camera pane (Flight Data tab> Edit Camera button). If the necessary camera metadata fields are not populating then you can try to manually define the camera model information, assuming you know what the values should be. See this blog for more info: target=_blankhttps://www.esri.com/arcgis-blog/products/drone2map/imagery/using-a-custom-camera-model...


I take it you are talking about distortions in the output true orthos edges? If so, then you could try to increase the Matching Neighborhood for images to see if Drone2Map will detect more tie points between neighboring images during the adjustment. (Home tab> Options> Adjust Images> Matching Neighborhood) Otherwise, if you are not able to get a better reconstruction on the edges and they still appear distorted you could clip those areas out: https://doc.arcgis.com/en/drone2map/latest/help/clip-area.htm Feel free to share the processing report for the project in your reply.


No, Drone2Map does not have georeferencing tools. Those tools are available in ArcGIS Pro and we allow you to quickly create an ArcGIS Pro project with the Open in ArcGIS Pro button on the Home tab.


Not sure if I fully understand this question, please let me know if the following explanation helps. The camera model used is whatever Drone2Map detects, or you define in the Edit Camera pane. For coordinate systems we use the same ones that are found in ArcGIS Pro, along with the same available transformations. You can install an optional ArcGIS coordinate system package that adds many more coordinate systems and transformations. These coordinate systems work for 3D or 2D datasets and you can overlay 2D layers within a 3D scene.

Regards,

Mark

View solution in original post

1 Reply
MarkBarker1
Esri Contributor

Hi @İlaydaUzel ,

For your first question on adjusting the resolution, I am assuming you are talking about adjusting the output product resolution. If so, that is a setting under the processing options. (Home tab> Options> General> Project Resolution. You can manually define a pixel size to make the output products lower resolution.


The camera model is identified through the image exif. Drone2Map will automatically detect the camera model based off the specific necessary metadata fields. If they are not present, then you should get a warning when you try to process the imagery that no camera model is present. If you are able to get to the point of processing the imagery then something must be getting populated in the Edit Camera pane (Flight Data tab> Edit Camera button). If the necessary camera metadata fields are not populating then you can try to manually define the camera model information, assuming you know what the values should be. See this blog for more info: target=_blankhttps://www.esri.com/arcgis-blog/products/drone2map/imagery/using-a-custom-camera-model...


I take it you are talking about distortions in the output true orthos edges? If so, then you could try to increase the Matching Neighborhood for images to see if Drone2Map will detect more tie points between neighboring images during the adjustment. (Home tab> Options> Adjust Images> Matching Neighborhood) Otherwise, if you are not able to get a better reconstruction on the edges and they still appear distorted you could clip those areas out: https://doc.arcgis.com/en/drone2map/latest/help/clip-area.htm Feel free to share the processing report for the project in your reply.


No, Drone2Map does not have georeferencing tools. Those tools are available in ArcGIS Pro and we allow you to quickly create an ArcGIS Pro project with the Open in ArcGIS Pro button on the Home tab.


Not sure if I fully understand this question, please let me know if the following explanation helps. The camera model used is whatever Drone2Map detects, or you define in the Edit Camera pane. For coordinate systems we use the same ones that are found in ArcGIS Pro, along with the same available transformations. You can install an optional ArcGIS coordinate system package that adds many more coordinate systems and transformations. These coordinate systems work for 3D or 2D datasets and you can overlay 2D layers within a 3D scene.

Regards,

Mark