![\"\"](\"https:\/\/www.keris-studio.fr\/blog\/wp-content\/titre-1.jpg\")
<\/a><\/p>\nWhen information about a site is not available, it is interesting to make a survey by drone to realize a photogrammetry. The photogrammetry will allow to build a 3D model and to extract the necessary information to understand the site.<\/p>\n
The preparation of the drone flight is essential. The flight must be able to cover the entire site from different angles. Otherwise, there would be deformations or missing information.<\/p>\n
The example presented here presents the workflow for creating a 3D model from an aerial film shot with a drone and the processing of the data for further work.<\/p>\n
Here, the drone was forced to work in sections. Here are the different views. The work to come will be to assemble the 3D file fragments.<\/p>\n
Location<\/h1>\n
We will work from a drone survey done at the location Batterie de la Reine situated here : -20.375756949812796, 57.728543193103164<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
<\/p>\n
extraction of images<\/h1>\n
For each segment, we use the free software Free Video to JPG Converter (https:\/\/www.dvdvideosoft.com\/products\/dvd\/Free-Video-to-JPG-Converter.htm<\/a>) to convert the movie to jpg images.<\/p>\n The choice of the number of images depends on the length of the sequence and the movement of the drone. It is therefore necessary to make tests according to what has been filmed. The idea is to have images that follow each other in a fluid way, but still, not to have too many so as not to penalize the calculations.<\/p>\n In short, you have to test.<\/p>\n The extraction creates a large collection of images. Beware, photogrammetry consumes a lot of storage space, it must be planned in advance.<\/p>\n We will use Metashape, Agisoft’s flagship software, one of the best for photogrammetry. It can be loaded here. https:\/\/www.agisoft.com\/<\/p>\n The workflow is quite simple, just let yourself be guided by the menus.<\/p>\n Then compute Align Photos<\/strong><\/p>\n Operate as followed<\/p>\n You get a light CloudPoint. To view the camera path activite Show Camera<\/strong> button.<\/p>\n Go back to workflow and activate Build Dense Cloud<\/strong><\/p>\n To visualize the result Click on Dense Cloud <\/strong>button<\/p>\n Then back to Workflow and choose Build Mesh<\/strong><\/p>\n To visualize the result Click on Model Shaded<\/strong> button<\/p>\n Last step, go back to Workflow and click on Build Texture<\/strong><\/p>\n To visualize the result long Click on Model Shaded button to activate Model Textured<\/strong> button.<\/p>\n In the Workspace, add new chuck, and then restart the process with a new set of photos<\/p>\n To assemble, we need to choose common geometry between the two Chunks.<\/p>\n We have this in the first one<\/p>\n And that in the second chunk<\/p>\n We then add markers on both common geometry on each chunk, with the same name and same order. Right Click, add marker.<\/strong><\/p>\n When done go to workflow, Align Chunk, <\/strong>choose Marker Based<\/strong><\/p>\n It results a Merged Chunk<\/strong> in the WorkSpace.<\/p>\n The two parts are combined<\/p>\n Before, clean up what is not too good by using selections tools and delete to erase unwanted polygones.<\/strong><\/p>\n To have a Merged model go to WorkFlow Merge Chunks<\/strong><\/p>\n Then, remesh and recalculate the texture.<\/p>\n We can do a bit more of cleaning<\/p>\n The easiest way is to select what needs to be kept and then invert selection<\/p>\n Remove all markers (select and right click remove markers) and add new one on the known geometry.<\/p>\n We add 3 markers which will be working as orthonormal reference frame.<\/p>\n Then we change the values of the points to input the ones we want<\/p>\n Update !!<\/p>\n If not it won\u2019t work<\/p>\n The model is ready to be exported. Sketchfab\u00a0<\/a><\/p>\n<\/a><\/p>\n<\/p>\n<\/p>\nMetashape<\/h1>\n
Open Workflow Menu<\/h2>\n
<\/p>\n<\/p>\n<\/p>\n<\/p>\n<\/p>\n<\/p>\n<\/p>\n<\/p>\nCalculating other parts<\/h1>\n
<\/a><\/p>\n<\/p>\n<\/p>\nAssembling parts<\/h1>\n
<\/p>\n<\/p>\nAdding Markers<\/h2>\n
<\/p>\n<\/p>\n<\/p>\n<\/p>\n<\/p>\n<\/p>\n<\/p>\n<\/p>\nScaling the model<\/h1>\n
<\/p>\n<\/p>\n<\/p>\nExploring in VR and AR<\/h1>\n