The Wizart Vision Reconstruction API makes it possible to obtain information about the 3D dimensions (real sizes) and positions of objects in the scene in the photo.
In cases with indoor photos it allows you to get an idea of the dimensions and distances in the interior, and can also help solve problems with choosing the correct scale for new added objects in AR mode. You can use it to some 3d room reconstruction tasks. Based on predicted values of walls, windows, doors and etc. it is possible to create completely identical 3d models of room from the photo. Which as a result can be used in various AR applications.
see documentation
[
{
"wall_id": 0,
"points": [
[
-1.3342448472976685,
1.499555230140686,
-1.9448838233947754
],
[
-0.04529324546456337,
1.4999998807907104,
-4.439100742340088
],
[
-1.3342448472976685,
1.499555230140686,
-1.9448838233947754
],
[
-0.04529324546456337,
1.4999998807907104,
-4.439100742340088
]
],
"wall_normal": [
0.8883863091468811,
8.334720291713893e-08,
0.45909667015075684
],
"width": 2.807581663131714,
"height": 1.68436075411843
},
{
"wall_id": 1,
"points": [
[
-0.04529324546456337,
1.4999998807907104,
-4.439100742340088
],
[
1.9216951131820679,
1.5,
-3.379462718963623
],
[
-0.04529324546456337,
1.4999998807907104,
-4.439100742340088
],
[
1.9216951131820679,
1.5,
-3.379462718963623
]
],
"wall_normal": [
-0.4742699861526489,
-5.33553787818164e-08,
0.8803794384002686
],
"width": 2.234250545501709,
"height": 1.68436075411843
},
{
"wall_id": 2,
"points": [
[
1.9216951131820679,
1.5,
-3.379462718963623
],
[
2.9297468662261963,
3.7625319957733154,
-4.28175687789917
],
[
1.9216951131820679,
1.5,
-3.379462718963623
],
[
2.9297468662261963,
3.7625319957733154,
-4.28175687789917
]
],
"wall_normal": [
-0.9116857647895813,
0.0,
-0.41088807582855225
],
"width": 1.3528869152069092,
"height": 1.68436075411843
}
]