Professional arch-viz artist, Indre Kuzminiene, shows how to use 3ds Max and Photoshop to create a 3D object for integrating into a 2D scene...
We are often faced with the task of integrating a 3D object into a photograph. This technique is very often used in architectural visualization for analysis of how a new object is going to blend into the existing environment. It can also be used in video montage when we integrate a 3D object into a real scene that was captured using a static camera (a different technique is used for moving camera though).
For this tutorial we will use 3ds Max
and Google Satellite views to create a 3D object and integrate it into a scene.
Detailed analysis of photographs is required to analyze and implement a camera positioning method. There are 2 methods in the 2014 version of 3ds Max.
1. Camera match (existed in previous versions of 3ds Max): This method is more suitable when it is difficult to identify 3 pairs of perpendicular lines corresponding to the XYZ axes and we can at least partially reconstruct the existing situation.
2. Perspective match (new method introduced in 3ds Max 2014). This is more suited to interiors and freestanding buildings when it is easy to distinguish perpendiculars corresponding to the co-ordinate axes.
In this tutorial we are going to explain how to match the camera in photomontage using method 1.
An example of analyzing an interior scene to determine and co-ordinate the axes
Locate the photo-image
If the task is to place the object into a photo of existing landscape satellite imagery, such as Google Maps or Google Earth (higher resolutions), then it is very easy to locate and place existing objects.
Finding an image to place our object in
Scale the image
Satellite images need to be scaled to match the size of the 3D scene. Refer to existing markers and height points
in the scene.
Scaling the photo-image to match the object's scale
Place the CamPoints
Locate the CamPoints on existing objects that are best visible on the satellite image and photograph. The accuracy of the result will depend on the amount of camera points chosen. It is recommended to choose CamPoints on different height levels, too.
Locating the CamPoints on the 2D image/3D image