Forensic identification often involves examining an image and comparing a candidate object against the one shown in the image. Matching the pose of the two objects and overlaying them allows for highlighting possible differences, leading to a more effective comparison. The term “superimposition” refers to the task of laying a real object or a 3D model over an image showing another object so that the pose of the two is the same.

The real object and the one shown in the image can be of identical or different natures. The former case is that of Forensic Facial Comparison, where a 3D model of the face of a candidate is superimposed over a facial photograph of interest. As for the latter case, in Craniofacial Superimposition, the skull of a candidate is superimposed over a facial photo of a missing person.

Correct superimposition requires producing a new image of the object matching all the characteristics of the provided photograph. This means not only reproducing position, scale, and orientation but also matching the amount of perspective distortion, which can drastically change the appearance of an
object.

This seminar will introduce the audience to the superimposition process and its application in multiple forensic identification techniques. We will review the photographic formation process and examine the factors that influence the produced image. This will also help clarify common misconceptions about perspective distortion. We will also introduce 3D scanning methods and best practices in handling photos. Finally, we will show how recent advances in Artificial Intelligence allow practitioners to perform superimposition automatically with the help of Artificial Intelligence.