3D similarity analysis

Here is an introduction to the 3D similarity judgment technology that we have pioneered in the world.

What is 3D similarity analysis?

3D similarity analysis means to search for a shape similar to the input 3D shape in the 3D shape database. We are the pioneer in the world in developing 3D AI models that can make 3D similarity analysis.

Conventional similarity analysis technology

In many design fields such as manufacturing and construction, work is shifting to be based on 3D geometry using 3D CAD. Computer-aided technologies such as CAD, CAM, and CAE are also based on 3D shapes. Therefore, it is necessary for AI to handle 3D data as well.

The conventional method of determining similarity is to use text-based AI to search for similar shapes based on character information such as part numbers. However, this method has its own problems, such as the huge amount of time required to prepare the text information for the search.

Conventional similarity judgments use a group of snapshots of 2D images as an alternative, even when identifying 3D shapes, which makes it difficult to maintain recognition accuracy, among other problems.

For details on the problems of conventional AI, please refer to the activity outline.

If the 3D shape can be recognized directly, as in our method, it can reduce the time and effort of preparation and avoid the problem of shooting.

Mechanism of 3D similarity analysis

The core mechanism of the 3D similarity analysis is the feature extractor.

It extracts a lower-dimensional feature vector from a more complicate raw input data such that the extracted vector encodes useful information of the input data which is suitable for further application.

Take the bolts as an example, instead of storing a bolt using its CAD data, which may require a large amount of memory and have no quantitative characteristics, we can choose to extract and store only useful features of the bolt such as its type, diameter, and length.

Intuitively, the two similar bolts should have two similar features. By utilizing such feature vectors, we can search for the bolts with similar geometry in the database.

A simple example of extracting bolts’ features by their length and diameter.
Each dot represents the features of a bolt. Intuitively, when doing similarity analysis for the bolt represented by the red dot, the two bolts shown by green dots are found as the most similar bolts.

The methods of feature extraction varies from plain mathematical processes like the Principal Component Analysis, t-Distributed Stochastic Neighbor Embedding to deep neural networks like autoencoder.

Introducing products for 3D similarity analysis

We have prepared a demo site of 3D matching AI PLM equipped with a 3D similarity analysis function.

At the 3D matching AI PLM demo site, you can try out matching examples of bolts and brackets.

In 3D matching AI PLM of bolts, classification is performed by bolt head shape and similarity analysis is performed considering the size of bolt length and diameter.

650

In addition, in the 3D matching AI PLM of the bracket, the similarity is judged by the overall shape of the bracket.

On the demo site, you can try out an example of bolt and bracket similarity analysis. For bolts, determine the similarity by considering the head shape, bolt length, and bolt diameter size, and for brackets, considering leg length, width, plate thickness and holes, slits, and so on.