From additive manufacturing and aerial imagery to augmented reality and custom-fit production, the pace of technological innovation is quickening. Companies can no longer afford to skip a generation of technology and expect to remain competitive or capture market share.
Some businesses are adept at using technologies available to them to create and improve products. But many others are having difficulties because they rely on technology that cannot keep up with the fast-changing environment. The difference often lies in how companies approach research & development and software development which we discussed in our previous article.
Areas of R&D expertise
As a software consulting company, Aksiio specializes in AI/ML, Computer Vision, and 3D technologies. With 12+ years of practical experience in R&D and software development, we know first-hand how to solve complex challenges in niche areas. This article highlights the most recent niche R&D project Aksiio undertook and what our team achieved as a result.
3D Scanning & Reconstruction
One of the most promising aspects of Industry 4.0 is the production of personalized goods that enable compelling use cases in orthopedic, prosthetic, fashion, cosmetology, automotive, and many other fields. There is an initial vital task of scanning a physical object and reconstructing it in digital form for most of these applications.
Specialized scanning systems offer great accuracy but come at a high cost and size restrictions, restricting their use for the wider audience. Smartphones and cameras with depth sensors can potentially be used to replicate the results of the specialized scanners. Our team researched and developed a pipeline for reconstructing a full 3D torso model using one RGB-D image and then measuring the parameters of the reconstructed model.
Starting with models of the human body randomly generated using open-source software, our engineers extracted 3D meshes to create a statistical shape model that contains information about the variations in the shape of the human torso. Using that information, the team made an equation to generate new 3D torso models.
This approach enables us to reduce the number of scanned points necessary to reconstruct a 3D model because we can determine weight vectors and match them to a torso that we want to reconstruct. Thus, using the corresponding points on the generated models, we calculate the Euclidean distances between the closest pairs and calculate the human body measurements that we are interested in.
Points on a statistical shape model, corresponding points on the raw image, and the resulting reconstructed torso model with measurement lines.
The solution was tested using an Intel Depth Camera to generate a raw model of a person and the surrounding environment. By manually selecting key points on the raw model and corresponding points on the statistical shape model, the team reconstructed the model using the statistical shape model of a torso.
As a result, the model was used to determine body measurements with a high degree of precision, opening up new possibilities and a wide range of applications. For example, customers could find the perfect clothing size based in the fashion industry. While prosthetic and orthotics fittings could be improved with more precise measurements. The next R&D goals concern the automation of key point selection from images and 3D model reconstruction using just 2D images to expand the scope even further.