Deep Learning based Optical Image Super-Resolution via Generative Diffusion Models for Layerwise in-situ LPBF Monitoring

TL;DR

This paper introduces a generative diffusion model that enhances low-resolution optical images of laser powder bed fusion builds into detailed high-resolution images, improving defect detection and process monitoring.

Contribution

It presents a novel conditional latent diffusion approach for super-resolution in L-PBF monitoring, integrating 3D morphology reconstruction and zero-shot generalization capabilities.

Findings

High PSNR and SSIM scores indicate accurate image reconstruction.

Effective 3D morphology and surface roughness analysis using the framework.

Model generalizes well to unseen geometries with synthetic data.

Abstract

The stochastic formation of defects during Laser Powder Bed Fusion (L-PBF) negatively impacts its adoption for high-precision use cases. Optical monitoring techniques can be used to identify defects based on layer-wise imaging, but these methods are difficult to scale to high resolutions due to cost and memory constraints. Therefore, we implement generative deep learning models to link low-cost, low-resolution images of the build plate to detailed high-resolution optical images of the build plate, enabling cost-efficient process monitoring. To do so, a conditional latent probabilistic diffusion model is trained to produce realistic high-resolution images of the build plate from low-resolution webcam images, recovering the distribution of small-scale features and surface roughness. We first evaluate the performance of the model by analyzing the reconstruction quality of the generated images using peak-signal-to-noise-ratio (PSNR), structural similarity index measure (SSIM) and wavelet covariance metrics that describe the preservation of high-frequency information. Additionally, we design a framework based upon the Segment Anything foundation model to recreate the 3D morphology of the printed part and analyze the surface roughness of the reconstructed samples. Finally, we explore the zero-shot generalization capabilities of the implemented framework to other part geometries by creating synthetic low-resolution data.