Deep learning driven multifeature extraction for quality evaluation of ultrafast laser drilled microhole arrays

Abstract

An efficient quality evaluation method is crucial for the applications of high-quality microhole arrays drilled with ultrafast lasers. The vision-based feature extraction was used as a data acquisition method to evaluate the drilling quality in terms of the geometric quality of the hole shape. However, the morphological features such as the recast layer, microcracks, and debris on the surface are difficult to consider in the quality evaluation since simultaneous recognition of multiple features remains challenging. Herein, we successfully recognized and extracted multiple features by deep learning, thus achieving the quality evaluation of microhole arrays in terms of both geometrical and surface qualities. Microhole arrays of various sizes and surface quality are fabricated on copper, stainless steel, titanium, and glass using different processing parameters. Then, the images of the microhole arrays are prepared as the dataset to train the deep learning network by labeling the typical features of microholes. The well-trained deep learning network has efficient and powerful recognition ability. Typical features such as the hole profile, recast layer, microcracks, and debris can be recognized and extracted simultaneously; thereby the geometric and surface quality of the microhole are obtained. We also demonstrate the implementation of the method with a fast quality evaluation of an array of 2300 microholes based on a statistical approach. The methods presented here extend the quality evaluation of microhole arrays by considering both geometric and surface qualities and can also be applied to quality monitoring in other ultrafast laser micromachining.