Multi-focus image fusion for microscopic depth-of-field extension of waterjet-assisted laser processing

The waterjet-assisted laser processing (WJALP) technology has a significant effect in reducing the damage to the workpiece caused by laser ablation. While, due to the limitation of the depth-of-field (DOF) of the microscopic magnification system, the processing quality and features of the material surface cannot be accurately identified during processing. In multi-focus image fusion (MFIF) according to different DOF images, the conventional wavelet transform in the condition of waterjet cannot accurately provide detailed features of microscopic images. In this work, in order to overcome the shortcomings of traditional wavelet transform, the method of discrete wavelet transform (DWT) based on the human vision principle (HVP) was proposed, and stretched the detail part while fully retained the source image information. Based on morphological theory and defogging algorithm, the waterjet interference was excluded by image pre-processing as much as possible. The approximate component and detailed component were obtained by DWT, and the detailed components were stretched utilizing HVP according to the brightness and darkness provided by the approximate component. Finally, the image brightness was adjusted using an adaptive gamma correction. According to the experimental results, in the case of complex flowing water films, SD , E ( F ), AG , and SF can reach 75.45%, 95.38%, 73.60%, and 77.45% of the ideal fusion, respectively, which indicated that the proposed method can achieve a better fusion effect in different waterjet situations.