Abstract
The increasing utilization of micro-nano particles across diverse fields requires precise characterization of their microscopic features, as these features directly influence product performance. Image-based detection offers a promising solution; however, variations in particle size and distribution can impact image focus, leading to distorted feature information. To address this challenge and ensure clear observation and accurate detection, we proposed a method for generating high-resolution particle images by leveraging focus stacking technology in conjunction with focused search method. After determining the image sharpness evaluation function and particle size measurement standard, this method combines multiple images captured at different focal points to reconstruct a single image with enhanced particle details, thereby improving the accuracy of detection features. Experimental validation was conducted using polystyrene particles of three mixed suspensions, the average errors in particle size for the reconstructed images were 2.23%, 1.39% and 1.89%, respectively, which are lower than those of the particle images before reconstruction, validating the effectiveness of this method. The reconstructed images showed clearer two-dimensional outlines of particles, with minimal errors in particle size and shape. Additionally, the method can differentiate particles of different sizes based on their texture features, resulting in a bimodal distribution that provides more accurate representation of particle size data.