Machine vision-based spangle rating method and device, terminal and medium

• Main Authors: MA HAINENG, WU YINGNA, ZONG DEXIANG
• Format: Patent
• Language: Chinese; English
• Published: 29.11.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention provides a spangle rating method and device based on machine vision, a terminal and a medium. The spangle rating method comprises the following steps: segmenting a plurality of spangle images according to a preset pixel size, and manually marking spangle nucleation points to obtain corresponding unit images and labels; inputting the unit images and the labels thereof into a neural network model for training and verification so as to output a density map of spangle nucleation points of each unit image, thereby obtaining the number of the nucleation points of each unit image; fitting the density map of each unit image by adopting a clustering algorithm to obtain a position coordinate of a nucleation point of each spangle in the unit image, so as to obtain a nearest distance of the nucleation point and calculate an average value of the nearest distance of the nucleation point corresponding to the unit image; and inputting the number of the nucleation points of each unit image and the average value