3D printing model defect pre-judgment system and method based on visual neural learning

Abstract

The invention discloses a 3D printing model defect pre-judgment system based on visual neural learning. The 3D printing model defect pre-judgment system comprises a scanning unit, an information transmission unit and an information processing module. The scanning unit comprises a camera and a grating projection device. The information transmission unit comprises a WiFi module and a controller. Thegrating projection device projects the grating on an object to be measured, carries out thickness change and displacement; a digital image captured by the camera is transmitted to the controller in cooperation with the camera; the controller transmits the digital processing information to the information processing module through the WiFi module, and the information processing module transmits the obtained digital processing information to the system online database and inputs frame-by-frame image data into the convolutional neural network model base trained by the system for matching and prediction. The method has the functions of photo type scanning, three-dimensional recognition, automatic defect pre-judgment, real-time 3D printing monitoring and algorithm model automatic optimization,the tedious step of manual re-printing of a user is omitted, and the printing efficiency and the yield are improved.

Description

technical field [0001] The present invention relates to 3D printing technology, more specifically, to a 3D printing model flaw prediction system and method based on visual neural learning. Background technique [0002] In recent years, 3D printing has been widely used, and the market demand is increasing, especially for some manufacturing enterprises, they need to use 3D printing to manufacture some parts that are difficult to manufacture by traditional techniques, but for enterprises, these Parts are market-oriented and must be mass-produced. At present, 3D printing has a relatively low yield rate and low yield in some finishing fields (precision instrument fields), and it is difficult to achieve large-scale production. There are long production times for 3D printing models and 3D printing conditions are controlled. Inaccurate, poor printing materials and processing accuracy. Contents of the invention [0003] In order to overcome at least one defect of the above-mentio...

AI Technical Summary

Problems solved by technology

[0002] In recent years, 3D printing has been widely used, and the market demand is increasing, especially for some manufacturing enterprises, they need to use 3D printing to manufacture some parts that are difficult to manufacture by traditional techniques, but for enterprises, these Parts are market-oriented and must be mass-produced

At present, 3D printing has a relatively low yield rate and low yield in some finishing fields (precision instrument fields), and it is difficult to achieve large-scale production. There are long production times for 3D printing models and 3D printing conditions are controlled. Inaccurate, poor printing materials and processing accuracy

Images