3D Full-Vision Inspection for Tires

Project Overview:

This project involves the development of 3D intelligent visual inspection and application for tires using 3D laser control technology, linear laser high-frequency control technology, combined with AI visual intelligence algorithms that incorporate template comparison and OCR character recognition. The aim is to provide tire manufacturers with an automated holographic inspection solution.

Tire size detection and tread section size detection, etc

• By capturing the dimensions and tread patterns of tires using 3D laser technology, 100% online inspection can be achieved. At the same time, it can accurately classify manufacturers and tire sizes, and improve inspection efficiency by connecting with tire manufacturer databases.

• Support the detection of out-of-roundness and dimensions of molded tires, including the detection of out-of-roundness, tire width, tire diameter, and other dimensions of the tread.

• Support the detection of tread section dimensions, including key dimensions such as tread thickness profile, overall width, shoulder width, mid-width, shoulder thickness, and mid-thickness.

• Through 3D laser vision to detect the physical and chemical indicators of tires, and deep learning of physical and chemical indicators, the ability to self-detect and grade is achieved.

Inspection of bead joint and tread forming joint

• The detection of bead joints and tread molding joints uses high-sampling-rate sensors combined with self-developed specific algorithms, based on the principle of laser triangulation, to capture the different shapes of laser structure lines caused by the ups and downs of the rubber strip surface during the molding process. The continuous high-speed shooting of the sensor during the rotation of the molding drum can form a set of laser structure lines with a certain rotation step length during this process, forming a cross-section, which can obtain 3D data of the lap joint, and analyze the shape and size of the lap joint.

Tire Appearance Defect Detection: Bulging, Indentation, Scratches, Cracks, etc.

• Through rapid 3D camera scanning and real-time data collection, a 3D point cloud map is generated, and the data is labeled and stored. With the core algorithm developed independently, defects are identified with a recognition rate of 99.96%. The recognition results are then fed back to the MES system of the tire manufacturer in real-time.

3D Detection Technology

Capable of producing highly accurate and repeatable measurement data, it can match the speed of online production to achieve automated non-contact 3D detection. This allows for 100% inspection of tires regardless of material or lighting conditions.

Deep Learning Algorithm

By training on the comprehensive and multi-dimensional boundary features of defective products, it can perform inspection tasks that many traditional algorithms cannot, continuously improving detection accuracy.

Data Analysis Technology

It enables precise quality statistics for each batch of production materials, providing detailed defect records and statistics. This facilitates the assessment of production processes and equipment status, effectively ensuring product quality.