Highly accurate recognition of the location and orientation of randomly piled parts in three dimensions facilitates the automation and acceleration of production lines for parts feeding.
While robots play an essential role in the manufacturing industry, there are some tasks for which they are ill-equipped. One such task is the selection and pickup of individual parts from randomly piled parts in a box or on a pallet.
This meant that workers would have to position each part at a designated spot for pickup by robot, creating a bottleneck amid efforts to streamline and automate production lines.
Canon’s 3D machine vision system solves this issue. Machine vision refers to the use of industrial image sensors to recognize the position and orientation of parts. The most common form of machine vision currently in use is 2D machine vision, which has difficulty with randomly piled parts. As a solution, Canon developed the RV Series of 3D machine vision system capable of high-speed, high-accuracy three-dimensional recognition of objects. This system enables the automation of parts supply on production lines, a task that conventionally has had to be performed manually, and opens up new possibilities on the front lines of manufacturing. Canon offers three models in its machine vision product lineup—the RV300, RV500, and RV1100—which can pick up parts from small to large to suit a wide range of production lines.
Canon’s 3D machine vision system projects recognition patterns onto randomly piled parts and analyzes the projected images. Based on the analysis of the differences between images of the parts and the multiple projected patterns, the systems are able to recognize targeted objects in three dimensions. Conventional 3D machine vision systems are very difficult to calibrate upon installation. Canon’s 3D machine vision system combines the pattern projector and imaging sensor into a single unit, which makes the system compact and lightweight, and eliminates the need for difficult calibrations. The compact design also allows the system to be moved easily when changes are made in a production line or the line is relocated.
Additionally, the system features a dust- and water-resistant body design for maintenance-free operation, and easy operation through a well-designed software interface.
Canon’s 3D machine vision system delivers unrivaled recognition precision. The system is applicable for parts with curved surfaces, parts with few distinguishing features, and even parts with more complex shapes. Users can easily register parts simply by inputting the CAD data and capturing images of the randomly piled parts. Also, a new approach that matches CAD data with distance measurement data and gray-scale images allows the system to recognize a wide variety of parts with high precision. The system is being adopted by manufacturers across a range of industries, including automotive, electronics, metal equipment, resin and chemical industries. Canon also aims to expand its business in the future through the introduction of systems for assembly processes and the automation of defect inspections.