Computer Vision

Part
01
of one
Part
01

Computer Vision

While there is no pre-existing information to fully answer your question, we've used the available data to pull together key findings. We have located a company that uses machine vision for 360 scans. It can detect if the correct text is put on labels on cartons and can scan glass jars for defects and cracks. We also located a robotic circular scanning system for 3D phenotyping of plants which is similar to scanning for text on the manufacturing floor. In addition, we also located a machine vision tool attached to a circular knitting machine to detect defects in the material. The processes that we have identified use multiple cameras taking images of the product and then doing text verification, so while it is a 360 machine vision scanning, it does not appear that anyone is doing circular text reading, but image scanning and verification. Additional companies or papers with this process were not publicly available.

Below you'll find an outline of our research methodology to better understand why the information you've requested is publicly unavailable, as well as a deep dive into our findings.

METHODOLOGY

An extensive search of the public domain revealed no news articles or companies using circular scanning via computer vision or artificial intelligence.

Since searching news articles directly was fruitless, we side-stepped the issue by instead searching for reports and scholarly articles on computer vision, artificial intelligence, and circular scanning.

HELPFUL FINDINGS

Integro Technologies has a machine vision inspection system that can do a 360-degree scan. It currently can scan ice cream containers to make sure the type of ice cream is labeled on the carton. It can also inspect glass jars for defects and cracks. The system stores data and the comes with a login system for security. Integro uses a 2 axis robot system and scans 360-degrees in 3D form. The system includes multiple cameras that capture the image and verify the labels. They are located in Salisbury, North Carolina. The company started in 2001 with 1 employee and now have 48, including engineers in machine vision integration. It has several products for various applications that do machine vision inspections. Information on specifics of when the applications were developed is publicly unavailable. With its current products, Integro continues to be labeled one of the fastest growing companies and the 2017 Integrator of the Year.

CircumSpect- Machine Vision Company offers this product that does a 360 package and label inspection. It does not require contact with the product and can verify 750 products per minute if they are round or cylindrical. Multiple image scans are put together to do one image verification to make sure the product passes the inspection criteria before being passed to the next conveyor. It is now owned by VAIA and is located in Massachusetts. We were unable to locate exact specifics on when the 360 machine vision inspection was started, but we did locate a technical overview of the 360 vision and how it works. The computer is loaded with an image of what the label is supposed to look like, it then scans the product label to verify text matches as well verification that the label is not crooked or misplaced on the package. The module has 4 cameras mounted on the side that produce side images and a 5th that looks down over the top to make sure lid is applied. MVC has in-house machine vision engineers for both software and mechanical systems. It also has a support team and a full-service machine shop.

Robotic circular scanning- This is being done with plants, but it includes a robot scanning in a circular trajectory and taking images. The scan uploads and then can be analyzed to make sure the proper data is captured. The system determines the center point and then scans. It then waits 10 seconds before scanning again. This is a paper on 3D phenotyping, so it does not have a specific location in the paper.

Circular knitting machine vision inspection- The tool is attached directly to the circular knitting machine and constantly scans for defects in the material online. It is used for quality control and fabric grading. The system includes both hardware and software and it calculates a 93% detection percentage. This is being used in Italy and was presented at the 5th WSEAS International Conference on Applied Computer Science. This was developed by the Department of Mechanical Engineering at the University of Florence in Italy. Additional details were not publicly available.

CONCLUSION

Despite the limited capabilities of internet archiving on the subject of computer vision being used for circular scanning of text on manufacturing floors, we were able to find information on a machine vision system for 3D plant phenotyping that has a similar process. In addition, a machine vision defect detection tool attached to a circular knitting machine to identify defects in materials, a product called CircumSpect that is also a 360-degree machine vision scanning tool that can detect errors in text on product labels and pre-printed packages. Additional information on companies or papers using the circular reading of text was publicly unavailable.
Sources
Sources