Anyone in the supply chain, mail order, order fulfilment and logistics business would know that warehouses play a key part in the customer’s experience and in ensuring the company meets its productivity goals.
But all over the world, warehouses face huge challenges in processing the millions of incoming packages received each day. First, packages need to be identified accurately and registered properly, then they have to stored efficiently for onward delivery, to prevent them from becoming lost inside the warehouse.
Quelle GmbH, based in Germany, is one mail order company that has to deal with 80,000 incoming packages a day. Upon arrival, each package must be identified, registered and marked by a specific label generated by Quelle.
Due to the large number of incoming packages, such a task cannot be performed manually. Compounding the difficulty, some packages do not have any identification labels, or exhibit incorrect font or are labelled by hand.
Often times, the labels are squashed, crumpled or difficult to read, resulting in enormous difficulties in bar code and OCR reading, making package identification by machine vision almost impossible.
To turn the situation around, Quelle turned to – Machine Vision with advanced OCR.
Machine Vision can acquire images for processing, analysing and measuring of various characteristics for decision making, while Halcon Deep OCR, released in November 2020,
can localise characters much more robustly regardless of their orientation, font type and polarity. Big images can be handled more readily, and its list of character candidates contains corresponding confidence values which can be used to achieve high quality recognition results.
MACHINE VISION WITH ADVANCED OCR IN BAR CODE READING
To implement the Machine Vision system at Quelle’s warehouse, MVTec Halcon OCR was deployed together with a network of intelligent i-Cam 108 cameras at the entrance of the warehouse.
An architecture was further designed to guarantee the calculation power of label detection and high resolution OCR reading.
During identification operations, if the bar code reading of the vendor label is identical to the OCR reading, the package is classified as identified.
In cases where the vendor label was damaged during transport or non-specified fonts were used on the vendor label, MVTec Halcon stepped in with an added powerful co-processing software, run on a regular desktop computer.
The results from the OCR software run on the desktop computer are then compared with the OCR results of the i-Cam. Where at least two results of the three OCR measurements match, the package is classified as identified.
“Using the i-Cam OCR with the Halcon OCR software, we achieved a significant increase in the degree of automation and we are pleased to report this is the desired level we specified,” said Kerstin Mehlhose, plant engineer at Quelle.
And adding a feather to the cap of this solution:
- No structural alterations are required to existing warehouse operations
- Only simple installation is needed for the i-Cam cameras
- A regular desktop computer can be used for the Halcon OCR software
- Additional software may be added in parallel
- No maintenance is required of the system
IS MACHINE VISION WITH HALCON OCR POSSIBLE IN OTHER PRODUCTION LINES?
Today, machine vision systems are widely used in many industries. Along with this, artificial intelligence (AI) is increasingly deployed in machine vision, enabling judgement based part location, inspection, classification and character recognition to be performed many times more effectively than human effort or traditional machine vision solutions.
For the package and logistics industry, advanced OCR and deep learning are having a huge impact on material location identification, pick and place and checking of information on the package.
Further afield, deep learning is benefitting factory automation and production lines that use machine vision, especially when parameters cannot be clearly defined. As a result, product identification, sorting, measurement, visual quality checks and inspection can be executed at extremely fast speed, with high accuracy.
To top it all, the Halcon Deep OCR released in November 2020, includes a holistic deep-learning-based approach to OCR, and has the capability to address a much wider range of applications, thanks to its additional character support.
Furthermore, MVTec works with most vision cameras, and can extend the power of traditional machine vision in the Food and Beverage industry, as well as in automotive, pharmaceutical, and semiconductor production.
Wonder how this is done?
Come, discover how classification and character recognition challenges can be effectively solved with the power of Halcon Deep Learning at the LINXDays Southeast Asia 2021 Webinar.
All Webinar participants are entitled to an exclusive ‘Halcon Orientation Training’ worth $800, which comes with a 1-month trial of Halcon and immediate hands-on training.