Poland: Vision system developed to improve sorting
Visort is a vision system which recognizes extrinsic characteristics of sorted products such as size and colour, as well as external damage and deformities (external quality control of the product). Sorter also designed a module to examine internal quality which can recognize flesh or seed damage. Up to now, there hadn't been a vision system on the market for quality control which was able to detect all changes under the fruit-peel during sorting process. It was difficult due to the fact that those structural changes appeared on the fruits or vegetables only after few days (e.g. it may have been caused by Brix sugar content ratio). This often only become visible on arrival to the final consumer, which threatened the return of the entire batch.
Another new solution in the Visort system is ESIQ module (External and Semi – Internal Quality module), detecting fresh bruising during sorting which is not yet visible under normal conditions. Sorter is the first company who have developed innovative technology of vision control to use in sorting machines, which enables recognition of under-fruit-peel bruising, which could otherwise only be identified after one or two days. Sorter created this new method by comparing products at different angles using two matrix cameras, making it possible to penetrate into the product up to few millimeters, as shown in the pictures below. By using cameras and unique image-processing algorithms generating proper code, the vision verification process is more precise and efficient.
This option means that the final client will receive top quality products, without any defects, even those which are likely to occur after few days. This technological solution ranks the company in ‘pole position’ among companies producing vision systems for fruit and vegetables sorting in terms of the system functionality.
(Top row) The colour camera image of analysed object taken up to 1 hour after bruising. (Bottom The image of analysed object with marked algorithm result (external outline means object and internal contour means bruising).
The image of object as a result of innovative algorithm:
The applied algorithms ensure very thorough product analysis. Visort system will meet the demands of special and complicated requests.
During the classification process according to colour, the cameras installed in Visort system analyze a series of images and enable recognition of the product according to:
• the percentage content of colouring (e.g. apple blush)
• the estimation of the average colour value or maximum colour value.
Below is an example of apple classification according to colour:
Using the electronic strain gauges, digital signal filtration and the reduction of noises, fast and modern weighing module enables precise measurement of sorted product in real time, which is several times shorter than it used to be.
Example of product classification according to real-time weight measurement:
VISORT enables the user to analyse the width and height of the sorted products through to the analysis of a series of images including characteristics of each product (e.g. its shape).
Example of apple classification according to the shape:
During the classification process according to external quality, Visort system analyses the product surface and recognizes defects seen in the visible light spectrum or intra-red light (damages such as scabs, hail damages, russeting, mould, rotting). The applied algorithm also recognizes the positioning of the stalk and core, ensuring precise measurement of product. This function ensures that errors are not repeated and allows the users to decide which percentage of defect they are willing to accept. Thanks to this function, it is possible to specify classes of sorted products.
Example of apple classification according to external quality measurement:
The Visort system enables tracing product defects on the whole surface of the scanned product, which is rotated by 360 degrees in order to exclude double identification.
Analysing internal quality through the infrared cameras, Visort scans the products and recognises internal damage which can also eliminate the risk of products containing extraneous objects. In the case of fruit, it is also possible to the analyse the total sugar content. Visort can also check if the fruits contain a high or low Brix value.
Initially, the system was designed to only sort fruits and vegetables, but thanks to its functionality, Visort can be also implemented in other industries (e.g. for sorting glasses, ceramic tiles).
For more information:
Edgar Klusa
Sorter
Tel: +48 377 9999
Fax: +48 377 9998
Email: [email protected]
www.sorter.pl