Actualités Moba

« Nous vous tiendrons informés »

Egg inspector40: identifying different defects

With the help of multi-sensor, full-color cameras



Moba egg grading, packaging and processing machines are packed with technology. Some designs are worth further investigation as they represent crucial customer value. In this series of articles, we will be zooming in on these "hidden gems." One year ago, Moba introduced the Egg inspector40, with the EggQualizer-tool, which tunes the systems to your own egg quality standards for any individual defect type. In this article, we take a closer look at the lighting aspect of the new Egg inspector and how multi-sensor, full-color cameras make it possible to identify different defects.

news-video

In 2006, Moba introduced the Egg inspector, an advanced vision system that detects leaking and dirty eggs using image processing on the infeed of Omnia egg graders. This has been a game-changer in the candling process, as by using a combination of monochrome cameras and narrow bandwidth LED lighting, the most accurate results were achieved. In comparison to detection systems that used single-chip color cameras, the revolutionary technology the Egg inspector uses, makes it possible to check the spectrum of the deviations on the eggshell closely and thereby gives the most accurate results. This has been the basis for the succeeding generations Egg inspector10, 20 and 30. 

 

However, in our never-ending search for innovations, Moba has found a way to take detecting defects to the next level. With the introduction of the fourth generation of the Egg inspector last year, Moba made significant changes to ensure even more accurate results. In the Egg inspector40, the monochrome cameras have been replaced by high-quality multi-sensor, full-color cameras. With these new cameras, even the slightest variation between defects can be detected, thus improving detection rates. This means you can be sure that every egg that ends up in consumer packs meets the quality standards of the target market.

 

Images within seconds

How does image processing exactly work? While the eggs are rotating, images are created by a balanced combination of cameras and flashing high-intensity LEDs. This combination produces special frequencies in the color spectrum making it possible to detect different defects on the eggs. These images are then processed on a high-performance computer using an Artificial Intelligence (AI) model. This AI model is trained by hundreds of thousands of images of all possible defects. This guarantees that the algorithm will recognize the defect, and is trained to identify what specific defect it is. The Egg inspector can even handle both brown and white eggs, even randomly mixed on a grading machine.

 

The output of the AI system is transmitted to the grading system. The Egg inspector "tells" the grader which eggs are dirty or leaking. Depending on how the grader is programmed, the leaking and dirty eggs can be sent to an outlet in the infeed. Now, it is possible to remove leaking and dirty eggs early in the grading process, preventing downstream contamination. This configuration contributes strongly to a high level of food safety and ensures the high quality of the eggs that end up at the consumer's table. Removing leaking eggs as early as possible also ensures that the equipment stays clean, as a clear separation can be created between "incoming eggs" and the "clean zone."

Full-color cameras

Until recently, monochrome cameras have been the industry standard for detecting defects on eggs. Even though color (RGB) cameras have been available for a while. Monochrome cameras are still commonly used due to the large overlap of colors in the lighting spectrum of traditional color cameras. This is shown in the graph on the right (indicated by the black-circled areas). In practice, this means that specific dirt spots on eggs are not visible using these traditional one-sensor RGB cameras.

However, with the multi-sensor, full-color cameras, you can count on:

  • Even better detection performance as even smaller defects can be detected;
  • Not only detecting the dirt but also categorizing the dirt spots thanks to the separate lighting signals.

news-video

Setting your standard

Up to 20% of all unwashed eggs have spots that should not be there. If you ask ten different consumers all around the world what they regard as acceptable in the end product, the chances are that you will get ten different answers. That is because worldwide there are huge differences in what is acceptable regarding the eggs you find in the consumer pack. Hence why setting your detection system to a specific sensitivity level per defect type is of such added value. It increases efficiency and food safety as higher detection rates are achieved and false rejects are reduced. And that is all possible with the EggQualizer tool.

 

This is where the completely new camera and lighting systems come into play. By utilizing the full separation of the light spectrum, many different defect types can be both detected and identified. Knowing exactly which defect is present on the eggshell opens up a valuable addition to the Egg inspector. Namely, setting your own individual requirements for which defect can or cannot be present on the egg, thus improving the net efficiency of the system.

test-1

In short: The multi-sensor, full-color cameras make it possible for the Egg inspector40 to perfectly identify different defects, allowing the customer to define their own egg quality standards! If you want to know more about the Egg inspector40, please contact your local sales representative.

left_image

Contacts et service Notre service fait toute la différence !

img