Raytec Vision

Over the past 3 years, Raytec focused much of its resources on the development of a selection-optical processing technique that has led Raytec to implement the Blueight technology on its SPRAY machines.

The achieved goal of the project was to:

Include in one machine the technologies that other manufacturers typically do on separate platforms.
Simplify at the same time the use of the system that, due to its nature, was not easy to simplify operationally.
Improve continuity in time of selection efficiency, avoiding continuous calibrations.
Overcome the limit of LASER technology without losing its advantages.
Significantly reduce yearly maintenance costs.

The HSL Colour function

HSL is a color analysis function. Processing of the monochrome RGB signals allow to obtain the HSL values (Hue, Saturation, Lightness).

Compared to RGB, HSL is a color representation similar to human perception.

A possible spatial representation of HSL is represented by the solid "double cone" in Figure.

"H" value is 0° with the red hue, 120° with the green hue, 240° with the blue hue and 360° of the red hue again.
"S" is represented by the distance of points from the vertical axis of the solid. On the vertical axis “S” value is 0 while on the circumference of the circular cross-section S has value 100.
“L” is the distance of points from the maximum circular cross section. On the lower vertex of the solid L value is 0, on the HS plane L value is 50, on the upper vertex of the solid L value is 100.

This representation allows identification of portions of the volume of the solid whose coordinates are referred to colors accepted by the sorting recipe.

The CIT function

Visible color analisis is often insufficient for the identification of foreign bodies. Raytec invented the CIT analysis, that stands for: Chlorophyll, Infrared, Transparency. CIT analisis hilights foreign bodies on the basis of the following caracteristics:

low fluorescent light emission;

low infrared light reflection;

low infrared light transmission (Transparency). One threshold combines together the power to:

analyze the presence of chlorophyll to clearly distinguish organic green products (containing chlorophyll, e.g. leafs) from products not containing chlorophyll (e.g. green plastic or insects);

analyses products using the infrared technique to determine the morphological structure of the product/foreign material, in other words to understand if a product is good or rotten, if it is organic (e.g. mais) or inorganic (e.g. wood), if it is green organic (e.g. bean) or animal organic (e.g. mouse);

cancel the background, the conveyor belt on which the product runs, as if it would be suspended in the air (in flight) so that the changes in colour of the conveyor become unrelevant as time goes by. The conveyor may even become of the same colour of the product and the sorting accuracy doesn’t change.

The BLOB function

A "Blob" is a group of pixels next to each other (with contact on at least one vertex) that are recognized as incorrect by the sorting recipe. The system allows to associate a weight to all the various types of defects that the sorting recipe is able to discriminate. The BLOB function rejects only the groups of pixels (blobs) whose sum of weights is higher than a defined value. The logic behind this function is that the defects are not all of the same importance. Insects are certainly undesirable contaminants in fresh-cut salad, so their pixels are associated with an high weight. In this way the "insect blob" is rejected even if it is composed of a few pixels only. Color defects are often due to climatic conditions. For many varieties of fruits and vegetables color defects are particularly evident at the beginning or end of the harvest season. Such defects have to be tolerated in certain parts of the year for not having a too high percentage of rejected products. Pixels of those defects are associated with a low weight.

In this way the "defect of color blob" is rejected only if it is composed of a sufficient number of pixels. Small color defects are accepted by the system.

The SHAPE function

The SHAPE function calculates nine mathematical parameters for each inspected object. Those parameters are representative of geometric characteristics of a two-dimensional object such as eccentricity, compactness, or the presence of protuberance.

If at least one of the nine parameters is characteristic of a defect, then its elimination is possible by using the SHAPE function.

Function BIG STOP

This function allows controlling the presence of dangerous foreign bodies that we can find. The typical samples are the frog or the mouse.

The "Big Stop" allows a series of programmable choice as: object calculation allocated in three different classes of dangerousness.

Automatic imagine capture of dangerous contaminates, alarm generation (with option to stop the machine).

Function STEM