EP0661108B1 - Method for optically sorting bulk material - Google Patents

Abstract

The invention relates to a method for optically sorting bulk material in a colour sorting machine, the material being conveyed by way of a conveyor belt and moving past an observation head having a light source and a product signal receiver arranged in the vicinity of the light source, the reflected light from the image points of the material to be tested being decomposed by means of various coloured filters of adjacent detection elements of a row of the receiver into a plurality of spectral regions and the material to be tested being sorted on the basis of the colour values (measured value of the intensity in the respective colour). According to the invention, to improve the detection rate, it is provided that in the case of material to be tested which is mixed with reject parts, in each case the colour values of the product are investigated in a plurality of selected subregions, in that in each subregion a classifier determines associated areas of image points having colour values falling into the respective subregion and carries out a classification according to predetermined criteria from the geometry and the size of these detection areas. <IMAGE>

Description

The invention relates to a method for optical sorting of bulk goods according to the preamble of claim 1.

Such a method is known from US-A-5,085,325.

It is already known that test material is conveyed on tapes and its image for testing with a diode line camera or TV camera is recorded. The signal is preferably recorded in flight if e.g. the test material from one belt to one other band is translated. When recording signals in flight the test material can be viewed from several sides with a defined background be assessed.

In modern systems, the color is also used for image acquisition detected. The color is used to identify striking areas in the Detect image.

The image of the test material keeps pace with the image scan evaluated so that immediately after passing through a test part this can be classified by the measuring station. So that is ejection of the parts in flight using flaps or air nozzles possible.

A disadvantage of the known methods is that the Detection rate for color heterogeneous products is low, if you focus on the detection of conspicuous pixels Detection of color values limited that are not included in the product because there are many different color values in the product occurrence. If one extends the detection to color values, the are also included in the product is generally already at Extensions to colors rarely found in the product are unbearable high proportion of the faultless product as rejects detected.

The object of the invention is the method for optical sorting of bulk goods to improve that in color heterogeneous bulk material to be detected with a foreign body very low error rate can be recognized.

The method with the characteristic is used to solve this task Features of claim 1.

When capturing an image, the light from each pixel is transmitted Color filter in front of the detection elements of a line e.g. in the three color components red (R), green (G) and blue (B) disassembled. This ensures that detection of conspicuous pixels (Points with color values that rarely occur in a faultless product) by evaluating those measured by the row elements Color values (intensities of the respective color components) possible is. Then an evaluation of the geometry with regard on local clusters of conspicuous pixels.

First, the entire range of possible color values in the Color space divided into several sub-areas, the color space is spanned by the different color components that be measured for each pixel. For example, by the three color components red, green and blue are three-dimensional Color space formed. Classifiers, i.e. Means for evaluation of the measured values based on specified criteria a classification of the measured color values, whereby a Classifier focused on only one sub-area and included in this sub-area for the color heterogeneous product Detects detection areas, i.e. contiguous areas of eye-catching pixels.

If the color values of a color-homogeneous reject part are preferred in the selected sub-area, the committee part as a relatively large area of pixels of the color values of the selected sub-area is detected and can be evaluated this detection area can be recognized by the classifier. On the other hand, the faultless product within a such sub-area in general only in rare cases large areas of conspicuous pixels found, and the number of false detections remains small. This improvement classification is used in practical application by dividing the committee into typical classes and a classifier is set up for each class, where the classifiers work in parallel during the test.

In a preferred embodiment, in the sub-areas, in which committee parts are suspected, by showing learned from committee parts the distribution of their color values.

Fig. 1

zeigt beispielhaft eine eindimensionale Farbwertverteilung mit den Bereichen für gutes Prüfgut.

Fig. 2

zeigt ein eindimensionales Beispiel zur Klassifizierung mit parallelen Klassifikatoren bei der Erkennung von Ausschuß, dessen Farbwerte sich mit den Farbwerten des Produktes überlappen.

Fig. 3

zeigt ein eindimensionales Beispiel zur Korrektur eines Klassifikators durch das Nachlernen.

Fig. 4

zeigt ein Beispiel für die Farbverfälschung an Objektkanten bei Verwendung einer Kamera mit Bildpunkten, bei denen die Farbsensoren nebeneinander liegen.

The invention is explained in more detail below with reference to drawings:

Fig. 1

shows an example of a one-dimensional color value distribution with the areas for good test material.

Fig. 2

shows a one-dimensional example of classification with parallel classifiers in the detection of rejects, whose color values overlap with the color values of the product.

Fig. 3

shows a one-dimensional example for the correction of a classifier by re-learning.

Fig. 4

shows an example of the color distortion at object edges when using a camera with pixels in which the color sensors are next to each other.

The bulk material moves preferentially in a color sorting machine in flight on an observation head with a light source and a product signal receiver arranged in the vicinity of the light source past. The reflected light of every pixel of the The test material is made more adjacent by different color filters Line elements of a camera line, e.g. a CCD line, of the receiver in the three colors red (R), green (G) and blue (B) disassembled. The row elements thus measure in their respective Spectral ranges the brightness of the pixels, also color values called. This results in a three-dimensional distribution of color values, the evaluation of which is based on one-dimensional examples is discussed.

With reference to FIG. 1, the test material is in a pre-learning process without Measure committee parts and the frequency distribution 1 of Color values determined.

In a re-learning process, the test material is also without rejects measured and in a first step a color value range determined for good test material by an experience Threshold 2 is placed over the frequency distribution 1 of the color values , the intersection between the threshold 2 and the curve of the frequency distribution 1 the limits of the test material color value range surrender.

With the selected setting of threshold 2, also at error-free test sample pixels that appear as conspicuous get ranked. These pixels would, however, if they were agglomerate into large areas, mistakenly as a committee classified. Experience now shows that such Agglomeration in turn occurs preferentially in certain color value ranges. In order to measure these color value ranges, in the re-learning process a large area detected in the error-free test material Area saved and the distribution of its color values measured. This distribution is called a standardization Threshold 3 introduced. All color values where the threshold 3 the color value distribution exceeds 1 of the test parts, i.e. the Color values in the interval between the intersections of the Threshold 3 with the curve of the color value distribution 1 are called interpreted according to the test material and thus do not lead to an error detection.

When measuring test material with rejects the color value ranges of the product are divided into sub-ranges. Referring to Fig. 2, everyone is concentrating in this example of the classifiers A, B and C working in parallel only a sub-area. Are the color components of the color homogeneous Committee part preferably in the selected sub-area, the reject part is detected as a relatively large area and can be recognized by evaluating the detection areas become. Again, the distributions of the color values of this large areas measured and after their normalization as Thresholds introduced. All color values at which these thresholds 4, 5 and 6 exceed the color value distribution 1 of the test parts, are interpreted as rejects and lead to error detection.

It is also possible that with a faultless product large-scale detection areas in one by a classifier covered color value range are detected and thus the faultless product is classified as a committee. In one further learning processes are specifically these color values, which too large detection areas in the flawless product area lead, learned and by changing the thresholds as good Test material recognized. 3, the threshold 8 shows the Color value distribution of a reject part. Within the by the Threshold 8 of certain color value range becomes error-free test material classified as a committee. Through the learning process the color value distribution of this large-area detection area measured in error-free test material and according to a standardization as Threshold 7 introduced. All color values with threshold 7 the threshold 8 of the reject part is considered to be Test material is interpreted appropriately and does not lead to one Fault detection.

After learning, the product is automatically checked passed over.

With the test, which can drag on for days, is with systematic drift-like changes in the product. These changes result in a deterioration over time System performance. To avoid this, the classification system doubled. A system takes over the test task while the other system the current color value distribution of the product measures. The measurement of the current color value distribution is carried out by the testing classifier is monitored so that during this measurement no color values of the committee are recorded. After capture The learning classifier becomes a representative number of measured values with the newly measured distribution for the test task activated while the classifier currently set to check takes over the learning task.

This adjustment is only possible if a detected as noticeably classified color point does not always become one Committee decision leads. Would be a detected color point the learner could always lead to a committee decision Classifier do not take on new color values, because with one Committee decision rejected the newly learned color value distribution becomes. Since the system only detected color dots then be classified as a committee if they are a larger one form a contiguous area, the measured frequency can also be adjusted for detected color values. Conversely, can the system with this adjustment color values belonging to the committee detect that in a previous measurement in the color value distribution of the product were represented and at the currently measured Distribution is no longer included.

When the signal is recorded, the test item is, for example, of two Lamps illuminated from the direction of the line scan camera. Between Both lamps have the optical axis of the line scan camera. At this arrangement comes with the design of the background essential importance because of the background the color value distribution if possible, do not expand the faultless product should. An extension would lower the detection performance.

This requirement cannot be met if the test item is recorded lying on the conveyor belt. Because of pollution and wear, the tape has no uniform color. In addition shadows form on the conveyor belt what overall to a substantial expansion of the color value distribution leads when measuring the error-free test material. For this The test material is therefore observed in flight.

In a first embodiment, the background has the Color of the test material, which has the advantage that the contrast between Background and test material is low and therefore the color value distribution of the test material due to edge effects at the transition from the background is not significantly expanded to test material. This variant provides in terms of color and spatial resolution the best results.

The disadvantage of pollution is avoided by the background running as a rotating roller, which deposits flung away immediately. The shadow of the test specimen on the background becomes diffuse and, depending on the bulk density, harmless if the rotating roller at an adjusted distance from the test material is installed. When the bulk density of the test material is high an excessive darkening of the background by an additional Backlighting avoided. Alternatively, you can the background can be a cylindrical spotlight that is in the Color of the test specimen shines and from a transparent rotating Surrounded role, which throws away the deposits.

In a second embodiment, the background is on dark hole, which has the advantage that the test material from Background segmenting and no interference Pollution and shadow formation arise. With segmentation of the test material can, for example, be the form for separating Good sharing and committee are used.

In order to realize the dark hole, one is as large as possible Containers built with low-reflection walls. The line scan camera looks through a slit into this container. The slot is regarding its width at the aperture and focal length of the camera lens and adapted to the distance to the focus plane.

When capturing an image, the light of each pixel is reflected in the three colors red (R), green (G) and blue (B) disassembled. Depending on the selected scanning principle and the adjustment of the camera the color components not ideally in the same place, but measured at different locations. The color sensors are in common color cameras even side by side so that the color sensors with regard to a pixel different local areas of the See the target. 4 are the color sensors (R, G, B) arranged horizontally while the measurement object is from moved up and down this horizontal line. Of the In this example, the background is generated by the respective color sensors the signal levels R = 0, G = 0 and B = 0, while the Object causes the signal levels R = 100, G = 50 and B = 20. In Fig. 4 only measures the sensor triple Xn, Yn the correct color of the measurement object. Color values are used for all other triples measured that contain at least one color value, the darker than the corresponding color value of the test material. So measure to Example the triple Xn, Yn-1 the levels R = 50, G = 25 and B = 10. To avoid these disturbances, pixels, whose color values are darker than that by an adjustable factor of the corresponding neighboring point are suppressed by the Signal levels are stored and a comparison of the horizontal and vertical neighboring points is performed.

Key to the figures Figure 1

occupation

density

error-free test material

fault-free examination material

Reading

measured value

Figure 2

occupation

density

Classifier

classifier

detected

detects

Committee

reject

Reading

measured value

Figure 3

occupation

density

Detection of

detection of reject

Committee

Reading

measured value

Figure 4

Pixel grid

image point grid

background

background

Test object

examination subject

Claims (10)

1. A process for the optical sorting of bulk material mixed with reject parts, such as agricultural products, drugs, ores, etc. in a colour sorting machine by this bulk material being conveyed over a transport belt and being moved past an observation head with a light source and a product signal receiver arranged in the vicinity of the light source, whereby the reflected light of the image points of the examination material is broken down by various colour filters of detection elements lying next to one another in a line of the receiver into several colour components and the examination material is sorted on the basis of the colour values corresponding to the measured intensities of the respective colour components,
   characterised in that in each case the colour values of the product are examined in several selected sub-regions of the colour space by, in each sub-region, a classifier ascertaining connected areas of image points with colour values falling into the respective sub-region and carrying out a classification according to preset criteria from the geometry and size of these detected areas.
2. A process according to Claim 1,
   characterised in that

   the examination material is surveyed in a prelearning process without reject parts and its colour value frequency distribution is determined dependent on the colour;

   in a relearning process, in a first step using examination material without reject parts, a colour value region is defined for good examination material by laying a threshold based on experience over the frequency distribution of the colour values, wherein the limits of the colour value region of the examination material result from the intersection points between the threshold and the curve of the frequency distribution;

   ascertained in the relearning process, using examination material without reject parts, are measured values suspected of containing foreign objects in dependence on colour, which according to the limits of the colour value region of the examination material from the first step of the relearning process would mistakenly be classified as reject, and the size of the local accumulation of these measured values is determined; and

   in the relearning process, using examination material without reject parts, when a preset magnitude of this local accumulation of measured values suspected of containing foreign objects is exceeded, the threshold value decision of the first step of the relearning process is changed dependent on colour so that for these measured values a decision is made for good examination material.
3. A process according to one of the preceding Claims,
   characterised in that during examination of the examination material mixed with reject parts, the classifiers operating in parallel only analyse sub-regions of the colour space in which reject parts are suspected.
4. A process according to Claim 3,
   characterised in that in the sub-regions of the colour space in which reject parts are suspected, the colour value distribution of reject parts is learnt by displaying them.
5. A process according to one of the preceding Claims,
   characterised in that the examination of the examination material mixed with reject parts takes place with a first classification system, while the current frequency distribution of the colour values of the examination material is measured for adaptation to systematic drift-like changes of the examination material with a second classification system, the measurement by the examining first classifier being monitored so that no reject values are detected during the measurement.
6. A process according to Claim 5,
   characterised in that both classification systems alternate in their function.
7. A process according to one of the preceding Claims,
   characterised in that by comparing colour signals of adjacent image points, large gradients and these generally disrupted colour values are not taken into account during the measurement.
8. A process according to one of the preceding Claims,
   characterised in that the sorting of the material takes place in flight, when the material passes from one belt to another, for example.
9. A process according to one of the preceding Claims,
   characterised in that the measurement background is a dark hole.
10. A process according to one of Claims 1 to 8,
    characterised in that the measurement background is designed as a cylindrical radiator with a rotating transparent roller surrounding the radiator, wherein the radiator transmits light in a colour matched to the examination material.

Images