SI20441A - Method for monitoring a two-dimensional or three-dimensional distribution process - Google Patents

Abstract

The invention relates to a method for monitoring the distribution of structures on a surface or of particles in space. An optical image of the distribution is produced using video technology, for example. This image is then broken down into pixels, the average brightness of each pixel is determined, and the differences between the brightness values of adjacent pixels are calculated along pre-selected rows. These differential values are recorded on a data carrier and/or output in such a way that a correlation of the differential values with the position of the pixels is preserved on the image. This is used to determine at which points there are inhomogeneities in the distribution of the structures on the surface or the particles in space. The method can be used for example for assessing the homogeneity of surfaces, for detecting surface flaws and for monitoring the opening angle and the homogeneity of a directed spray.

Description

A place in a row

PROCEDURE FOR SURFACE OR SPATIAL DISTRIBUTION CONTROL

The invention describes a process by which the distribution of structures on a surface or particles within a space can be automatically controlled. In the event of deviations from the desired state, alert messages may be automatically introduced locally or at a remote location and / or corrective measures. For example, the process is suitable for the automatic control of the results of cleaning, chemical transformation and / or coating of surfaces such as a metal or plastic surface. This does not require human intervention. The area of application is, for example, the control of surface transformation and / or surface transformation. surface coatings in the steel and automotive industries. Furthermore, the process is suitable for controlling the spatial distribution of particles in a spray jet. The discharge angle of the spray jet and the homogeneity of particle distribution in the spray jet are automatically controlled on the one hand. The scope of the process, on the one hand, is spray drying or spray drying. spray curing of solutions, suspensions or melt with a view to optimizing the time-consuming process of these processes. However, if the spray jet serves to coat the surface as evenly as possible, then a process may be used to keep the coating application in the desired area.

In the present state of the art, it is known how images of a surface are stored in digital form on a storage medium and how image analysis is performed automatically from different points of view. For example, patent US-A4 878 114 describes a computer-aided optical system for assessing the roughness of flat surfaces of a product. This system comprises an adjustable light source for surface illumination, a video camera and a device for storing the video signal output in digitized form, and a processor for analyzing these digitized signals so as to identify a parameter that illustrates surface roughness. As a result of the image analysis, a single number is obtained that characterizes the roughness of the surface cutout analyzed. This analysis result does not provide any explanation as to whether there are particularly large deviations from the average roughness on the surface being investigated and at which point.

EP-B-255 177 discloses the process of automatically recognizing an object, which contrasts with a digitally stored video image. For this purpose, threshold values above the background values (Hintergrundwerte) are defined above which the image element is considered to belong to the contrasting object. Therefore, the process is so balanced that it recognizes the pattern in the video image. However, it is not appropriate to analyze the occurrence of the spatial distribution of disturbances on an otherwise largely homogeneous video image, if these disturbances do not describe any common object.

EP-B-428 751 patent is a measurement method for assessing paper quality. It estimates fluctuations in the light transmittance of paper. It is a texture measurement process that captures an image from a transmitted light of light source on a paper surface by means of a camera in order to mark the reflection of transmitted light on a marking device of a computational element that transforms the image of transmitted light on a marking device into windows of a certain size and of a certain number, denoted by calculating an average value of the intensity and some primary variance of the intensity of each window from the intensity of each individual pixel in each window, calculating the average value of the primary variance for all windows and the secondary variance of that first variance for all windows, and using the secondary variance for all windows as texture factor. As a result, this procedure gives a single characteristic that describes the uniformity of the transmitted light and thus the homogeneity of the paper structure. This procedure does not recognize any of the locations of the fault and its position on the paper surface.

EP-B-159 880 relates to a device for evaluating the density and uniformity of a sample of a printed sample, such as a letter. The apparatus is indicated by a means for calculating the density distribution of a plurality of image data in each of a plurality of sample segments by typing the density of each image data particle; means for testing the density distribution within each segment and for those segments in which the density distribution satisfies a predetermined function of normalizing the density distribution in each such segment by dividing each density distribution value by a value representing the total density in the named segment; and a means for quantifying the density and uniformity of the sample based on the density distribution, which are normalized by a determination and normalization means, in which uniformity is defined as the rate of change of the normalized density within the sample. This procedure is not intended to identify the occurrence and location of a local disturbance in the sample.

There are a variety of technical procedures in which the quality of the result is characterized by the surface having as homogeneous appearance or as possible after this technical process. that the space cutout is as evenly filled as possible with the particles. An example of this is cleaning and / or hydrophilizing metal or plastic surfaces. The result of this machining step is to ensure that the surfaces are homogeneously coated with water film. The formation of droplets, which appears in some image of the surface as a disturbance of uniformity, indicates the lack of results of the Purification or. hydrophilization.

A further group of technical processes, such as phosphating, aims at the chemical transformation of a metal surface in order to protect it, for example, from corrosion. Disturbances of these processes become noticeable as interfering sites in an otherwise uniform surface coating. The surface image shows these opacities by deviating their brightness from the average brightness of the surface. The same applies to coatings with anti-corrosion coatings such as varnishes. Disturbances in the even varnish layer, such as bubbles or craters, are also noticeable as places whose brightness deviates significantly from the average brightness of the surface.

Other technical processes in which liquid or solid particles are dispersed over several nozzles in a room are to determine the exit angle of the spray jet at the nozzle as well as to distribute the particles in the spray jet uniformly. For example, a clogging of a single nozzle in the nozzle head is shown so that the particle space is less densely filled than in front of a properly functioning nozzle in front of each clogged nozzle. If spray is used to treat the product by spray drying or spray curing, clogged nozzles lead to reduced space-time efficiency. However, if the spray is used to apply the particles as evenly as possible, the clogged nozzles lead to an uneven coating of the surface. When applying a coating to a surface by spraying liquid or solid particles, it may be important for the spraying jet to have a defined exit angle for the economy and efficiency of the spray process.

Surface uniformity in surface treatment processes or homogeneity of particle distribution in a spray jet or. its exit angle is generally estimated by visual assessment of the surface or surface. spray jet directly or. from a photo shoot. Therefore, human intervention is needed for the first time. Secondly, visual assessment may only be feasible at certain times when the workforce is not bound by other obligations, unless staff are set solely to continuously monitor the production result. If, as a rule, normal controls are performed only at certain moments, there is a risk that a defective product will emerge between the two control moments.

The invention sets itself the task of working or. the manufacturing process shall be continuously monitored automatically, showing the quality of the results, either as a surface with a homogeneous appearance or as a uniform filling of the particle space, and to send an alert message and / or the causes of the disturbances in the event of disturbances and, if possible, to eliminate them.

The invention therefore relates to a method for controlling the distribution of structures on a surface or particles in a space, characterized in that

a) at least one two-dimensional image of the distribution shall be produced optically or electronically, divided into pixels and the luminance of each pixel stored in digital form on a data medium,

b) divides the image or cutout of that image into a specified number into sorted image elements, each image element comprising at least four pixels,

c) determine the average luminance of each pixel by determining the average luminance of each pixel of that pixel,

d) determine the difference between the mean luminosities of the adjacent image elements along the first set of image elements and write n ^ the data carrier legibly and / or be printed in the form of a diagram in order to obtain a local correlation between the differences and the position of the corresponding image elements , and if desired

e) Step d) Repeat for a pre-selected number of the following types of image elements that essentially run in parallel with the first given row.

In this case, the formulation in partial step a) of at least one two-dimensional image means that an image or more images of a distribution are taken. As a rule, one image is sufficient to evaluate one mainly flat surface. However, in order to estimate the spatial distribution or strongly curved surface, it is preferable to take several images whose image planes form predefined angles. With this, the surface or the spatial distribution is estimated from different angles. It is preferably used to form a video camera image. In this case, the image clip is adjustable by the focal length of the video camera and / or by the distance of the camera from the subject being evaluated. If too small cutouts of the object are evaluated, such as for controlling the conversion of metal surfaces, it may be necessary to equip the video camera with a microscopic device. Of course, this procedure assumes that the surfaces or surfaces are painted. the recorded spatial areas are sufficiently illuminated. For example, a device described in US-A-4 878 114 may be used to record and store image information in digital form.

In partial step b), the image or the selected image clip is decomposed into exactly the previously selected number into sorted image elements. In this case, the maximum number of image elements is conditional on the decomposition of the camera used. Further, the pixel is marked with the lowest possible decomposition-dependent pixel that can be attributed to luminosity (Grauwert). Therefore, an image can be decomposed into up to as many pixels as the pixels allocated to it. In the process of the invention, however, image elements comprising multiple pixels are selected by the preferred method. For each pixel, the luminance is determined by adding the luminosities of the individual pixels and dividing the luminosity by the number of pixels. With such computational coverage of multiple pixels in a single pixel, background noise is reduced by the formation of a mean. Preferably, an image element is selected at least so large that it comprises at least 4 pixels, preferably arranged in pairs in pairs. Depending on the problem being posed, the image element may also have significantly more pixels.

In order to maintain a meaningful message in the next partial step d), each type of image element must contain at least 2 elements. Preferably, however, the number of pixels in a row is substantially larger, such as from about 10 to about 200 pixels, especially from about 15 to about 100 pixels in a row. In order to perform the process according to the invention, it is sufficient to work in principle with a single type of image elements. However, a more reliable message is obtained if the image or image cut is broken down into several types, which can include a larger cut from the surface or surface. from a spatial scale. Preferably, an image or a cutout is decomposed into as many types of image elements as there are images of each type. This means that the image or the exhibit the cutout in some 10 to some 200, preferably in 15 to 100 types.

As large as possible the image elements are selected, i.e. the more pixels they span, the better the background noise will be detected. In any case, the choice of image elements larger than the size of the expected error locations should be avoided. If the image element is larger than the error location, there is a risk that the error location will not be recognized. With the number and size of image elements, the correlation relation is superficial or surface. a space cutout that we take as a picture. As a rule, surface or surface imprints are preferred. a space cutout having a side length between about 1 mm and about 5 m. By adjusting the distance of the camera from the subject, the focal length of the lens, and possibly using a microscopic device, the individual image cut-off can be adjusted. If a procedure is used, for example, to evaluate the results of a chemical surface treatment, it is advisable to capture and paint a surface cutout of approximately 10 x 10 cm. Particularly advantageous are 30 x 30 cm imaging cutouts for evaluating the surface coat or the homogeneity of the spray jet.

In partial step c), the average luminance of each pixel is determined by subtracting the luminosities of each pixel of the pixel and dividing the resulting sum by the number of pixels. This reduces background noise. If this is not sufficient, a computational filtering process such as a Gaussian filter or a Fourier filter can be used before or after the formation of the mean. Gaussian filtering is also carried out, for example, in the previously cited EPB-255 177. In the process of the invention, such filtering processes are only necessary in exceptional cases.

In partial step d), the right criterion for assessing the homogeneity or the to determine the unevenness. Further, from the results of partial step d), the outlet angle of the spray jet can be determined, as will be explained in more detail below. It is important for the invention that the differences between the luminosity of adjacent image elements are indicated or marked. they print it like this. which location on the mapped spatial cutout fits in to be recognized. in the distinctively large difference between the luminosities of adjacent image elements. Disturbances in homogeneity are reflected in the particularly high difference between the luminance of adjacent image elements. In the case of point disturbances such as bubbles or craters in the lacquer layer or some freckles (bright crater defects) in the crystalline layer after phosphating, the occurrence, number and spatial distribution of these disturbances can be recognized. Line or surface disturbances, such as uncovered or otherwise covered sites on the surface, are distinguished by showing the edge of the disturbed surface in each type of image element as a particularly large difference between the luminance of adjacent image elements. The turbidity line is recognized in such a way that the locations of particularly large differences in individual types of image elements form continuous lines. A surface disturbance that is smaller than the selected cutout is recognized by identifying in each row of image elements, as a rule, 2 a particularly large difference between the luminance of adjacent image elements, each indicating the beginning and end of the disturbed surface. In a graphical two-dimensional representation of the luminosity differences of adjacent image elements, the particularly prominent differences in disturbed surfaces are described.

It is, of course, therefore necessary to determine beforehand what differences in luminance of adjacent image elements can be accepted as noise, and from which threshold values the differences are further characterized as disturbances.

In the process according to the invention, the basic luminance of an image, which depends, for example, on the intensity of the illumination, does not play for up to 4 recordings, so-called no role can be played, since in determining the luminance differences of adjacent image elements, the absolute brightness disappears. However, it is preferable in each case to ensure that the painted surface or the illuminated surface is as uniform as possible. mapped space cutout. For this purpose, one or more light sources may be used whose light cones are preferably oriented parallel to the direction of the shot. Thus, shading will be minimal. Preferably, light sources are grouped around for the cameras used. The overlapping of individual light cones leads to a particularly uniform illumination surface. space cut. If, however, such a preferential arrangement of light sources is not possible, the uneven illumination of the recorded surface or the spatial cut-out shall be computationally corrected by the illumination correction. The assumption is that, before determining the differences between the mean luminosities of adjacent pixels, the previously selected correction value is subtracted from the luminance of each individual pixel before determining the luminance value of the pixels or from the luminous intensities of the individual pixels or before the selected correction value is added, whereby individual pixels or pixels of the corresponding correction value are described by a plot over an image or a cutout. In this case, corrective values or. the surfaces to which these values relate, to be determined or from substantially all pixels or points. image elements, or from any statistically relevant selection of flights. These corrective values or. the surfaces to which they refer can be determined, for example, by the method of minimum square deviation of the individual luminosities observed from the luminosity of the leveling surface, which thus represents the surface of the correction value. In this way, the effect can be computationally compensated by defining it in oblique illumination

Part 1 of the picture looks systematically darker than the other.

In order not only to draw a line cutout of the captured image to identify the disturbance, but to attract a larger flat cutout for this purpose, in step e), a partial step d) is repeated with a certain number of further types of image elements that are predefined above. mainly in parallel with the first given row. The parallel course of the species becomes preferred if the analyzed image is rectangular and especially square. As a rule, this is preferred. However, if the camera's viewing angle is not at least substantially perpendicular to the recorded surface, on the axis of the recorded spray jet, the image taken contains a perspective distortion. In this case, it would be preferable to analyze the image cutouts and also individual image elements, not rectangular but trapezoidal so that each image element corresponds to the same surface of the object being recorded or. space cut. In such cases, the types of pictorial elements do not necessarily run in parallel, but may form a (small) angle to each other, illustrating the perspective perspective distortion of the reflected surface cutout or. spray jet.

In this analysis of the image, along the generally parallel lines of the image elements, no linear or surface disturbance can be recognized, the edge of which is, respectively. the edges randomly run approximately parallel to the selected image element types. In order to avoid this danger, it is proposed to analyze the cutout at least along one or more types of image elements (ie, to determine the differences in luminance of adjacent image elements), which, with the first pre-selected type of image elements, is determined as, preferably in the range of about 60 ° to about 120 ° and especially in the range of about 90 °. This recognizes the particularly large differences in luminance between adjacent image elements along different directions on the surface, making the estimation process more reliable.

The method of the invention is used, for example, to control the distribution of structures on any metal or plastic surface. Preferably, it is selected as far as possible from a straight cutout of the surface, which is taken from as near a vertical angle as possible. As mentioned above, the estimation of a curved surface or oblique angle can be adjusted to take a perspective perspective when selecting individual image elements. In this case, therefore, the size of the individual image elements is preferably selected such that each image element represents essentially the same large area of the recorded object. Furthermore, a previously described illumination correction may be advantageously used on a curved surface. In any case, it is advisable to adjust the angle of view and / or the illumination

so that cutouts. se we avoid it shading of the evaluated pictorial For example se can use a procedure to control what kind metal or plastic surface. at which controls

success of cleaning and / or hydrofishing. Successful cleaning and / or hydrophiling is recognized by the fact that the surface is covered with a uniform liquid film after leaving the treated zone. Insufficiently cleaned and / or hydrophilized sites, on the contrary, become recognizable by breaking the water film and pouring it into a droplet. a puddle. In the captured image of the surface, they can be recognized again in suitable illumination as being from a uniform background of differing luminosity. According to the invention, the difference between the luminosity of adjacent image elements can be identified.

The controlled surface may further be a metal or plastic surface that is subjected to a chemical treatment (such as cleaning or conversion treatment that alters the chemical nature of the surface) or coating. For example, it may be a metal surface that is subjected to chemical treatment in the form of chromatization, treatment with an acidic solution of simple and / or compound fluorides, treatment with a solution of metal transition compounds, or layered or non-acidic phosphating. Such machining steps are well known in the field of corrosion protection - preferably against subsequent varnishing - of technically important metal surfaces such as the surface of iron, steel, galvanized or zinc alloy steel or aluminum and its alloys. Such processes are routinely used in the metalworking industry, such as the automotive industry or the manufacture of household machinery, to improve corrosion protection. A particular field of application of the process according to the invention is to evaluate the quality of some layered phosphating in the automotive industry. In this way, deficiencies in quality in the form of disturbances in the uniformity of the phosphate layer become noticeable. These disorders can be mainly point-like, such as Phosphatierstippen. Disturbances may also be superficial and may occur, for example, in non-phosphated or less densely phosphated metal sites. For such a particular scope, a surface cutout having a size of approximately 10 x 10 cm is preferably selected. This image is divided into so many image elements that some 0.5 to 2 mm large phosphating spots are still recognizable as individual disturbances.

Furthermore, a process may be used to control those metallic or plastic surfaces that are subject to a network of organic matter. A typical example of this is lacquering, which is carried out, for example, in the form of electric immersion lacquer, as well as by dipping into a lacquer bath or by lacquering. Typically, such a lacquer hardens after application by heating, infrared irradiation, or photochemical, with the corresponding reactive molecules interconnecting and networking. Such varnishes may have different interferences. This may be more of a point disturbance than, for example, bubbles or craters in the lacquer layer, or also applied dust particles or dirt, which becomes noticeable as a bulge in the lacquer layer. Because such interferences in the appropriate illumination reflect light differently than a undisturbed layer of lacquer and thus become visible in the image of the lacquer layer as sites of different illumination, they can be recognized as outstanding differences in the luminance of adjacent image elements in the process of the invention. However, they may also be surface disturbances that become noticeable by the systematic distribution of the striking differences between luminosity in the individual types of image elements described above.

In principle, it is sufficient for a process according to the invention to take a picture of a controlled surface after completing a controlled work step (cleaning, chemical treatment, coating) and analyze this image. In processes that change the surface only slightly (cleaning, conversion processing by layer transformation in the submicrometer area), the luminance distribution of the image can be impressed mainly by the structure of the metal surface itself. For example, this can be the case when dipping steel into molten zinc. when the luminous distribution of the metal surface is strongly influenced by the crystalline structure of the zinc layer. In such cases, it may be advisable, in order to improve the performance of the process according to the invention, that the structure of the substrate be computationally removed prior to true image analysis. This compares the image of the surface before the controlled work step with that of the same surface cutout after that work step. This is done in such a way that at least roughly equal positions of the metal or plastic surface form any first image before and any other image after chemical treatment or application of the layer in accordance with a partial step a) before or after performing partial steps b) and c) some other image that, at least approximately, overlaps the two images by finding characteristic locations of a metal or plastic surface that are recognizable in both images, transferring those characteristic points to at least approximately the overlaps in the two images and then the brightness of the pixels or medium subtracts the luminance of the pixels in the first image from the luminosities of the corresponding pixels or pixels in the second image, before partial step d) and, if desired, e) is performed with the second image.

Therefore, it is necessary to bring the two images at least so far as to cover them so that the corresponding image elements in the two images overlap at least substantially. As a rule, however, it is not compulsory for the two images to overlap so closely that each pixel of the first image fits in with the corresponding pixel of the second image. Depending on the size of the image elements, deviations of the size range from millimeter particles to several millimeters are therefore permissible. Characteristic points that can be drawn into the computational overlap of the two images are, for example, angles or edges obtained by bending or folding, or small points of contact such as welding points, for example.

Alternatively, both images can be computationally covered so that one image moves relative to the other image until the squares of the luminance differences of the cutouts imprinted on each other become minimal. This process, however, is computationally intensive and therefore less suitable for rapid quality control.

The process can be used, for example, to control the quality of cleaning processes, chemical conversion processes and layering processes. For example, it can be determined which area of difference in luminance of adjacent image elements is considered as the normal range. An area of major difference can be defined as a control area. However, it is permissible if the differences in luminance of adjacent image elements of an individual image up to n times the size are in this control range, where n> 1 represents a predetermined number. If more than n differences are found lying in the control area, or if at least one difference exceeds the control area, one of the following actions can be activated automatically:

i) broadcast a warning message locally or at a remote location;

ii) start testing at least one of the machining or application means by which metal or plastic surfaces come into contact before partial step a);

iii) switching off the device used for cleaning and / or hydrofishing, chemical treatment or coating.

Of course, both measures i) and ii) or i) and iii) can be triggered at the same time. Furthermore, it is preferably provided in the method according to the invention that the number of differences between the luminance of adjacent image elements lying in the standard area, in the control area and outside the control area, is recorded continuously on a data carrier. Preferably, the recording is made by attributing differences to the treated object to which they relate. Therefore, machine-readable, machine-readable code may be used. As a workpiece, for example, it may be a steel coil or a steel coil. what a cut out of it or in the case of the car industry a particular vehicle. This makes it possible for labels to automatically manage and archive those labels that are strictly necessary for quality control and assurance.

Further, it can be envisaged that during the implementation of the process according to the invention, trends in the movement of differences between the luminances of adjacent image elements are recognized. For example, the fact that a process step according to the invention (cleaning, conversion, coating, coating, etc.) may be increasingly lost in quality, which can be recognized by always moving from image to image, such as bodywork to bodywork more differences beyond the standard range. If such a phenomenon is detected, it can be envisaged that the control system, within the process of the invention, emits a warning message locally and / or at a remote location.

In the context of this discovery, the local broadcast of a warning message is intended to mean that a warning message is sent within the drive in which the process of the invention is performed. By remote location, however, the opposite is meant outside the plant in which the process of the invention is carried out. By transmitting to a remote location, those machining steps that must be controlled by the process of the invention can be controlled and monitored from some location outside the associated plant. For example, this remote location may be located with the manufacturer of the chemicals used for controlled surface treatment steps. In this way, the manufacturer of processing (process) chemicals is promptly informed if the appropriate processing steps for the user of these chemicals are within the standards.

If the maximum allowable number of differences in the control range of differences in luminance of adjacent image elements is exceeded or differences outside the control area are detected, the computerized control system conducting the process according to the invention according to alternative ii) may trigger the verification of at least one machining means or coating agents which contact the metal or plastic surface prior to the partial step a). This can also be triggered automatically if the control system in the process according to the invention determines that the number of differences lying in the control range increases by some setpoint.

It is also possible to take into account whether the greater number of prominent differences between the luminosities of adjacent image elements are independent of one another and probably represent many dots of interference, or whether these prominent differences lie on a line or describe a surface and therefore belong to a single but widespread the jamming spot. Depending on the results of such an analysis, different measures can be envisaged. For example, some jamming line may be allowed, but more point defects may not.

If a method according to the invention is used, for example, to control the quality of some layered phosphating of a metal surface, it can be envisaged as a measure ii) checking the predefined parameters of the phosphating bath or the treatment baths included therein, such as cleaning baths or activation baths. For example, it may be possible to start by automatically analyzing the predefined parameters of the phosphating bath, the activation bath or the cleaning bath. For cleaning baths, for example, one or more of the following parameters can be automatically analyzed: alkalinity, surfactant content and / or fat load of the cleaning bath. To this can be added automatic analysis procedures, such as those described in German patent reports 198 02 725. 198 14 500 and 198 20 800.

In connection with these bath control measures, it can be re-envisaged that corrective measures will be automatically introduced at certain deviations of the bath parameters from the desired values.

By combining the control method of the invention and the analysis of the previously included treatment baths, which is triggered at certain deviations by the control system, which again, if necessary, results in automatic corrective measures for the assembly of the treatment baths, it is possible to ensure the result of the surface treatment automatically and continuously, without the need for human intervention. It is recommended that the results of the control procedure according to the invention, as well as the results of deviations of the analytical and corrective actions taken, be stored for later evaluation.

As the most effective measure (iii), the control system may be automatically switched off if the number of differences between the luminances of adjacent image elements in and / or out of the control range exceeds the predetermined number. Of course, it is of course preferable for the system to issue a local message and / or at a remote location an appropriate message, allowing the device's operating personnel to intervene as quickly as possible and remove the resulting error.

The method of the invention can further be used to control the distribution of particles in any particle stream, in which at least one image of a scattered particle jet spray is controlled by means of one or more nozzles. For this purpose, it is advantageous to capture an image substantially perpendicular to the spray axis in order to prevent any major perspective distortion. As emphasized above, minor perspective distortions can be compensated by dividing the image into so many image elements that each image element comprises an equally large cutout of the spray jet and / or a jet of surrounding space.

Particularly for a fan spray jet, it may be sufficient to mark a single image of the spray jet, which is taken preferably perpendicular to the plane of the fan. However, in the case of a conical spray jet, it may be advantageous to have multiple spray jet images taken from different angles. That is, in this case, the partial steps a) to e) of the process according to the invention are repeated one or more times with images whose image planes (and thus their plane normals) form certain angles. In any major cone spray jet, it may be sufficient to carry out the process of the invention by means of 2 images, the image planes of which are substantially orthogonal to each other.

This may be, for example, a jet whose particles consist of droplets of some solution or suspension which solidify into a solid in the jet. Or they are droplets of some kind that melt into solid particles in a jet of particles. Example no. 1 describes a typical spray drying process in which a solution or suspension of a substance is sprayed into a vacuum and / or into a zone of elevated temperature, whereby the solvent and / or the suspending agent evaporates. This gives the substance a powder form. Such processes are used in different industries. For example, mention should be made of the use in the food industry where food and edible powders are produced in this way. Milk and coffee powder should be mentioned as an example. Further, such spray drying processes are commonplace, for example, in the detergent industry in order to obtain detergents in powder form.

The process according to the invention allows to determine whether the spray device is operating according to the regulations, whether the spatial arrangement of the spray jet falls outside the standard range due to the occurrence of nozzle wear, or if the spray nozzles become clogged due to the clogging of individual nozzles. optimally or if it is possible to expect a reduced flow due to disturbances.

Further, the spray jet particles may be droplets of any solution, suspension or melt sprayed onto the surface in order to form a layer on the surface. Liquid varnishes or varnish varieties are an example of this. However, they may also be solid particles with which a surface is covered. Powder coating (Pulver lackierung) is an example of this. In both cases, the irregularities of the spray jet lead to uneven coverage of the surface and thus to poor quality.

Analogous to the measures described above in the control of surface treatment processes, it may also be envisaged in controlling the distribution of particles in a spray jet that an automatic message is sent automatically or at a remote location if the difference in luminance of adjacent image elements inside a particle jet exceeds a certain value. How it can be determined that the image elements lie inside or outside the spray jet will be further explained in the case of controlling the exit angle of the spray jet. Thus, continuous and automatic control of the spray process is possible without the need for human intervention.

In the following embodiment, a process for controlling the exit angle of a particle jet, which is performed by means of a spray with one or more nozzles, may be used. This may be important, for example, in spray drying or spray curing processes. Therefore, the spray jet should not be too fluffy, since otherwise it is drying or drying. the solidification is unreliable and there is a risk that the particles will interfere with each other. On the other hand, the spray jet should not be branched too wide, as it must prevent it from sticking or sticking. viscous droplets reach the wall of the spray tower and firmly adhere to it. The outlet angle of the spray jet is, however, somewhat predetermined by the arrangement of the nozzles, but may be used in the use of the nozzle. change the nozzle change. Further, the outlet angle of the spray jet depends on the correct setting of the spray pressure.

If a spray jet is used to cover surfaces such as lacquer application or to protect the floor of a car, the correct spray spray angle must also be observed. Only at a properly adjusted exit angle of the spray jet does the overlay surface be covered sufficiently evenly and with the correct thickness of the layer.

In order to determine the angle of the spray jet, the extracted nozzle of the spray jet is first of all based on the width of the spray jet. For mass production by means of spray drying, such as detergents, spray jets have a considerable dimension - within one meter. A cutout must therefore be taken to determine the exit angle. A cutout with a side length of up to 5 m is suitable for this. In the case of spray jets used for covering surfaces with, for example, lacquer or bottom protection, it is generally sufficient to have a cutout with a side length in the range of a few centimeters to one meter. In the specific case of a bottom protection order, for example, a suitable cutout with a side length in the range of about 10 to about 50 cm, especially in the range of about 30x30 cm.

In the process for controlling the angle of the spray jet according to the invention, it is preferable to disassemble an image into rows perpendicular to the axis of the spray jet. A sufficient number of pixels in each row is selected (in the range of from about 10 to about 200, preferably from about 15 to about 100 pixels per row) and differences between the luminances of adjacent pixels are formed, starting from any edge of the image. The differences in the luminance of image elements that represent a pure background are relatively small in adequate lighting. The differences in luminance of adjacent image elements, the first of which mainly comprises a background image and the second mainly a spray jet image, have an unusually high value. The differences in the luminance of adjacent image elements, which consist solely of the spray jet, in turn become relatively small (until the spray jet becomes too inhomogeneous). Thus, during the formation of differences, the tipa goes from the background to the spray jet, it moves over the spray jet until it reaches the background area again. The last major difference between the luminosities of adjacent image elements then corresponds with a very high probability of the edge of the spray jet. Thus, along one line, small differences in the adjacent image elements are detected first, a first high difference in the capture of a single spray jet edge, a slight fluctuating difference in the spray jet area, and once again a particularly high difference relating to the second spray jet edge. Of course, this procedure assumes that an image cutout so large that a background image is captured on both sides besides the spray jet image is selected. This can be automatically controlled by determining whether the image leads to the previously described pattern of differences.

Preferably, by determining differences in the luminosity of adjacent elements, it starts in the image row closest to the scattering head. Because its dimensions are known, it is best to be sure that the image type contains, in its center region, a spray jet image and a background image in both edge regions. Now, the row-by-line evaluation continues from the nozzle head away, and is controlled if, in each successive row, extreme differences between the luminosities of adjacent image elements move outwards. This will continue for as long as the image type has a background spray and a spray jet in both its ends. This procedure is interrupted when the extreme differences between the luminosities of adjacent image elements reach the edge region of their image species, because then there is a danger that the image species only comprise a spray jet but no background. However, the information requested is no longer in this image type.

Based on these observations, it is possible to determine the outlet angle of the spray jet if a straight line through at least 2 characteristic points, preferably a straightening line through several points, is passed through the points of the first large difference between the luminosities of adjacent image elements. Similarly, a different line or line is placed through the points of each last high difference in the luminance of adjacent image elements in each species. balancing line. The course of the two lines in the figure can be represented as a vector. Using vector algebra, the discharge angle of the spray jet is thus calculated.

In many cases, however, it is sufficient not to accurately determine the exit angle of the spray jet, but to control if the exit angle remains within the desired range over time. For this purpose, the evaluation apparatus of the method according to the invention can be set in which area of the image must first have the first and last high differences between the luminosities of adjacent image elements on the selected image row. If the extreme values of the luminance differences found are found deep inside, the exit angle of the spray jet decreases, but if they lie far outside, the exit angle increases.

If the evaluation device in the process according to the invention calculates the spray jet angle, for example by means of a vector calculus or, lastly, by comparing the desired positions of the extreme differences in the individual image types with their actual positions, it is found that the spray jet angle is beyond some allowable areas changed, ie has not reached or exceeded a certain angle range, one or more of the following actions may be taken:

i) delivering a warning message;

ii) changing the spray pressure in the direction that returns the spray jet exit angle to the predetermined angle range;

iii) a change in the viscosity of the composition from which the spray jet is generated in the direction that returns the exit angle of the spray jet again to the specified angle range;

iv) a change in the electric charge of the particles of the spray .curk or the electric field near the nozzles in the direction that returns the exit angle of the spray jet again to the specified angle range;

v) shutting off the spray jet.

The simplest measure, then, is a warning message sent locally or to a remote location. It may be envisaged that, with a slight change in the exit angle of the spray jet, a warning message may be emitted, but no further action will be taken until only one of the cut-off thresholds for the spray jet angle has been reached, one of the actions ii) to iv). Of course, it can be envisaged that the spray process will be completely interrupted if the spray jet angle does not reach or exceed a certain limit value.

The control system may be varied by changing the spray jet angle by changing the spray pressure, by changing the viscosity of the spray component (for example, after changing the temperature of the component in the spray head) or by changing the electric charge. Further, the control system may be asked which change in the spray conditions should lead to an increase or decrease of the spray jet angle. However, the control system can also be designed to teach itself what measures return the exit angle of the spray jet with high probability to the desired area. For this purpose, the control system can change the possible spray parameters one by one and, each time, by analyzing the spray jet image, examine the consequences of a given measure for the size of the spray jet exit angle.

It may further be provided that the control system, in cases where a change in the scattering parameters does not lead to any desired result, emits an alarm message (local and / or remote location) and at the same time proposes a possible action. Possible action options include cleaning or replacing nozzles. It is additionally preferred that the control system be shut off in such cases.

This process of controlling the exit angle of the spray jet can be used, for example, for painting surfaces and especially for applying a protective layer to the bottom of a vehicle. As a measure for correcting the exit angle of the spray jet, in particular in the case of application of the bottom protection, variation of the temperature of the applied protective substance for the bottom in the area of the spray head is offered. The changes in the viscosity of the protective substance for the bottom thus obtained have a particularly noticeable effect on the exit angle of the spray jet, and corrective action can be achieved with this measure.

Also preferred in this embodiment of the invention is the storage of the results of the exit angle control of the spray jet for later evaluation and / or quality control. Preferably, the correlation between the exit angle determined at a given moment and that moment and / or between the angle and the object coated at that moment is recorded.

Implementation examples

Example 1:

Zinc phosphate layer homogeneity control on steel sheet

Figure 1 shows in the upper half a video of a zinc phosphate sheet of phosphated steel sheet. In the lower part of the figure, the phosphate layer was deliberately damaged in order to simulate the phosphate defect. The inverted raster is, in both cases, a division into individual image elements for which differences in average luminance according to the invention are to be determined.

Both cutouts have dimensions of approximately 11x8 cm.

Figure 2 presents the result of identifying differences in the average luminosities of adjacent image elements of each individually longitudinally in a horizontal line image. The diagram on the upper half of the picture corresponds to the intact phosphate sheet in the upper half of FIG. 1 and the diagram in the lower half corresponds to the damaged phosphate layer in the lower half of FIG. 1. Differences between the luminosities of adjacent image elements along horizontal lines in FIG. 1 are shown as horizontal lines in FIG. 2. The individual horizontal types of image elements in FIG. 1 correspond to the individual species in FIG. 2.

From the upper half of FIG. 2 it can be clearly discerned that only small differences between the mean luminosities of adjacent image elements result from a seamless and uniformly phosphated sheet. In contrast, disturbances in the phosphate layer in the lower half of FIG. 1, there are large differences between the luminosities of adjacent elements, which is reflected in the lower half of FIG. 2. The phosphating disturbance can thus be automatically recognized so that the differences between the luminances of the adjacent image elements exceed a predetermined minimum value. At the same time, the lower half of FIG. 2 so that the extension and location of the phosphating defect can be recognized.

Example 2:

Control of the exit angle of the spray jet for the protection of the bottom of the vehicle

FIG. 3 shows in its lower half a video of a spray jet for applying a protective layer to the bottom of a vehicle body. The video is optically filtered, making the boundaries of the spray jet and its inhomogeneity clearly recognizable. The grid specifies the selected image elements, with a total image cutout of approximately 16x12 cm. The upper half of the figure shows the result of determining the differences in the mean luminance of adjacent image elements along each of the horizontally lying types. One above the other lying rows in the upper half of the image correspond to one above the other lying rows of image elements in the lower half of the image. The external limitations of the spray jet as well as the inhomogeneities within the spray jet are distinguished by the particularly large differences between the luminosities of adjacent elements.

At the same time, the homogeneity of the spray jet can be evaluated in this way. It can further be controlled whether the discharge angle of the spray jet changes over time. This compares the recorded videos to each other at different times and verifies that the edge of the spray jet on each clip falls on the corresponding image elements each time.

The exit angle, however, can be directly determined from the representation of the differences between the luminosities of adjacent image elements according to the upper half of FIG. 3. This can be done using the following algorithm:

If the spray jet boundary lines described above are given by the following linear equations y = a * x + b and y = c * x + d, the exit angle of the spray jet a is given by the formula:

- (1 + a * c) cos a = __.

sqrt (l + a * a) * sqrt (l + c * c) where sqrt is a function of square root. The connection point and orientation of the linear equation system in the coordinate system used to determine the exit angle are irrelevant and can be chosen from practical aspects. However, both boundary lines must be represented in the same coordinate system and must have the same distances on the x-axis and on the y-axis correspond to the same distances on the object (in our case, the spray jet). Typically, the coordinate system is chosen so that, for example, the x-axis runs exactly horizontally and the y-axis vertically on the recorded video.

FIG. 4 shows the application of this method to a spray jet according to FIG. 3. In the upper part of FIG. 4 is a schematic representation of the location of each species in FIG. 3 (above), each time the outermost major differences between the luminosities of adjacent image elements are determined. These correspond to the spray line boundary lines. Below it, linear equations are given for each of the two boundary lines of the spray jet and the exit angle of the spray jet is calculated from it by vector analysis. In our case, we get 47 ^0,

List of pictures

FIG. 1

Video of a cut-out of a layered zinc phosphate solution of phosphated steel sheet.

Above: Phosphate layer as obtained

Below: phosphate layer subsequently scratched with damage The size of each of the two cutouts is approximately 11x8 cm.

FIG. 2

A graphical representation of the differences between the luminosities of adjacent image elements of FIG. 1. The upper and lower half of the image are mutually compatible. One above the other lying species in FIG. 2 correspond to one above the other a series of image elements in FIG. 1.

FIG. 3

Below: An optically aligned video of a spray jet for applying a bottom protection layer. The cutout size is approximately 16x12 cm.

Above: By type, the difference between the luminosities of adjacent image elements in the lower part of the image is applied. One above the other lying rows correspond to one above the other lying array of image elements.

FIG. 4

Above: The schematic position of each individual externally significant difference between the luminosities of adjacent image elements in FIG. 3 - above indicating the spray jet limitation.

Below: Calculation of the spray jet exit angle.

Claims (19)

Patent claims A method for controlling the distribution of structures on a surface or particles in a space, characterized in that a) at least one two-dimensional image of the distribution shall be produced optically or electronically, divided into pixels and the luminance of each pixel stored in digital form on a data medium, b) divides the image or cutout of that image into a specified number in sets of sorted image elements, each image element comprising at least four pixels, c) determine the average luminance of each pixel by determining the average luminance of each pixel of that pixel, d) determine, along the first set of image elements, the difference between the mean luminosities of the adjacent image elements and, for the machine, be legibly written to the data medium and / or displayed in a diagram in order to obtain a local correlation between the differences and the position of the corresponding image elements, and in case of desire e) Step d) Repeat for a pre-selected number of the following sets of image elements, which essentially run in parallel with the first set. Method according to claim 1, characterized in that, before determining the differences between the mean luminances of adjacent pixels, an illumination correction is provided by subtracting the previously selected corrective value from the luminance of each individual pixel before determining the mean luminance of the pixels or from the average luminances of individual pixels, or to those luminosities, a pre-selected correction value is added, whereby individual pixels or pixels of the corresponding correction value are described by a plot over an image or a cutout. Method according to one or both of Claims 1 and 2, characterized in that rectangular or trapezoidal cutouts whose side lengths correspond to lengths in real space in the range of 1 mm to 5 m are selected. Method according to one or more of Claims 1 to 3, characterized in that the partial steps d) and e) are repeated with another predefined set of image elements, forming with the first predefined set of image elements pre-selected as in the range from 60 ° to 12O ^0, Method according to one or more of Claims 1 to 4, characterized in that the distribution of the structures on a metal or plastic surface is controlled. Method according to claim 5, characterized in that it is a metal or plastic surface on which the formation of water droplets is controlled after preliminary cleaning and / or hydrofirming. A method according to claim 5, characterized in that it is a metal or plastic surface that is subjected to chemical treatment or coating. Process according to claim 7, characterized in that it is a metal surface that is subjected to chemical treatment in the form of chromatization, treatment with an acidic solution of simple and / or compound fluorides, treatment with some solution of transition metal compounds or layered or non-acidic phosphating. A method according to claim 7, characterized in that it is a metal or plastic surface which is subjected to a coating of organic organic matter. Method according to one or more of Claims 6 to 9, characterized in that at least roughly equal positions of the metal or plastic surface form the first image before and any other image after cleaning and / or hydrofishing, chemical treatment or application of the layer in a partially partial manner steps a) before or after performing partial steps b) and c) for at least the second image, in order that the two images be computationally overlapped by finding the characteristic locations of the metal or plastic surface that are recognizable in both images, transfer these characteristic spots on both images at least approximately to the overlap and then subtracting the luminance of the pixels or medium luminance of the pixels in the first image from the luminosity of the corresponding pixels or pixels in the second image, before partial step d) is carried out with the second image and optionally a partial step e ). Process according to one or more of Claims 5 to 10, characterized in. to automatically initiate one or more of the following actions if the difference between the luminance of adjacent pixels exceeds a predetermined value at least once or at least n times, with n being a predetermined number: i) delivering a warning message; ii) start testing at least one of the machining or application means by which metal or plastic surfaces come into contact before partial step a); iii) switching off the device used for cleaning and / or hydrophilizing. chemical treatment or coating. Method according to one or more of Claims 1 to 4 for controlling the particle distribution in a particle stream, characterized in that at least one spray image is controlled by means of one or more nozzles of the executed particle jet to be recorded substantially vertically to the spray os. A method according to claim 12, characterized in that the partial steps a) to e) are repeated one or more times with images whose image planes form certain angles with each other. Process according to one or both of Claims 12 and 13, characterized in that the particles and droplets are of a solution or suspension which is dried into particulate matter in a jet of particles or that they are droplets of a certain melt which is particle jet. solidifies into solids. Process according to one or both of Claims 12 and 13, characterized in that the particles are droplets of a solution, suspension or melt that is dispersed on a * surface 'in order to form a coating on that surface. Method according to one or both of Claims 12 and 13, characterized in that the particles are solid particles that cover a surface. Method according to one or more of Claims 12 to 16, characterized in that an alert message is automatically issued if the difference between the luminosities of adjacent image elements inside the particle jet exceeds a predetermined value. value. Process according to one or more of Claims 12 to 16, characterized in that the outlet angle of a particle jet is controlled, which is spray-supported by means of one or more nozzles. A method according to claim 18, characterized in that one or more of the following actions is automatically initiated if the outlet angle of the spray jet does not reach or exceed a predetermined angle range: i) delivering a warning message; ii) changing the spray pressure in the direction that returns the spray jet exit angle to the predetermined angle range; iii) a change in the viscosity of the composition from which the spray jet is generated in the direction that returns the exit angle of the spray jet again to the specified angle range; iv) changing the electric charge of the spray jet particles or the electric field near the nozzles in the direction that returns the spray jet exit angle back to the specified angle range; v) shutting off the spray jet.