DISHWASHER AND METHOD FOR CLEANING WASHWARE IN A DISHWASHER
20200281440 · 2020-09-10
Inventors
- Harald DISCH (Elzach, DE)
- Björn S. HIMMELSBACH (Offenburg, DE)
- Markus Heidt (Kehl, DE)
- Verena WIEDENHÖFER (Rheinau, DE)
Cpc classification
A47L2401/30
HUMAN NECESSITIES
Y02B40/00
GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
A47L15/0021
HUMAN NECESSITIES
G01N21/255
PHYSICS
A47L15/46
HUMAN NECESSITIES
A47L2501/05
HUMAN NECESSITIES
A47L15/4295
HUMAN NECESSITIES
A47L2401/04
HUMAN NECESSITIES
A47L2501/30
HUMAN NECESSITIES
A47L15/0047
HUMAN NECESSITIES
A47L15/0081
HUMAN NECESSITIES
A47L15/0078
HUMAN NECESSITIES
International classification
A47L15/46
HUMAN NECESSITIES
A47L15/00
HUMAN NECESSITIES
G01N21/25
PHYSICS
Abstract
A dishwasher (1), in particular a commercial dishwasher (1), which is in the form of a box-type dishwasher and has a treatment chamber (2) which can be closed and in which washware can be treated in accordance with a predefined or predefinable treatment program, wherein a washware detector apparatus (51) is designed to detect the type of washware to be, wherein the washware detector apparatus (51) has an optical identification system, which is oriented toward the washware to be treated, and an evaluation device, wherein the optical identification system is designed to record at least a portion of the light reflected by the washware to be treated as a reflection image, and wherein the evaluation device is designed to automatically distinguish between different types of washware on the basis of the recorded reflection image, wherein the extent of the reflection provides information about the type of washware.
Claims
1-18. (canceled).
19. A box-type dishwasher, comprising: a treatment chamber which can be closed and in which washware can be treated in accordance with a predefined or predefinable treatment program; a washware detector apparatus positioned and configured to detect the type of washware to be treated in the treatment chamber, wherein the washware detector apparatus has an optical identification system, which is oriented toward the washware to be treated in the treatment chamber, and an evaluation device, wherein the optical identification system is designed to record at least a portion of the light reflected by the washware to be treated as a reflection image, and wherein the evaluation device is designed to automatically distinguish between different types of washware on the basis of the recorded reflection image, wherein an extent of reflection in the reflection image provides information about the type of washware.
20. The box-type dishwasher as claimed in claim 19, wherein the optical identification system is further designed to record the color of the light reflected by the washware to be treated, and wherein the evaluation device is further designed to distinguish between different types of washware on the basis of the recorded color.
21. The box-type dishwasher as claimed in claim 19, wherein the optical identification system has a camera which is arranged within the treatment chamber in or in the vicinity of an upper end region on a rear wall or a front wall of the treatment chamber, and wherein the optical identification system further has a lighting system with at least one light source which is designed to allow homogeneous and diffuse illumination of at least a portion of the washware to be treated with light at a color temperature of between 5200 and 5700 K.
22. A method for cleaning washware in a box-type dishwasher that has a treatment chamber which can be closed and in which washware can be treated in accordance with a predefined or predefinable treatment program, wherein the method comprises the following method steps: i) a type of washware to be treated in the treatment chamber is automatically detected; ii) a predefined or predefinable treatment program for treating the washware is automatically selected depending on the type of washware detected; and iii) the process parameters associated with the selected treatment program are automatically adjusted, wherein, in method step i), data in the form of a 2-dimensional image, relating to the washware to be treated is recorded and digitized with the aid of an optical identification system, and patterns are obtained from the data, which patterns serve to classify and/or detect the type of washware to be treated, wherein a reflection of light on the washware to be treated is used as at least one measure for detecting the type of washware to be treated.
23. The method as claimed in claim 22, wherein method step i) comprises the following: generating, on the basis of the recorded data, a corresponding grayscale value histogram; and detecting the type of washware to be treated on the basis of at least one predefined or predefinable value range of the generated grayscale value histogram.
24. The method as claimed in claim 22, wherein the data recorded in method step i) is a 2-dimensional color image, and wherein the color is used as a criterion for classifying and detecting the washware to be treated, wherein, for this purpose, a color histogram is formed for each object detected in the recorded color image and washware is categorized into types of washware on the basis of the color histogram.
25. The method as claimed in claim 22, wherein the data recorded in method step i) is a 2-dimensional image, and wherein the shape and/or size of detected objects is used for classifying and/or detecting the washware to be treated, wherein, for this purpose, edge detection is carried out on the recorded image and objects are detected and/or classified according to object surface area, roundness and/or rectangularity.
26. The method as claimed in claim 22, wherein the type of washware to be treated is automatically detected after the treatment chamber is closed and after a prewash phase, in which the washware loaded into the treatment chamber is prewashed, is complete.
27. The method as claimed in claim 22, wherein method step i) comprises the following substeps: a) the recorded 2-dimensional image is in the form of a grayscale value image or in the form of a color image; and b) if the 2-dimensional image is in the form of a color image, the color image is preprocessed, wherein the preprocessing in method substep b) comprises filtering and/or converting the color image recorded in method substep a) into a grayscale value image, wherein method step i) further comprises the following substep: c1) detecting whether there are reflections in the recorded 2-dimensional image by generating a corresponding grayscale value histogram preferably on the basis of the grayscale value image generated in method substep b) or recorded in method step a), wherein, when corresponding reflections are detected in method step c1), washware is categorized into a cutlery type of washware or into a pots/pans type of washware depending on a degree of the detected reflection and/or the size of a detected, average reflection surface area.
28. The method as claimed in claim 27, wherein, when no corresponding reflections are detected in method step c1), glass rack detection takes place, and, wherein, when it is identified during the course of glass rack detection that a glass rack is present, washware is categorized into a glass type of washware, wherein, however, when a glass rack is not identified during the course of glass rack detection, washware is categorized into a porcelain type of washware.
29. The method as claimed in claim 22, wherein method step i) comprises the following substeps: a) a 2-dimensional image, in the form of a grayscale value image or a color image, is recorded with the aid of an optical identification system; b) the recorded image is preprocessed; and c2) edge detection takes place on the preprocessed image for the purpose of object detection, wherein the preprocessing in method substep b) comprises filtering.
30. The method as claimed in claim 29, wherein a color image is recorded in method substep a), and wherein the recorded color image is additionally stored as a grayscale value image in method substep b), and wherein, when at least one object is detected in method substep c2), method step i) further comprises the following substep: d) a HSV histogram is formed for each object detected in the color image, and washware is categorized into types of washware, including into a porcelain type of washware and into a pots/pans type of washware, on the basis of said HSV histogram.
31. The method as claimed in claim 29, wherein, when no object is detected in method substep c2), method step i) further comprises the following substep: e1) detecting whether there are reflections in the recorded 2-dimensional image by a corresponding grayscale value histogram being generated on the basis of the grayscale value image stored in method substep b), wherein, when corresponding reflections are detected in method substep e1), washware is categorized into a cutlery type of washware.
32. The method as claimed in claim 21, wherein, when no corresponding reflections are detected in method substep e1), method step i) further comprises the following substep: f1) adopting an HOG method and determining HOG feature vectors for categorizing washware into a glass type of washware if corresponding HOG feature vectors can be determined.
33. The method as claimed in claim 32, wherein, when no corresponding HOG feature vectors for categorizing washware into the glass type of washware can be determined in method substep f1), washware is categorized into a porcelain type of washware.
34. The method as claimed in claim 29, wherein, when no object is detected in method substep c2), method step i) further comprises the following substep: e2) adopting an HOG method and determining HOG feature vectors for categorizing washware into a glass type of washware if corresponding HOG feature vectors can be determined.
35. The method as claimed in claim 34, wherein, when no corresponding HOG feature vectors for categorizing washware into the glass type of washware can be determined in method substep e2), washware is categorized into a porcelain type of washware.
36. The method as claimed in claim 34, wherein, when no corresponding HOG feature vectors for categorizing washware into the glass type of washware can be determined in method substep e2), method step i) further comprises the following substep: f2) detecting whether there are reflections in the recorded 2-dimensional image by a corresponding grayscale value histogram being generated on the basis of the grayscale value image stored in method substep b), and wherein, when corresponding reflections are detected in method substep f2), washware is categorized into a cutlery type of washware, but wherein, when no corresponding reflections are detected in method substep f2), washware is categorized into a porcelain type of washware.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
[0031] The invention is described in more detail below using exemplary embodiments with reference to the drawings in which:
[0032]
[0033]
[0034]
[0035]
[0036]
[0037]
[0038]
[0039]
[0040]
DETAILED DESCRIPTION
[0041] The dishwasher 1 according to the invention has a treatment chamber 2 for cleaning washware, not illustrated in the drawings. A washing tank 4 in which liquid can flow back out of the wash chamber 2 due to the force of gravity is arranged beneath the treatment chamber 2. The washing tank 4 can be covered at the transition to the treatment chamber 2 with the aid of a screen, not illustrated in the drawings. The washing tank 4 contains washing liquid 6 which is usually water to which, if appropriate, detergent can be supplied automatically in a controlled manner by a detergent metering device, not illustrated in the drawings. The washing liquid 6 can be conveyed by a washing pump 8 via a washing line system 10 to washing nozzles 33 and 34 and can be sprayed through these washing nozzles 33, 34 in the treatment chamber 2 onto the washware to be cleaned. The sprayed washing liquid 6 subsequently flows back into the washing tank 4. The lower end 5 of the washing tank 4 can have connected to it a discharge line 3 with an outflow pump 11 in order to empty the washing tank 4 as required.
[0042] In the embodiment, illustrated in
[0043] In the embodiment, illustrated in
[0044] The fresh water container 14 has an inlet 37 which is connected to a fresh water feed line 31. Either pure fresh water or fresh water with rinse aid which is added in a metered manner is supplied to the fresh water container 14 via said fresh water feed line 31.
[0045] However, it is, of course, also conceivable to arrange the abovementioned rinse aid metering device, not illustrated explicitly in the drawings, downstream of the outlet 36 of the fresh water container 14.
[0046] The washing nozzles 33, 34 and final rinse nozzles 7, 9 are in each case preferably arranged above and below the washware region and are directed toward the washware region of the treatment chamber 2. In the embodiment, illustrated in
[0047] In the embodiment, illustrated in
[0048] The non-return device 16 has an outlet 40 which is connected to a water softener device 39 via fresh water feed lines 24 and 41. The water softener device 39 has, firstly, a salt container 42 which is connected to the fresh water feed line 41 and, secondly, first and second water softeners 39a, 39b which are arranged parallel to one another. The two water softeners 39a, 39b which are arranged parallel to one another are connected to the outlet 40 of the non-return device 16 via a corresponding fresh water line system and the fresh water feed line 24. The water softeners 39a, 39b of the water softener device 39 can be operated alternately by suitable actuation of valves 43, in order to soften the fresh water which is supplied to the boiler 21 via the fresh water feed lines 26 and 28 and the fresh water supplied to the fresh water container 14 via the fresh water feed lines 26, 29 and 31.
[0049] The salt container 42 which forms part of the water softener device 39 can contain a suitable salt or a suitable chemical. The salt or chemical serves, as required, for regenerating a water softener agent which is added in a metered manner to the fresh water or a decomposition product which is produced after the metered addition. The salt container 42 can be refilled with the salt or the chemical from the treatment chamber 2 of the dishwasher 1 via an opening which can be closed by way of a lid 44.
[0050] In the illustrated preferred embodiment of the dishwasher 1 according to the invention, the boiler 21 has an integrated steam generator 20. A corresponding steam outlet 23 of the steam generator 20 is formed at the upper region of the boiler 21. The steam outlet 23 of the steam generator 20 is connected via a steam line 32, at a point 37 located above the washing tank 4, to the treatment chamber 2, in order, as required, to introduce the steam which is generated in the steam generator 20 into said treatment chamber. The outlet opening of the steam line 32 is preferably located between the upper nozzles 7, 33 and the lower nozzles 9, 34 of the washing line system 10 or of the final rinse line system 15. However, other positions are, of course, also possible.
[0051] A heater 45 is located in the boiler 21 which serves not only for heating the final rinse liquid, but also for generating steam. Furthermore, a level sensor 46, which, for example, controls a valve 19 of the fresh-water line 17, may be arranged in or on the boiler.
[0052] In the preferred embodiment, illustrated in
[0053] A program control device 50, illustrated merely schematically in the drawings, serves for controlling at least one cleaning program and is designed and connected to the controllable components of the dishwasher 1, such as, for example, to the valves 19, 27 and 30, to the washing pump 8, to the first final rinse pump 12, to the second final rinse pump 13 and/or to a heater control circuit (not explicitly illustrated) for controlling the heater 45, in order to implement a cleaning method (cleaning cycle) comprising the following steps (phases) which are to be carried out in succession: [0054] 1. a washing phase, in which washing liquid 6 is sprayed out of the washing tank 4 through the washing line system 10 into the treatment chamber 2 by means of the washing pump 8 and can then flow back from the treatment chamber 2 into the washing tank 4 due to the force of gravity; [0055] 2. a fresh water final rinse phase, in which heated or unheated fresh water or fresh water to which rinse aid has been added in a metered manner is sprayed as final rinse liquid into the treatment chamber 2 by means of the first final rinse pump 12 or the second final rinse pump 13 and can then flow from the treatment chamber 2 into the washing tank 4 due to the force of gravity; and [0056] 3. (optionally) a drying phase, in which drying air circulates within the treatment chamber 2.
[0057] In a preferred implementation of the program control device 50, said program control device is designed in such a way that it automatically activates the controllable components of the dishwasher 1 as a function of the detected type of washware, such that, after the fresh water final rinse phase, a steam final rinse phase is carried out, in which steam is generated by means of the steam generator 20 which is connected to the fresh water feed line 28 and is conducted into the treatment chamber 2.
[0058] The dishwasher 1 according to the inventionas illustrated, for example, in
[0059] As an alternative or in addition to this, it is conceivable that the washware detector apparatus 51 has at least one detector device 52 by way of which the type of washware can be detected indirectly. This implementation is particularly suitable when the dishwasher 1 is designed in such a way that at least one washware rack (not explicitly illustrated in the drawings), which receives the washware to be treated, can be inserted into the treatment chamber 2 of the dishwasher 1. In this case, the washware to be treated should preferably be sorted according to types of washware in such a way that only washware of a single type is inserted into each washware rack. In this case, provision is preferably made for each washware rack to have an identification feature, such as an identification code for example, which can be detected by the detector device 52. The identification feature which is detected by the detector device 52 then provides information on the type of washware which is received by the washware rack.
[0060] On the other hand, however, it is also conceivable that the detector device 52 is designed to detect the size and/or type of the at least one washware rack which can be received in the treatment chamber 2 of the dishwasher 1 and, on the basis of this information, to draw a conclusion as to the type of washware which is received by the washware rack. In one possible implementation, provision may be made, for example, for the following sorting operation to be performed before the actual treatment of the washware in the dishwasher: [0061] plates which are manufactured from porcelain or a porcelain-like material are inserted into a washware rack which is designed as a plate rack; [0062] cooking utensils which are manufactured from a metal, in particular from stainless steel, are inserted into a washware rack which is designed as a flat rack; [0063] cutlery or items of cutlery which are manufactured from metal, in particular from stainless steel, is/are inserted into a washware rack which is designed as a cutlery rack; and [0064] drinking glasses which are manufactured from glass or a glass-like material are inserted into a washware rack which is designed as a glass rack.
[0065] The remaining types of washware can in this case be inserted, for example, into a washware rack which is designed as a plate rack. When a presorting operation of this kind is performed, it is conceivable, in one possible implementation of the washware detector apparatus 51, that said washware detector apparatus has a detector device 52 which is designed, for example, for detecting the type of washware rack (plate rack, flat rack, cutlery rack, glass rack, etc.) which is inserted or to be inserted into the treatment chamber 2 and thereby to indirectly determine the type of washware which is inserted into the treatment chamber 2 of the dishwasher 1. The type of washware rack can be detected, for example, optically by way of the size, height and/or by way of the shape with the aid of the detector device 52. However, in this case, detection of another suitable identification feature of the washware rack may, of course, also be considered.
[0066] As already mentioned, in the dishwasher 1 according to the present invention, the program control device 50, which is illustrated merely schematically in the drawings, is also provided. According to the invention, the program control device 50 is designed to actuate different actuable components of the dishwasher 1, such as, for example, the respective pumps and valves, in accordance with a predefined or predefinable program sequence in order to thereby be able to adjust suitable process parameters during the individual treatment phases (washing phase, final rinse phase and drying phase). The program control device 50 is connected, in particular via a suitable communication link, to the washware detector apparatus 51 in order to continuously or at predetermined times or events, preferably before the start of the washing cycle, check the type, detected by the washware detector apparatus 51, of washware received in the treatment chamber 2 of the dishwasher 1.
[0067] The program control device 50 is designed, on the basis of the type of washware which is detected by the washware detector apparatus, to automatically identify the washware to be treated. The program control device 50 can preferably automatically identify at least the following washware: [0068] plates which are manufactured from porcelain or a porcelain-like material; [0069] cups which are manufactured from porcelain or a porcelain-like material, glass or a glass-like material; [0070] bowls which are manufactured from porcelain or a porcelain-like material, glass or a glass-like material; [0071] trays or tray-like articles which are manufactured from a plastic material; [0072] containers, in particular food-grade containers, which are manufactured from a metal, in particular from stainless steel; [0073] pots which are manufactured from a metal, in particular from stainless steel; [0074] pans which are manufactured from a metal, in particular from stainless steel; [0075] cutlery and items of cutlery which is/are manufactured from a metal, in particular from stainless steel; and [0076] drinking glasses which are manufactured from glass or a glass-like material.
[0077] In the event that the program control device 50 does not identify the washware on the basis of the detected type of washware, the washware in question is identified as other washware.
[0078] The solution according to the invention is distinguished not only, firstly, by the automatic detection of the type of washware to be treated and, secondly, by the automatic identification of the washware to be treated, but also by the additional functionality of the program control device 50, whereby said program control device is designed to automatically select, as a function of the detected type of washware, a predefined or predefinable treatment program according to which the washware which is received in the treatment chamber 2 is to be treated during at least one treatment phase, and for adjusting the process parameters of the selected treatment program by corresponding actuation of the actuable components of the dishwasher 1. For this purpose, it is conceivable in principle for the program control device 50 to have, for example, a storage device, not explicitly illustrated in
[0079] The invention is based, amongst other things, on the knowledge that the process parameters which are optimal for the individual treatment phases are dependent on the type of washware to be treated. The different types of washware require suitable treatment programs according to which the washware is to be treated, for example, during the washing phase, the fresh water final rinse phase and/or the drying phase. Therefore, for example, only lightly soiled items of tableware, such as, for example, drinking glasses, in contrast to heavily soiled items of cookware with burnt-on food residues which are difficult to remove, such as, for example, cooking utensils, keeping-warm containers, baking molds, etc., require, during the washing phase, only a relatively small quantity of washing liquid sprayed per unit time and only a relatively low nozzle pressure with which the washing liquid is sprayed. Furthermorein comparison with heavily soiled items of cookwareonly a shorter cycle duration of the washing phase is necessary for lightly soiled items of tableware.
[0080] On the other hand, the different types of washware also require coordinated treatment programs for the fresh water final rinse phase which follows after the washing phase. Adaptation, carried out in terms of the type of washware, of the process parameters for the steam final rinse phase and drying phase, which are to be carried out, if appropriate, after the fresh water final rinse phase, is also likewise conceivable.
[0081] As a result, overtreatment, in particular of only lightly soiled items of tableware, such as, for example, drinking glasses, can be efficiently prevented in a simple, but effective, way, so that no more resources in terms of energy, water, chemicals, etc. than are necessary are used during the cleaning process (and, if appropriate, during the drying process) for only lightly soiled items of tableware of this kind.
[0082] The consumption of fresh water, which is sprayed in pure form or mixed with further additives onto the washware, for example, during the fresh water final rinse phase, can also be reduced. As a result of the reduced consumption of fresh water by the dishwasher 1, the consumption of chemicals, in particular rinse aids and/or detergents, can also be reduced, without there being a change in the concentration of chemicals in the respective liquid (washing liquid, final rinse liquid). Owing to a lower fresh water supply, the quantity of water to be heated is lower, with the result that energy can be saved.
[0083] In this case, the invention is based, inter alia, on the knowledge that, in conventional dishwashers, the washing and final rinse cycles of which proceed according to a program which is preset at the factory, overtreatment of the washware often takes place. It is frequently sufficient for an adequate washing and rinsing result simply to have, for example, a shorter cycle duration of the washing phase, a smaller quantity of washing liquid which is sprayed per unit time during the washing phase, a lower nozzle pressure with which the washing liquid is sprayed during the washing phase, a smaller quantity of final rinse liquid which is sprayed overall during the fresh water final rinse phase, and/or a lower temperature of the liquid (washing liquid, final rinse liquid) to be sprayed during the washing phase and fresh water final rinse phase.
[0084] According to the invention, the program control device 50 is accordingly designed in such a way that said program control device automatically actuates the washing pump 8 as a function of the detected type of washware in such a way that at least one of the parameters given below is adapted to the detected type of washware for the washing phase: [0085] the cycle duration of the washing phase; the quantity of washing liquid sprayed per unit time during the washing phase; and [0086] the pressure with which the washing liquid is sprayed during the washing phase.
[0087] A dishwasher 1, in particular a commercial dishwasher, in the form of a box-type dishwasher according to the second preferred embodiment of the solution according to the invention is provided schematically in
[0088] The method according to the invention and, in particular, the automatic detection of the type of washware will be explained in more detail below using various examples with reference to the schematic illustrations of
[0089] Specifically,
[0090] In the following explanations, the washware to be treated is subdivided into four different categories (types). These categories include porcelain, glasses, cutlery and pots and pans. It goes without saying, however, that the present invention is not restricted to these four types of washware but these are types of washware which occur most often in commercial dishwashing. Said washware includes white porcelain and pots and pans made from stainless steel. Furthermore, it is assumed below that glasses are loaded only in the glass racks provided especially for them. This assumption is based on the conventional procedure in commercial dishwashing.
[0091] The objective of the concepts described below is automatic program selection, specifically depending on the type of washware to be treated. In the first concept, as is schematically shown with reference to the flowchart according to
[0092] In the first exemplary embodiment of the method according to the invention, a 2-dimensional image relating to the washware to be cleaned or to be classified is recorded and preferably digitized with the aid of an optical identification system for the purpose of detecting the type of washware to be treated. Patterns are then obtained from this data, which patterns serve to classify and/or detect the type of washware to be treated. The 2-dimensional image which is detected with the aid of the optical identification system is preferably a grayscale value image and further preferably a color image. The recorded image is possibly preprocessed and, in particular, filtered. The preprocessing can also comprise converting a previously recorded color image into a grayscale value image.
[0093] The subsequent evaluation of the recorded 2-dimensional image initially involves detecting whether there are corresponding reflections in the recorded 2-dimensional image. This is preferably performed by a corresponding grayscale value histogram being generated on the basis of the previously generated grayscale value image or the previously recorded grayscale value image. When corresponding reflections are then detected, washware is automatically categorized into the cutlery type of washware or into the pots/pans type of washware, specifically depending on the degree of the detected reflection and/or depending on the size of an average reflection surface area.
[0094] When, however, no corresponding reflections are detected, glass rack detection takes place in the method according to the first exemplary embodiment. When it is identified during the course of glass rack detection that a glass rack is present, washware is automatically categorized into the glass type of washware. When, however, a glass rack is not identified during the course of glass rack detection, washware is automatically categorized into the porcelain type of washware.
[0095] In other words, according to the first exemplary embodiment of the method according to the invention, which method is (at least partially) illustrated in the schematic flowchart according to
[0096] In the exemplary embodiment, the average surface area is measured in order to distinguish between cutlery and pots and pans. It is assumed here that these surface areas are significantly smaller in the case of cutlery than in the case of pots and pans. The reflection is measured by means of the associated grayscale value histogram. The idea here is that reflections on cutlery and pots and pans have a considerably higher peak in a predefined value range than white porcelain in the histogram. This value range preferably corresponds to a pure white in the grayscale value image.
[0097]
[0098] In order to prevent accompanying influences of the stainless steel hood of the dishwasher or the stainless steel inner wall of the treatment chamber of the dishwasher in the histogram, only the region of the dish rack which is received in the treatment chamber should be taken into consideration for forming the histogram.
[0099] It is possible to identify on the basis of the exemplary grayscale value histograms according to
[0100] In order to optimize detection about the extent of the reflection, it is possible to perform classification by means of the object shape and object size. To this end, the following criteria are preferably defined: firstly the surface area, in pixels, of an object found. Furthermore, the roundness of the object is also recorded as a criterion. This is established by way of the smallest possible circle being drawn around the object found and a check then being made in respect of how much is covered by the object surface. The same is then once again additionally calculated using the smallest encompassing rectangle. These give the following three properties in respect of the shape and size of the objects: object surface area, roundness (=circular surface area of the smallest enclosing circle/object surface area), and rectangularity (=rectangular surface area of the smallest enclosing rectangle/object surface area).
[0101]
[0102] A second exemplary embodiment of the method according to the invention for automatically detecting and classifying washware is described below with reference to the schematic flowchart according to
[0103] For the purpose of automatically detecting/classification washware, a 2-dimensional image, in particular a grayscale value image and preferably a color image, is first recorded with the aid of an optical identification system. The recorded image is then possibly preprocessed, wherein the preprocessing comprises, in particular, filtering. Subsequently, edge detection takes place on the preferably preprocessed image for the purpose of object detection.
[0104] Locations in an image at which the intensity values increase sharply along a specific direction in a small space are called edges in image processing. In the present case, the so-called canny operator is preferably used for edge detection. The canny operator attempts not only to detect edges as reliably as possible but at the same time to precisely determine the position of the edges. Furthermore, said canny operator attempts to minimize the number of incorrect edge detections. To this end, smoothing of the image using a Gaussian filter takes place in the first processing step. This is carried out in order to reduce noise in the original image.
[0105] Once the edge strength and edge direction have been determined for each pixel in the image, a check is once again made as to whether the edge is actually an edge. To this end, a check is made with the aid of the second derivative as to whether the pixel is at a local maximum along its edge direction.
[0106] Finally, the canny operator attempts to join all remaining pixels to form a contour. This is preferably performed with the aid of a threshold operation in which a check is made as to whether the edge strength lies between a lower and an upper limit value. If the edge strength lies above the limit value, the pixel is accepted as an edge. If the gradient is between the upper and the lower value, the pixel is recognized only as an edge if it adjoins a pixel with a value above the limit value. If the edge strength lies below the limit value, it is rejected.
[0107] Once the edge detection has taken place in the method according to the flowchart in
[0108] When, however, an object has not been detected during the course of object detection, detection is performed as to whether there are reflections in the recorded 2-dimensional image.
[0109] In order to be able to detect reflections, a corresponding grayscale value histogram is preferably generated on the basis of a previously stored grayscale value image. When it is detected that there are corresponding reflections in the recorded 2-dimensional image, washware is automatically categorized into the cutlery type of washware.
[0110] When, however, no corresponding reflections are detected, an HOG method is adopted and HOG feature vectors for the glass type of washware are determined. If corresponding HOG feature vectors for the glass type of washware can be determined, washware is automatically categorized into the glass type of washware.
[0111] When, however, no corresponding HOG feature vectors for the glass type of washware can be determined, washware is categorized into the porcelain type of washware.
[0112] The flowchart according to
[0113] At the start, the image which is recorded with the aid of the optical identification system and preferably stored on a computer has to be read into a corresponding evaluation device. After it has been read in, the image is ready for processing as a multidimensional array. This array describes the image height, the image width and also the number of color channels. Due to the fact that edge detection which is to be carried out later expects a grayscale value image as an input for object detection, the color image is additionally stored as a grayscale value image. Filtering of the grayscale value image constitutes a further step in the preprocessing. A Gaussian filter is preferably used for said filtering. This operating step is mainly necessary in the case of pots and pans composed of stainless steel. Said pots and pans have thick structures on the surface owing to their material and owing to the reflection. These structures can be considered to be undesired high-frequency interference. These can be reduced by filtering using a low-pass filter, such as a Gaussian filter.
[0114] Edge detection takes place after filtering. The abovementioned canny edge detector is adopted for this purpose. The minimum and maximum values for the threshold value function are passed on as parameters. It has been found that the best results are achieved with a minimum threshold value of 0 and a maximum threshold value of 100.
[0115] In the binary image which is produced as a result of the edge detection, the relevant objects then have to be segmented for evaluation. To this end, the contours are found. Contours are understood to be all points along a boundary with the same color or intensity. When applied to a binary image, the white regions are understood to be objects and the black regions are understood to be background. In order to segment the relevant objects, the binary image is inverted in a first step. This step is necessary so that the object surface area is displayed as white objects and the background is displayed as a black area. In a next step, the background is separated from the desired object with the aid of morphological operations. To this end, all objects are reduced in size by applying erosion. Owing to suitable selection of the structural element, it is possible, as a result, to reduce the size of the non-relevant surface areas to such an extent that they are deleted from the image. The remaining elements are then increased back to their original size by way of dilation.
[0116] In a next step, the features have to be extracted from the detected objects. In this concept, the color is first used as a property. Owing to the preceding object detection, a mask can be created for each object found. With the aid of this mask, it is possible to record the histogram only for the object found. A mask is understood to be a binary image which contains only the object found.
[0117] In this context, it is advantageous when the RGB image, which has been recorded by the camera of the optical recording system, is further converted into an HSV image beforehand.
[0118]
[0119] After extraction of the properties, the object classification is divided into a training phase and a test phase.
[0120] In the same way as a human, the computer naturally cannot classify any objects which it does not identify. For this reason, a training phase has to take place. During this phase, the human has to provide the computer with information as to which class the recorded object properties belong to.
[0121] During the test phase, classification of the unknown object is carried out on the basis of the learnt properties from the training phase.
[0122] In the case of object detection with a rack loaded with glasses, no objects are found. However, due to the program status, only loading with glasses or an empty rack can be present at this point. In order to distinguish between these situations, glasses are sought with the aid of an HOG method (HOG=Histogram of Oriented Gradient).
[0123]
[0124] The invention is not restricted to the exemplary embodiments but is the result of looking at all of the features disclosed in this document together. Rather, the invention is the result of overall consideration of the patent claims and of the description of the exemplary embodiments by a person skilled in the art.
[0125] In particular, it is conceivable that the dishwasher 1 has, furthermore, a detergent metering device for adding detergent in a metered manner to the washing liquid. The program control device 50 should in this case particularly preferably be designed to adjust the quantity of detergent added to the washing liquid in a metered manner depending on the type of washware to be treated. In particular, it is conceivable that an increased quantity of detergent is added in a metered manner when a type of washware which is associated with the second washware group is detected, while a standard value is selected for the quantity of detergent which is added in a metered manner in the case of the other washware groups.
[0126] As an alternative or in addition, it is further conceivable that the dishwasher 1 has a rinse aid metering device for adding rinse aid in a metered manner to the final rinse liquid, wherein the program control device 50 is preferably designed to adjust the quantity of rinse aid which is added in a metered manner depending on the type of washware to be treated. In this case it is, in particular, conceivable that, when washware which is associated with the first group of washware is detected, the quantity of rinse aid which is added in a metered manner is automatically reduced in comparison to the other washware groups.