Method, device and dishwasher for dishwasher

Abstract

The embodiments of the present invention provide a method, an apparatus and a dishwasher for a dishwasher. The dishwasher includes an image acquisition device and a fill light arranged in the dishwasher, and the method includes: controlling the fill light to perform fill light at different brightnesses; controlling the image acquisition device to capture images at different brightnesses; inputting the images into a prediction model to predict the image information of each image; comparing the images according to the image information to determine the target image; and using the brightness corresponding to the target image as the target fill light brightness of the fill light. The embodiments of the present invention utilize the idea of ​​integration, and cleverly combine deep learning with this special scenario, so as to achieve real-time and fast precise control of the light brightness in the dishwasher, and can continuously perform iterative optimization to save network bandwidth and reduce server pressure, thereby bringing a better experience to users.

Description

Technical Field

[0001] The invention relates to the technical field of kitchen and bathroom appliances, and in particular to a method and a device for a dishwasher and the dishwasher. Background Art

[0002] With the booming development of artificial intelligence, users have an increasingly strong demand for smart homes. A dishwasher with artificial intelligence functions will bring users convenient operation and greatly improve the quality of the product and user experience. Automatic identification of tableware placed in the dishwasher by users is a basic requirement for dishwashers. Accurate judgment of the items inside can enable the dishwasher to automatically adjust the water intensity, control oil stains, protect tableware and other functions. This requires adding cameras and fill lights to the dishwasher, especially in a relatively closed inner cavity, where the machine can only provide lighting through fill lights. The quality of the image is directly related to the effect of the fill light, which will directly determine the accuracy of the subsequent image intelligent algorithm. Therefore, it is particularly important to enable the machine to automatically adjust the brightness of the fill light according to the utensils placed inside.

[0003] In the prior art, the detection and segmentation model based on deep convolutional neural network is a common method for judging the items inside the dishwasher, and has been actually applied in many scenarios. Therefore, a feasible method is to transmit the image information of various light brightness back to the server, and use the powerful computing power of the server to drive the detection and segmentation model to analyze various brightness, but this will inevitably lead to an increase in network load and also increase the pressure on the server. Therefore, the prior art still needs to be improved. Summary of the invention

[0004] In order to solve the above technical problems, an object of the present invention is to provide a method and a device for a dishwasher and a dishwasher.

[0005] In order to achieve the above-mentioned purpose, in a first aspect of the present invention, a method for a dishwasher is provided, wherein the dishwasher comprises an image acquisition device and a fill light arranged in the dishwasher, and the method comprises: controlling the fill light to perform fill light at different brightnesses; controlling the image acquisition device to acquire images at different brightnesses; inputting the images into a prediction model to predict the image information of each image; comparing the images according to the image information to determine a target image; and using the brightness corresponding to the target image as the target fill light brightness of the fill light.

[0006] In an embodiment of the present application, the image information includes the number of tableware, and comparing the images according to the image information to determine the target image includes: sorting the images according to the number of tableware; determining the first image with the largest number of tableware from the images; and when the number of first images is only one, determining the first image as the target image.

[0007] In an embodiment of the present application, the image information also includes confidence, and comparing the images according to the image information to determine the target image also includes: when the number of first images is greater than one, sorting the confidences of the first images; and determining the second image with the highest confidence among the first images as the target image.

[0008] In the embodiment of the present application, controlling the fill light to perform fill light with different brightnesses includes: controlling the fill light to change the brightness step by step.

[0009] In an embodiment of the present application, the method also includes: determining relevant parameters of the control program based on the image information of the target image; controlling the dishwasher to operate according to the control program and obtaining the operating results; determining whether the image information matches based on the operating results; if the image information does not match, using the target image as a training sample to train the prediction model to optimize the prediction model.

[0010] In an embodiment of the present application, using the target image as a training sample to train the prediction model to optimize the prediction model includes: sending the target image to a cloud platform; training the prediction model on the cloud platform using the target image as a training sample; and downloading the trained prediction model to the dishwasher to optimize the prediction model.

[0011] In the embodiment of the present application, the prediction model includes any one of the following: Anchor-Based model and Anchor-Free model.

[0012] In a second aspect of the present application, a device for a dishwasher is also provided, including a fill light for filling light inside the dishwasher; an image acquisition device for acquiring images inside the dishwasher; and a processor configured to: control the fill light to fill light at different brightnesses; control the image acquisition device to acquire images at different brightnesses; input the images into a prediction model to predict image information of each image; compare the images according to the image information to determine a target image; and use the brightness corresponding to the target image as the target fill light brightness of the fill light.

[0013] In an embodiment of the present application, the image information includes the number of tableware, and the processor is configured to compare the images according to the image information to determine the target image, including the processor being configured to: sort the images according to the number of tableware; determine the first image with the largest number of tableware from the images; and when the number of the first images is one, determine the first image as the target image.

[0014] In an embodiment of the present application, the image information also includes confidence, and the processor is configured to compare the images according to the image information to determine the target image. The processor is also configured to: when the number of first images is greater than one, sort the confidences of the first images; and determine the second image with the highest confidence among the first images as the target image.

[0015] In an embodiment of the present application, the processor is configured to control the fill light to perform fill light with different brightnesses, including the processor being configured to: control the fill light to change the brightness step by step.

[0016] In an embodiment of the present application, the processor also includes: determining relevant parameters of the control program based on image information of the target image; controlling the dishwasher to run according to the control program and obtaining the running results; determining whether the image information matches based on the running results; if the image information does not match, using the target image as a training sample to train the prediction model to optimize the prediction model.

[0017] In an embodiment of the present application, using the target image as a training sample to train the prediction model to optimize the prediction model includes: sending the target image to a cloud platform; training the prediction model on the cloud platform using the target image as a training sample; and downloading the trained prediction model to the dishwasher to optimize the prediction model.

[0018] In the embodiment of the present application, the prediction model includes any one of the following: Anchor-Based model and Anchor-Free model.

[0019] An embodiment of the present invention provides a dishwasher, comprising the above-mentioned device for a dishwasher.

[0020] On the other hand, a machine-readable storage medium is provided. The machine-readable storage medium stores instructions. When the instructions are executed by a processor, the processor is enabled to execute the method for a dishwasher according to the above.

[0021] In an embodiment of the present invention, a computer program product is further provided, including a computer program, wherein the computer program implements the above-mentioned method for a dishwasher when executed by a processor.

[0022] Through the above technical scheme, the method provided by the embodiment of the present invention is different from the previous method of calculating the most reasonable fill light brightness through images. Instead, it first controls the fill light to perform fill light at different brightness values; controls the image acquisition device to acquire multiple images, each image corresponding to a brightness value; then compares the image information in the image to determine the target image, and uses the brightness value corresponding to the target image as the target fill light brightness value of the fill light, so that the most appropriate brightness value required by the dishwasher can be determined in real time and quickly. The embodiment of the present invention utilizes the idea of ​​integration, cleverly combining deep learning with this special scenario, and can achieve real-time and fast precise control of light brightness in the dishwasher, and can continuously iterate and optimize in a loop to save network bandwidth and reduce server pressure, bringing a better experience to users.

[0023] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0025] Figure 1 is a structural schematic diagram of a dishwasher provided by an embodiment of the present invention;

[0026] Figure 2 is a flow chart of a method of a dishwasher provided by an embodiment of the present invention;

[0027] Figure 3 is a flow chart of step S104 in the method for a dishwasher provided by an embodiment of the present invention;

[0028] Figure 4 is a schematic diagram illustrating step S104 in the method for a dishwasher provided by an embodiment of the present invention;

[0029] Figure 5 is a further flow chart of step S104 in the method for a dishwasher provided by an embodiment of the present invention;

[0030] Figure 6 is a further flow chart of a method for a dishwasher provided by an embodiment of the present invention;

[0031] Figure 7 is a further flow chart of step S109 in the method for a dishwasher provided by an embodiment of the present invention;

[0032] Figure 8 is a connection topology diagram illustrating steps S106 to S109 in the method for a dishwasher provided by an embodiment of the present invention;

[0033] Description of Reference Numerals

[0034] 100. Dishwasher; 10. Image acquisition equipment;

[0035] 20. Fill light; 200. Cloud platform. DETAILED DESCRIPTION

[0036] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0037] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0038] It should be noted that all directional indications in the embodiments of the present invention (such as up, down, left, right, front, back, etc.) are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0039] In addition, the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the meaning of "or" appearing in the full text is that it includes three parallel solutions. Taking "A or B" as an example, it includes solution A, or solution B, or a solution that satisfies both A and B. In addition, the technical solutions between the various embodiments can be combined with each other, but it must be based on the ability of ordinary technicians in this field to implement. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0040] The embodiment of the present invention provides a fill light solution for a dishwasher. The solution includes multiple embodiments, including image acquisition of a washing area in the dishwasher, image selection, and controlling a fill light to fill light. The contents included in the embodiments of the present invention are described in detail below:

[0041] The embodiment of the present invention first provides a method for a dishwasher. Specifically, in order to make the dishwasher more intelligent, a method for supplementing light in the cleaning area of ​​the dishwasher to better obtain a target image is provided. The method aims to solve the problem that the existing dishwasher transmits image information of various light brightnesses back to the server, and uses the powerful computing power of the server to drive the detection and segmentation model to analyze various brightnesses. However, this will inevitably lead to an increase in network load, thereby increasing the pressure on the server, and ultimately the required R&D cost is high. It is difficult to achieve accurate recognition and the user's expected experience cannot be achieved.

[0042] See also Figure 1 , Figure 1: is a schematic diagram of the structure of the dishwasher provided in an embodiment of the present invention. The dishwasher 100 has a cleaning area for placing dishes in the area for cleaning. The inner wall of the cleaning area may be provided with an image acquisition device 10 for acquiring images of the cleaning area, and further provided with a fill light 20 for filling light for the cleaning area of ​​the dishwasher; optionally, the fill light 20 is provided in the form of a ring fill light on the periphery of the image acquisition device 10 to save space in the cleaning area and facilitate wiring on one side of the inner wall. The positions of the fill light 20 and the image acquisition device 10 may not be limited, and may be provided on the top or side of the cleaning area. The dishwasher provided in the embodiment of the present invention serves only as an example and does not limit the dishwasher required by the embodiment of the present invention.

[0043] See also Figure 2 , Figure 2 1 is a flow chart of a method for a dishwasher provided by an embodiment of the present invention. The method for a dishwasher comprises:

[0044] Step S101, controlling the fill lights to provide fill light at different brightnesses;

[0045] Step S102, controlling the image acquisition device to acquire images at different brightness levels;

[0046] Step S103, inputting the image into the prediction model to predict the image information of each image;

[0047] Step S104, comparing the images according to the image information to determine the target image;

[0048] Step S105: taking the brightness corresponding to the target image as the target fill light brightness of the fill light.

[0049] Specifically, in the method provided in the embodiment of the present invention, it is necessary to use a fill light with adjustable brightness. For example, the fill light can be set to control the brightness of the fill light through PWM (Pulse Width Modulation) to achieve the control of the fill light to fill light with different brightness values ​​in step S101. In one embodiment, the duty cycle (duty cycle: the proportion of the time that a signal or system is in a conducting state to the overall cycle is called duty cycle or power cycle) of the electrical signal input to the fill light by controlling the PWM. For step S102, it is based on the fill light adjusting the brightness value while collecting images, thereby collecting multiple images with different brightness values, each image corresponding to a brightness value. In simple terms, steps S101 to S102 are to fill light by continuously adjusting the brightness value of the fill light, so that the cleaning area is located in different light environments, and cooperate with the image acquisition device to collect images, so as to obtain all images under different brightness values.

[0050] For step S101 to step S102, there are two ways. One way is to initialize the fill light and start the image acquisition device to collect images each time the brightness of the fill light is changed. The brightness value of the fill light can be set to multiple levels, such as 10 levels. Each level corresponds to a pulse width modulation waveform with a different duty cycle. Level 10 is the brightest duty cycle, that is, the fill light has the highest brightness when the duty cycle is the largest. Conversely, level 1 is the minimum brightness of the fill light. After each initialization of the fill light, the brightness value of the fill light is controlled to fill the light in sequence from high to low or from low to high. After each adjustment, the image acquisition device is triggered to collect an image, thereby collecting 10 images with different brightness values ​​from 10 to 1. It can be understood that the brightness value of the fill light can also be divided into more levels to improve accuracy, so as to obtain more images, which is determined according to user needs.

[0051] In another embodiment, a cycle is set, that is, within a preset cycle starting from the initial time, the fill light and the image acquisition device interact to acquire images, such as 0 to 0.1s, the fill light adjusts the brightness value once, and the image acquisition device acquires an image at 0.1s (the exposure time is short and can be ignored), and the fill light adjusts the brightness value once from 0.1s to 0.2s, and the image acquisition device acquires an image again at 0.2s, so that the fill light adjusts all the brightness values ​​within one cycle, and the image acquisition device can obtain images of the cleaning area at all brightness values.

[0052] It should be noted that the functions implemented by the above two methods are consistent, but they are inconsistent in terms of coding thinking and control logic. Both methods are within the scope of protection of the embodiments of the present invention.

[0053] In one embodiment, controlling the fill light to perform fill light at different brightness values ​​in step S101 may include: controlling the brightness value of the fill light to perform fill light in sequence from high to low or from low to high according to the level. Specifically, it is preferred to control the brightness value of the fill light to perform fill light in sequence from high to low according to the level, so that the subsequent image acquisition device can find the most suitable brightness value more quickly, thereby accelerating the reaction speed, reducing the calculation time, and improving the response speed.

[0054] Specifically, according to the above steps S101 to S105, the embodiment of the present invention provides a general solution (executed by the controller):

[0055] A trigger is set at the door opening mechanism of the dishwasher. When the door of the dishwasher is closed, the controller will receive a signal, and then start to initialize the camera. According to the N (N is greater than 2) levels of the fill light, N photos are quickly taken in the above two ways. Here, one image represents a photo of one brightness level (i.e., the corresponding brightness value), so color images of N brightness levels are obtained.

[0056] Then, these N color images are input into the pre-trained detection or segmentation model, and N output results, i.e., image information, can also be obtained. The image information can include the content required by the dishwasher during the dishwashing process, such as the number of tableware, the cleanliness of the tableware, the type of tableware, the confidence level, etc. After comparing and analyzing these N results, the best result, i.e., the target image, is selected, and then the best brightness level is determined through the target image. Then, the brightness of the fill light is directly adjusted to this brightness, providing high-quality images for subsequent tasks, significantly improving recognition precision and accuracy. Compared with previous methods, this method can greatly reduce computing power, thereby reducing R&D costs and speeding up response time.

[0057] See also Figure 3 as well as Figure 4 , Figure 3 It is a flow chart of step S104 in the method for a dishwasher provided by an embodiment of the present invention. Figure 4 is a schematic diagram illustrating step S104 in the method for a dishwasher provided by an embodiment of the present invention. Determining the target image according to the number of dishes in step S104 includes:

[0058] Step S1041, sorting the multiple images according to the number of tableware;

[0059] Step S1042, determining the first image with the largest number of tableware from the images;

[0060] Step S1043: When the number of the first images is one, determine the first image as the target image.

[0061] It is understandable that in general, in an integrated deep learning strategy, a detection segmentation model is usually required to identify the features of an image. The detection segmentation model runs in a chip or server with graphics acceleration. It can process multiple images of a batch at one time, which also allows N images to be transferred at one time in the embodiment of the present application without affecting the entire running time. At the same time, due to the image quality that is too dark or too bright, the edges of the tableware cannot be distinguished, which leads to a decrease in the number of tableware detected by the model, that is, the recall rate (detection rate) is reduced. However, if the tableware is not recognized, it is likely to cause the tableware to be missed and not washed, resulting in a very poor user experience, and the effect is contrary to the original intention of the smart dishwasher. Therefore, when the dishwasher expected by the embodiment of the present invention performs intelligent cleaning, it is necessary to ensure high recall, that is, the number of tableware that has not been detected must be as low as possible. Then it is only necessary to determine that for these N images, the brightness of the highest number of detections is the required brightness. The dishwasher has higher requirements for the recall rate when cleaning. As long as the number of detections is the largest, the brightness value that is considered to be the best is selected, and this feature is used to achieve more accurate and faster response recognition. The number of tableware can be identified by edge detection technology in image recognition. Of course, a large number of tableware detected, that is, a high recall rate, usually leads to a decrease in accuracy, that is, the tableware detected by the model may also be the background. In the embodiment of the present invention, the model can be improved by continuously mining difficult examples.

[0062] In one embodiment, taking the number of tableware as an example, see Figure 5 , Figure 5 This is a further flow chart of step S104 in the method for a dishwasher provided by an embodiment of the present invention. The image information also includes confidence. The image information also includes confidence. In step S104, extracting the target image according to the number of dishes also includes:

[0063] Step S1044: when the number of first images is greater than one, sort the confidences of the first images;

[0064] Step S1045: determine the second image with the highest confidence in the first images as the target image.

[0065] It can be understood that when there is more than one second image with the largest number of tableware, such as in two images corresponding to brightness values ​​of level 7 and level 8 (at this time, the two images are both referred to as second images), the number of tableware is the largest number of 6 tableware in the same batch of images, the confidence (also called reliability, or confidence level, confidence coefficient, that is, when sampling to estimate the overall parameters, due to the randomness of the sample, the conclusion is always uncertain. In the embodiment of the present invention, it can be understood as the accuracy estimated by the prediction model) of the two images, and the second image with higher confidence can be used as the target image.

[0066] See also Figure 6 , Figure 6 is a further flow chart of the method for a dishwasher provided in an embodiment of the present invention. In the embodiment of the present invention, the method further includes:

[0067] Step S106, controlling the operation of the dishwasher according to the image information of the target image and obtaining the operation result;

[0068] Step S107, determining whether the image information matches according to the operation result;

[0069] Step S108: When the image information does not match, the target image is used as a training sample to train the prediction model to optimize the prediction model.

[0070] Specifically, after the image information in the target image obtained in step S105 (at least one of the number of tableware, cleanliness, type of tableware, material of tableware, and position of tableware) is obtained, the control program of the dishwasher is determined, that is, the relevant parameters of the control program of the dishwasher are changed in a targeted manner according to the image information, such as adjusting the water pressure and cleaning time according to the number of tableware. The more tableware there are, the stronger the water pressure and the longer the cleaning time. After cleaning is completed according to the above control program, the effect after cleaning is judged to see whether it matches the expected result. When the operating result does not match the expected effect, that is, the image information is inaccurate, the target image is what is called a difficult sample in this field, and can be used as a training sample to train the prediction model to optimize the prediction model, thereby continuously optimizing the prediction model to make future recognition more accurate. Among them, in the embodiment of the present invention, whether the operation result matches the expected result can be determined by obtaining another image immediately after the cleaning is completed, and by performing image recognition on the image, that is, by comparing the similarity between the image after the cleaning is completed and the original target image, it is determined whether the tableware has been washed clean. If the similarity is higher than a preset threshold, that is, the similarity between the images before and after cleaning is very large, the tableware is determined to be not clean, that is, the image information does not match.

[0071] It is also possible to set up a feedback system in the dishwasher to provide manual feedback when the food is not cleaned thoroughly, thereby determining whether the operating results match the expected results.

[0072] See also Figure 7 as well as Figure 8 , Figure 7 It is a further flow chart of the method for a dishwasher provided by an embodiment of the present invention. Figure 8 is a connection topology diagram illustrating steps S106 to S109 in the method for a dishwasher provided by an embodiment of the present invention. In step S108, training the prediction model using the target image as a training sample to optimize the prediction model includes:

[0073] Step S1081: Send the target image to the cloud platform, and the cloud platform uses the target image as a training sample to train the prediction model;

[0074] Step S1082: download the trained prediction model to the dishwasher to optimize the prediction model.

[0075] Specifically, an embodiment of the present invention provides a method for training a prediction model using a target image as a training sample to optimize the prediction model, that is, optimization is performed through the cloud platform 200. First, a difficult sample set is established on the cloud platform 200, and the target image with mismatched operation results is uploaded to the cloud platform 200 via the Internet as a difficult sample. The prediction model is trained using the target image and a vector machine or a classifier. After optimization based on the training results, it is then sent to the dishwasher 100 via OTA (Over the Air Technology) to update and optimize the original model of the dishwasher 100, so that the user's dishwasher 100 is more perfect, the scalability and portability of the dishwasher 100 are increased, and the accuracy of subsequent recognition is increased.

[0076] Through the above method, the cloud platform 200 can be established to uniformly update users in various regions, thereby iterating the original prediction model to achieve the best usage effect for users. For the model, any detection model or instance segmentation model can be applied to the method provided in the embodiment of the present invention according to different tasks. On the other hand, it is also necessary to run the detection algorithm in the chip carried by the dishwasher. Considering the chip computing power and real-time requirements of the dishwasher, and the method provided in the embodiment of the present invention only estimates the optimal brightness, the accuracy of the detection is not so high. The above conditions make it unnecessary to design the model at the dishwasher end so complicated. Various streamlined one-stage anchor base models or fast ancho free models can be applied to the model deployment at the dishwasher end. The above model can also be further accelerated by using the optimization operator or by compressing, clipping or even quantizing and binarizing the model, so that the entire model can meet the requirements of fast and real-time. Then, according to the number of detected targets, the optimal judgment of the illumination is made in combination with the confidence. According to the above description, this method and strategy are applicable to various model optimization methods, such as depthwise separable convolution, grouped convolution, binary or ternary networks, int8 quantization, various pruning operations, model distillation and other methods.

[0077] In summary, the method provided by the embodiment of the present invention is different from the previous method of calculating the most reasonable fill light brightness through images. Instead, it first controls the fill light to perform fill light at different brightness values; controls the image acquisition device to acquire multiple images, each image corresponding to a brightness value; then compares the image information in the image to determine the target image, and uses the brightness value corresponding to the target image as the target fill light brightness value of the fill light, so that the most appropriate brightness value required by the dishwasher can be determined in real time and quickly. The embodiment of the present invention utilizes the idea of ​​integration, cleverly combining deep learning with this special scenario, and can achieve real-time and fast precise control of light brightness in the dishwasher, and can continuously iterate and optimize to save network bandwidth and reduce server pressure, bringing a better experience to users.

[0078] An embodiment of the present invention also provides an apparatus, including an image acquisition device for acquiring an image of a cleaning area of ​​a dishwasher; a fill light for filling light on the cleaning area; and a processor, wherein the processor is configured to execute some or all of the steps provided in the above method embodiment.

[0079] Those skilled in the art should also understand that if the method or device provided by the present invention is combined with the above-mentioned method after simple changes, or replaced on the device, such as the replacement of the model and material of each component, the replacement of the use environment, the simple replacement of the position relationship of each component, etc.; or the product constituted by it is set as an integral whole; or a detachable design; all the combined components can form a method / device / device with specific functions, and using such a method / device / device to replace the method and device of the present invention also falls within the protection scope of the present invention.

[0080] The present invention also provides a dishwasher, including the above-mentioned device. It should be understood that the dishwasher is not limited in size and shape, and only needs to utilize the corresponding elements of the device to achieve the same or similar functions, which should also fall within the scope of protection of the present invention.

[0081] The device further comprises a memory, and the method for a dishwasher can be stored in the memory as a program unit, and the processor executes the program unit stored in the memory to implement corresponding functions.

[0082] The processor includes a kernel, and the kernel calls the corresponding program unit from the memory. One or more kernels can be set, and the method for using the dishwasher can be used to fill in the light by adjusting the kernel parameters.

[0083] The memory may include non-permanent memory in a computer-readable medium, a random access memory (RAM) or a non-volatile memory in the form of a read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0084] An embodiment of the present invention further provides a machine-readable storage medium having a program stored thereon, and the program, when executed by a processor, implements the method for a dishwasher.

[0085] An embodiment of the present invention further provides a processor, which is used to run a program, wherein the method for a dishwasher is executed when the program is run.

[0086] An embodiment of the present invention further provides a computer program product, including a computer program, and when the computer program is executed by a processor, the computer program implements the above-mentioned method for a dishwasher.

[0087] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0088] The present application is described with reference to the flowcharts or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process or box in the flowchart or block diagram, as well as the combination of processes or boxes in the flowchart or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processor to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processor generate instructions for implementing the process in the flowchart or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0089] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processor to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0090] These computer program instructions can also be loaded onto a computer or other programmable data processor so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable device to implement the process. Figure 1 A process or multiple processes or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0091] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0092] The memory may include non-permanent memory in a computer-readable medium, in the form of random access memory (RAM) or non-volatile memory, such as read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0093] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0094] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0095] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for a dishwasher, the dishwasher comprising an image acquisition device and a fill light arranged in the dishwasher, characterized in that: The method comprises: Controlling the fill lights to provide fill light at different brightness; Controlling the image acquisition device to acquire images at different brightness levels; Inputting the images into a prediction model to predict image information of each image; Comparing the images according to the image information to determine a target image; Using the brightness corresponding to the target image as the target fill light brightness of the fill light; The image information includes at least one of the following: The quantity of cutlery, the cleanliness of the cutlery, the type of cutlery, and the confidence level; The image information is used to determine the target image corresponding to the target fill light brightness.

2. The method according to claim 1, characterized in that The image information includes the number of tableware, and comparing the images according to the image information to determine the target image includes: sorting the images according to the number of tableware; Determine a first image having the largest number of tableware from the images; When the number of the first images is one, the first image is determined as the target image.

3. The method according to claim 2, characterized in that The image information further includes confidence, and comparing the images according to the image information to determine the target image further includes: When the number of the first images is greater than one, sorting the confidences of the first images; The second image with the highest confidence in the first images is determined as the target image.

4. The method according to claim 1, characterized in that The controlling the fill light to perform fill light at different brightnesses comprises: The fill light is controlled to change brightness step by step.

5. The method according to claim 1, characterized in that The method further comprises: controlling the operation of the dishwasher and obtaining the operation result according to the image information of the target image; Determining whether the image information matches according to the operation result; In the case where the image information does not match, the prediction model is trained using the target image as a training sample to optimize the prediction model.

6. The method according to claim 5, characterized in that The training of the prediction model using the target image as a training sample to optimize the prediction model comprises: Sending the target image to the cloud platform, the cloud platform uses the target image as a training sample to train the prediction model; The trained prediction model is downloaded to the dishwasher to optimize the prediction model.

7. The method according to claim 1, characterized in that The prediction model includes any one of the following: Anchor-Based model and Anchor-Free model.

8. A device for a dishwasher, characterized in that include: A fill light, used for providing fill light inside the dishwasher; An image acquisition device, used to acquire images inside the dishwasher; The processor is configured to: Controlling the fill lights to provide fill light at different brightness; Controlling the image acquisition device to acquire images at different brightness levels; Inputting the images into a prediction model to predict image information of each image; Comparing the images according to the image information to determine a target image; Using the brightness corresponding to the target image as the target fill light brightness of the fill light; The image information includes at least one of the following: The quantity of cutlery, the cleanliness of the cutlery, the type of cutlery, and the confidence level The image information is used to determine the target image corresponding to the target fill light brightness.

9. The device according to claim 8, characterized in that The image information includes the number of dishes, and the processor is configured to compare the images according to the image information to determine the target image. The processor is configured to: sorting the images according to the number of tableware; Determine a first image having the largest number of tableware from the images; When the number of the first images is one, the first image is determined as the target image.

10. The device according to claim 9, characterized in that The image information further includes a confidence level, and the processor is configured to compare the images according to the image information to determine the target image. The processor is also configured to: When the number of the first images is greater than one, sorting the confidences of the first images; The second image with the highest confidence in the first images is determined as the target image.

11. The device according to claim 8, characterized in that The processor is configured to control the fill light to perform fill light at different brightnesses respectively, including the processor being configured to: The fill light is controlled to change brightness step by step.

12. The device according to claim 8, characterized in that The processor further comprises: controlling the operation of the dishwasher and obtaining the operation result according to the image information of the target image; Determining whether the image information matches according to the operation result; In the case where the image information does not match, the prediction model is trained using the target image as a training sample to optimize the prediction model.

13. The device according to claim 12, characterized in that The training of the prediction model using the target image as a training sample to optimize the prediction model comprises: Sending the target image to the cloud platform, the cloud platform uses the target image as a training sample to train the prediction model; The trained prediction model is downloaded to the dishwasher to optimize the prediction model.

14. The device according to claim 8, characterized in that The prediction model includes any one of the following: Anchor-Based model and Anchor-Free model.

15. A dishwasher, characterized in that: The method comprises a device for a dishwasher according to any one of claims 8 to 14.

16. A machine-readable storage medium, characterized in that: The machine-readable storage medium stores instructions, which are used to enable the processor to perform the method for a dishwasher according to any one of claims 1 to 7 when the instructions are executed by the processor.

Images