Variable partition autoencoder-based tooth color matching method, device, medium, and program

Abstract

The application discloses a tooth color matching method and device based on a variable division self-encoder, a medium and a program. The method comprises the following steps: segmenting maxillary tooth images and mandibular tooth images according to an intraoral scanner image of teeth; converting an RGB color space of a single tooth in the intraoral scanner image into an HSV color space and a Lab color space; merging the RGB color space, the HSV color space and the Lab color space to obtain color features of the single tooth; generating color information of all teeth in the maxilla and the mandible based on the color features of the single tooth by using a self-encoder model; and mapping the color information of all teeth in the maxilla and the mandible on the intraoral scanner image to generate a real tooth image. The tooth color matching method and device can obtain automatic, refined and accurate tooth color matching results.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, and in particular to a tooth color matching method, device, medium and program based on a variable division autoencoder. BACKGROUND

[0002] With the gradual improvement of living standards, people pay more and more attention to their own health problems, one of which is oral repair.

[0003] According to the statistics of global and Chinese denture sales from 2016 to 2020, the demand for oral repair in the domestic and foreign markets is increasing year by year, and China's market share of the world is also increasing. The 2020-2021 China consumer oral medical service type distribution table shows that residents' consumption of oral cosmetic services accounts for 43.56%, of which oral repair accounts for 31.29%. In addition, according to the 1998-2021 growth rate of the population aged 65 and above, the aging trend is deepening, accompanied by natural tooth loss, periodontal disease, tooth decay, and age-related tooth changes.

[0004] The State Office of the National Health Commission issued the Health Oral Action Plan (2019-2025) and clearly stated that the oral health level of the public should be improved, so oral repair has its mission and necessity. Color matching is an important part of the oral repair process. The advocacy of "dental aesthetics" requires that the repaired dentures not only be firm, suitable, and wear-resistant, but also match the overall aesthetic harmony of the patient's oral environment. Digitalization of the oral repair process is an effective means to improve the efficiency and effectiveness of oral repair.

[0005] With the development of computer technology, artificial intelligence is widely combined with other fields. The color matching model based on artificial intelligence algorithm is an application example of artificial intelligence technology in the field of oral repair. Existing models take the color number of the color chart as the learning goal to realize the classification and decision of the patient's tooth color. Although the existing tooth color matching model based on artificial intelligence technology has to some extent freed the artificial intervention in the color matching process.

[0006] In the process of implementing the technical solutions of the present application, the present application inventors have found at least the following technical problems in the prior art:

[0007] The current oral prosthetic process is not fully digitalized, and the specific manifestation is that the oral color matching link needs human participation. In the traditional color matching link, the clinician uses a color chart or a colorimeter to obtain a tooth color number, and then transmits the color number information and tooth shape information to the technician, and finally completes the production of the denture. The visual color matching method based on the color chart is easily affected by the lighting conditions, the color perception of different clinicians, visual fatigue, and color matching experience, resulting in deviation in the final color matching result. In addition, according to related research, the color of human teeth can be subdivided into thousands of colors, and the mainstream 26-color and 16-color color charts on the market have a relatively coarse granularity in color division, resulting in insufficient refinement of the color matching result. The color matching method based on the colorimeter has the problems of low color matching efficiency, poor accuracy, and low consistency of the color matching results obtained by different colorimeters.

[0008] Although the tooth color matching model based on artificial intelligence technology has to some extent eliminated human intervention in the color matching link, the problem of insufficient refinement of the color matching result still exists because the decision result of the model is the color number of the color chart.

[0009] In summary, the existing tooth color matching method has the technical problems of low automation degree, insufficient refinement of the color matching result, and poor accuracy. SUMMARY

[0010] The embodiments of the present application provide a tooth color matching method, device, medium and program based on a variable division autoencoder, which solves the technical problems of low automation degree, insufficient refinement of the color matching result, and poor accuracy of the existing tooth color matching method.

[0011] The embodiments of the present application provide a tooth color matching method based on a variable division autoencoder, which comprises: segmenting the upper jaw tooth image and the lower jaw tooth image according to the oral scanner image of the tooth; converting the RGB color space of a single tooth in the oral scanner image into the HSV color space and the Lab color space; merging the RGB color space, the HSV color space and the Lab color space to obtain the color feature of the single tooth; generating the color information of all teeth in the upper jaw and the lower jaw based on the color feature of the single tooth; and mapping the color information of all teeth in the upper jaw and the lower jaw on the oral scanner image to generate a real tooth image.

[0012] Optionally, the segmentation of the upper jaw tooth image and the lower jaw tooth image according to the oral scanner image of the tooth specifically comprises: obtaining annotation information facing the upper jaw and lower jaw tooth segmentation annotation of the oral scanner image; detecting the tooth, and obtaining the upper jaw tooth image and the lower jaw tooth image based on the tooth detection region.

[0013] Optionally, the converting the RGB color space of the single tooth in the intraoral scanner image into the HSV color space and the Lab color space specifically comprises: converting the RGB color space of the single tooth in the intraoral scanner image into the HSV color space; and converting the RGB color space of the single tooth in the intraoral scanner image into the Lab color space.

[0014] Optionally, the merging the RGB color space, the HSV color space and the Lab color space to obtain the color feature of the single tooth specifically comprises: superimposing the RGB color space, the HSV color space and the Lab color space in the channel dimension to obtain a feature aggregation image; and performing color representation on the feature aggregation image to obtain the color feature of the single tooth.

[0015] Optionally, the generating the color information of all teeth in the upper jaw and the lower jaw based on the color feature of the single tooth specifically comprises: encoding the color feature of the single tooth to obtain a high-level color feature of the single tooth; and generating the color information of all teeth in the upper jaw and the lower jaw based on the high-level color feature.

[0016] Optionally, before the separating the upper jaw tooth image and the lower jaw tooth image from the intraoral scanner image of the teeth, the method further comprises: generating a variable division autoencoder-based tooth color matching model based on the intraoral scanner image of the teeth and a digital photography image.

[0017] Optionally, after the mapping the color information of all teeth in the upper jaw and the lower jaw on the intraoral scanner image to generate a tooth real image, the method further comprises: manufacturing a denture based on the tooth real image.

[0018] In another aspect, the embodiments of the present application further provide a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the variable division autoencoder-based tooth color matching method when executing the computer program.

[0019] The embodiments of the present application further provide a computer readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the variable division autoencoder-based tooth color matching method when executed by a processor.

[0020] The embodiments of the present application further provide a computer program product comprising a computer program, wherein the computer program implements the steps of the variable division autoencoder-based tooth color matching method when executed by a processor.

[0021] The one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0022] A tooth color matching method based on a variable division autoencoder, the method comprising: segmenting a maxillary tooth image and a mandibular tooth image from an intraoral scanner image of a tooth; converting an RGB color space of a single tooth in the intraoral scanner image into an HSV color space and a Lab color space; merging the RGB color space, the HSV color space and the Lab color space to obtain color features of the single tooth; generating color information of all teeth in the maxilla and the mandible based on the color features of the single tooth; and mapping the color information of all teeth in the maxilla and the mandible to the intraoral scanner image to generate a real tooth image. The color matching process does not require human intervention, the entire process is fully digitized, the automation level is high, the color matching efficiency is improved, and on the basis of a tooth scanning image, a real tooth image generated by combining real color information of the tooth is a fine-grained and refined tooth color matching result, which can greatly improve the color matching accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 A flowchart of a tooth color matching method based on a variable division autoencoder in an embodiment of the present application;

[0024] Figure 2 A method flowchart of model training and model inference in an embodiment of the present application. DETAILED DESCRIPTION

[0025] Embodiments of the present application provide a tooth color matching method, device, medium and program based on a variable division autoencoder, which solve the technical problems of low automation, insufficient refinement and poor accuracy of existing tooth color matching methods.

[0026] The technical scheme of an embodiment of the present application is to solve the above problems, and the general idea is as follows:

[0027] A tooth color matching method based on a variable division autoencoder, the method comprising: segmenting a maxillary tooth image and a mandibular tooth image from an intraoral scanner image of a tooth; converting an RGB color space of a single tooth in the intraoral scanner image into an HSV color space and a Lab color space; merging the RGB color space, the HSV color space and the Lab color space to obtain color features of the single tooth; generating color information of all teeth in the maxilla and the mandible based on the color features of the single tooth; and mapping the color information of all teeth in the maxilla and the mandible to the intraoral scanner image to generate a real tooth image. The color matching process does not require human intervention, the entire process is fully digitized, the automation level is high, the color matching efficiency is improved; on the basis of a tooth scanning image, a real tooth image generated by combining real color information of the tooth is a fine-grained and refined tooth color matching result, which can greatly improve the color matching accuracy.

[0028] For better understanding of the above technical solutions, the above technical solutions will be described in detail below in combination with the drawings of the specification and specific embodiments. Obviously, the embodiments described in the present application are part of the embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor belong to the scope of protection of the present application.

[0029] Please refer to Figure 1 , a tooth color matching method based on a variable division autoencoder in an embodiment of the present application is described in detail.

[0030] Step 101: according to the dental intraoral scanner image, the maxillary tooth image and the mandibular tooth image are segmented out;

[0031] Step 102: convert the RGB color space of a single tooth in the intraoral scanner image into HSV color space and Lab color space;

[0032] Step 103: merge the RGB color space, the HSV color space and the Lab color space to obtain the color feature of a single tooth;

[0033] Step 104: based on the color feature of a single tooth, generate the color information of all teeth in the maxilla and the mandible;

[0034] Step 105: map the color information of all teeth in the maxilla and the mandible to the intraoral scanner image to generate a real tooth image.

[0035] The above will be described in combination with the drawings Figure 2 , the process of the tooth color matching method based on the variable division autoencoder in the embodiment of the present application is described in detail.

[0036] The training process of the tooth color matching model based on the variable division autoencoder includes the following steps:

[0037] S1, maxillary and mandibular tooth segmentation module. The module takes the oral scanner image and the digital photography image as input. First, the oral scanner image and the digital photography image are both labeled with maxillary and mandibular single tooth, which are used to train the maxillary and mandibular tooth segmentation model for the oral scanner image and the digital photography image. Then, the target segmentation model based on convolutional neural network is constructed to complete the maxillary and mandibular tooth segmentation model training. The segmentation results of the oral scanner image and the digital photography image are mapped by the two-part image maximum matching algorithm to calculate the formula: mapping_list=BiMaxMatch(x1,x2), where x1 represents the segmentation result of the oral scanner image, and x2 represents the segmentation result of the digital photography image. The segmentation result of the oral scanner is used as the input of the autoencoder, and the segmentation result of the digital photography image is used as the real label that the model needs to learn. Finally, based on the trained target segmentation model, the maxillary and mandibular tooth segmentation function of the oral scanner image and the digital photography image is realized.

[0038] S2, tooth color space conversion module. The module realizes the information conversion from the original image RGB color space to the HSV color space and the Lab color space for the color information of the single tooth in the oral scanner image.

[0039] S3, tooth color feature generation module. The module combines three kinds of color space information to obtain the final single tooth color feature expression, so that the single tooth color information in the oral scanner image is more abundant, and the generalization of the subsequent autoencoder model is improved.

[0040] S4, autoencoder module. The module is divided into an encoder submodule and a decoder submodule. The encoder submodule is used to encode the combined tooth color features to obtain the high-level features of the single tooth color information of the oral scanner image. The decoder generates the color information of all teeth in the maxilla and mandible based on the high-level color features, with all tooth color information segmented from the digital photography image as the learning target.

[0041] Further, in step S1, the maxillary and mandibular tooth segmentation model for the oral scanner image and the digital photography image includes the steps of maxillary and mandibular tooth segmentation labeling for the oral scanner image and the digital photography image, maxillary and mandibular tooth segmentation model construction and training based on convolutional neural network, etc. Step S1 specifically includes:

[0042] S11, maxillary and mandibular tooth segmentation annotation for oral scanning instrument image and digital photography image, the oral scanning instrument image and the digital photography image are in RGB mode and the size is 3*512*512, wherein 3 represents three channels of red, green and blue. When labeling data, labelme labeling software is used, tooth information is surrounded by a polygon, and then a corresponding single-channel binary mask is generated by a post-processing script. The labeled information with the size of 512*512 is obtained, wherein 0 pixels represent the background and 1 pixels represent the teeth.

[0043] S12, maxillary and mandibular tooth segmentation model construction and training based on convolutional neural network, the segmentation model is divided into two stages. The first stage is tooth detection, and the target detection algorithm yolov3 is used to detect the tooth, and then the second stage is entered. The feature map corresponding to the detected region is cut out, the FCN structure is used for deconvolution upsampling, and finally the tooth information output result of 512*512 is obtained. In particular, the target detection loss function and the segmentation loss function use IoULoss, which calculates the intersection over union between the model prediction result and the real artificial label and performs gradient back propagation on the result; the classification loss function uses BinaryEntropyLoss.

[0044] Specifically, the IoULoss loss function formula is Wherein A represents the output result box of the detection model, and B represents the real box of the artificial annotation. The BinaryEntropLoss loss function formula is Wherein y1 represents the real classification label, and y2 represents the class result predicted by the model.

[0045] Further, in step S2, two steps are included, i.e. conversion of oral scanning instrument image single tooth RGB color space to HSV color space and conversion of oral scanning instrument image single tooth RGB color space to Lab color space. Step S2 specifically includes:

[0046] S21, conversion of oral scanning instrument image single tooth RGB color space to HSV color space, from the perspective of hue, saturation and brightness to represent the RGB color space image. The hue H is from the perspective of the value range of red 0 degrees, yellow 60 degrees, green 120 degrees, cyan 180 degrees, blue 240 degrees and purple 300 degrees, the value range of saturation S is 0 to 1, the higher the S value, the deeper and brighter the color, the value of brightness V is from 0 (representing black) to 1 (representing white), the larger the value of V, the brighter the color.

[0047] S22, the single tooth RGB color space of the oral scanner image is converted to Lab color space. Lab uses digital information to describe human visual perception, where L represents brightness, ranging from 0 to 100; a represents the change from red to green, ranging from -128 to 127; and b represents the change from yellow to blue, ranging from -128 to 127.

[0048] Further, in step S3, in the tooth color feature generation module, specifically comprising:

[0049] S31, the resolution of the RGB image is 3x512x512, and after conversion in the S2 module, the resolution of the HSV color space and the Lab color space is 3x512x512. The RGB, HSV and Lab are stacked in the channel dimension to obtain the final output 9x512x512 of S31.

[0050] S32, the color feature of the feature aggregation image obtained in S31 is characterized, and the specific structure is average pooling, obtaining a color feature output of 32x32x32.

[0051] Further, in step S4, the construction and training of the autoencoder model include two steps, specifically comprising:

[0052] S41, autoencoder model construction. The autoencoder is composed of an encoder and a decoder, where the encoder has three fully connected layers and the decoder has three fully connected layers. More specifically, the first layer of the encoder has 2048 parameters, the second layer has 1024 parameters, and the third layer has 512 parameters. The first layer of the decoder has 1024 parameters, the second layer has 2048 parameters, and the third layer has 32768 parameters. Each layer of neurons has a 0.4 probability of random inactivation, and each layer has a LeakyReLU activation function.

[0053] S42, autoencoder model training. In S32, the color feature of the oral scanner image is input into the encoder, and before input, it is flattened into a 32768-dimensional vector, and the output of the decoder is a 32768-dimensional vector. The loss function is calculated by calculating the difference between the output of the decoder and the color feature of the digital photography image. The loss function is selected as the mean square loss function, and the formula is where x i is the model prediction value, x j is the true value of the digital photography image, and N is 32768.

[0054] The inference process of the tooth color matching model based on the variable division autoencoder includes the following steps:

[0055] S1, maxillary and mandibular tooth segmentation module. Unlike the training stage, the maxillary and mandibular tooth segmentation module of the model inference stage only takes the mouth scanner image as input, uses the trained convolutional neural network-based target segmentation model to realize the maxillary and mandibular tooth segmentation function of the mouth scanner image. The RGB image position area corresponding to the tooth segmentation result is cut out, the center coordinate position (xcenter, ycenter) is recorded, and the cut-out picture is input into the subsequent module.

[0056] S2-S3, the use method of the tooth color space conversion module and the tooth color feature generation module is consistent with the training stage.

[0057] S4, the autoencoder module. The autoencoder module of the model inference stage takes the merged mouth scanner tooth color feature as input, uses the trained autoencoder model to generate the color information of all teeth in the maxilla and mandible.

[0058] S5, maxillary and mandibular tooth image generation module. The module maps the color information of all teeth in the maxilla and mandible generated by the autoencoder module to the corresponding teeth in the mouth scanner image through the center coordinate information (xcenter, ycenter), and uses the tooth shape information obtained by the mouth scanner as the real tooth image.

[0059] S6, the tooth real image is sent to the technician, and the final denture is completed.

[0060] The positive progress effect of the present application is that:

[0061] 1. The missing tooth color can be generated.

[0062] 2. The whole process of oral color matching is digitized.

[0063] 3. The pixel-level oral tooth color matching result is generated.

[0064] Another embodiment of the present application provides a computer device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the steps of the tooth color matching method based on the variable partition autoencoder when executing the computer program.

[0065] Another embodiment of the present application provides a computer readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the steps of the tooth color matching method based on the variable partition autoencoder.

[0066] Another embodiment of the present application provides a computer program product comprising a computer program, the computer program being executed by a processor to implement the steps of the tooth color matching method based on the variable partition autoencoder.

[0067] The various changes and specific examples in the method in the foregoing embodiments are also applicable to the computer device, computer readable storage medium and computer program product of the present embodiment. Through the foregoing detailed description of the tooth color matching method based on the variable division self-encoder, those skilled in the art can clearly understand the implementation method of the computer device, computer readable storage medium and computer program product in the present embodiment. Therefore, in order to make the description brief, it will not be described in detail here.

[0068] The one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0069] A tooth color matching method based on a variable division self-encoder, the method comprising: segmenting maxillary tooth images and mandibular tooth images from an intraoral scanner image of teeth; converting the RGB color space of a single tooth in the intraoral scanner image into the HSV color space and the Lab color space; merging the RGB color space, the HSV color space and the Lab color space to obtain color features of the single tooth; generating color information of all teeth in the maxilla and the mandible based on the color features of the single tooth; and mapping the color information of all teeth in the maxilla and the mandible on the intraoral scanner image to generate a real tooth image. The color matching process does not require human intervention, the entire process is fully digitized, the automation level is high, and the color matching efficiency is improved; on the basis of the tooth scanning image, the real tooth image generated by combining the real color information of the teeth is a fine-grained and refined tooth color matching result, which can greatly improve the color matching accuracy.

[0070] Those skilled in the art will appreciate that embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage, etc.) containing computer-usable program code.

[0071] The present application is described with reference to flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as combinations of flows and / or blocks in the flowcharts and / or block diagrams, 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 processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one flow or multiple flows and / or blocks Figure 1 The functions specified in one flow or multiple flows and / or blocks

[0072] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the Figure 1 function specified in the flow or flows and / or blocks Figure 1 of the block or blocks.

[0073] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions that are executed on the computer or other programmable apparatus provide steps for implementing the Figure 1 function specified in the flow or flows and / or blocks Figure 1 of the block or blocks.

[0074] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

Claims

1. A tooth colorimetric method based on a variable divisible autoencoder, characterized in that, The method includes: Based on the intraoral scanning images of the teeth, the images of the maxillary teeth and mandibular teeth are segmented. The RGB color space of a single tooth in the oral scanner image is converted to HSV color space and Lab color space; By merging the RGB color space, the HSV color space, and the Lab color space, the color characteristics of each tooth are obtained; Based on the color features of each tooth, the autoencoder model generates color information for all teeth in the maxilla and mandible, specifically including: encoding the color features of each tooth to obtain high-level color features for each tooth; and generating color information for all teeth in the maxilla and mandible based on the high-level color features. The color information of all teeth in the maxilla and mandible is mapped onto the oral scanner image to generate a real image of the teeth. The construction and training of the autoencoder model are as follows: Autoencoder model construction: The autoencoder consists of two parts: an encoder and a decoder. The encoder has three fully connected layers, and the decoder also has three fully connected layers. More specifically, the encoder has 2048 parameters in the first fully connected layer, 1024 parameters in the second layer, and 512 parameters in the third layer; the decoder has 1024 parameters in the first layer, 2048 parameters in the second layer, and 32768 parameters in the third layer. Each neuron in each layer has a probability of random inactivation of 0.4, and each layer uses a LeakyReLU activation function. Autoencoder model training: The color features of the obtained intraocular scanner image are input into the encoder. Before input, the features are flattened into a 32768-dimensional vector, resulting in a 32768-dimensional vector as the decoder output. The loss function is calculated by determining the difference between the decoder output and the color features of the digital photographic image. The mean squared loss function is chosen, and the formula is as follows: ,in These are the model's predicted values. N represents the true value of the digital photographic image, and the number of N is 32768.

2. The method as described in claim 1, characterized in that, The step of segmenting the maxillary and mandibular tooth images based on the intraoral scanning images specifically includes: The maxillary and mandibular teeth of the oral scanner image are segmented and annotated to obtain annotation information; The teeth are detected, and images of the maxillary teeth and the mandibular teeth are obtained based on the detected tooth area.

3. The method as described in claim 1, characterized in that, The step of converting the RGB color space of a single tooth in the intraoral scanner image to the HSV color space and Lab color space specifically includes: Convert the RGB color space of a single tooth in the oral scanner image to the HSV color space; The RGB color space of a single tooth in the intraoral scanner image is converted to the Lab color space.

4. The method as described in claim 1, characterized in that, The process of merging the RGB color space, the HSV color space, and the Lab color space to obtain the color characteristics of a single tooth specifically includes: The RGB color space, the HSV color space, and the Lab color space are superimposed along the channel dimension to obtain a feature aggregation image; The color features of the individual tooth are obtained by performing color characterization on the aggregated feature image.

5. The method as described in claim 1, characterized in that, Before segmenting the maxillary and mandibular tooth images based on the intraoral scanning images of the teeth, the method further includes: Based on intraoral scan images and digital photographic images of teeth, a tooth colorimetric model based on a variable fractional autoencoder is generated.

6. The method as described in claim 1, characterized in that, After mapping the color information of all teeth in the maxilla and mandible onto the intraoral scanner image to generate a realistic image of the teeth, the method further includes: Based on the actual image of the teeth, a denture is fabricated.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-6.

9. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-6.

Images