A recipe input system for an intelligent cooking machine

Abstract

The present invention provides a recipe entry system for an intelligent cooking machine, comprising: a recipe collection module for obtaining recipe information; a recipe analysis module for analyzing the collected recipe information, extracting various attribute information in the recipe and saving it as a formatted recipe; a collection management module for collecting the formatted recipe through a user account and synchronizing it to the cloud network; and a recipe execution module for importing the formatted recipe selected by the user into the intelligent cooking machine for cooking work. Through the present invention, the intelligent cooking machine can independently obtain recipe information and generate a formatted recipe, enriching the types of recipes that the intelligent cooking machine can cook.

Description

Technical Field

[0001] The present invention relates to the technical field of automatic cooking, and in particular to a recipe input system of an intelligent cooking machine. Background Art

[0002] China's cooking culture is extensive and profound, with a wide variety of cooking methods. The main cooking techniques include stir-frying, stewing, frying, boiling, steaming, frying, and baking. The essence of various cooking methods is that after the cooked food is matched with different combinations and various expected heat states, various dishes with different tastes and flavor effects are obtained. However, cooking is a skill based on experience and manual work, and the technical requirements of the operator are very high. Nowadays, intelligent cooking machines have appeared. As long as the user puts the food ingredients into the cooking machine, the cooking machine can automatically complete the cooking process, so that the user can enjoy the food very conveniently. The working principle of the existing cooking machine is: put the food ingredients in the wok, when the cooking machine starts the cooking process, the control device controls the wok heating device to heat the wok, and the control device also controls the wok driving mechanism to flip the wok, so that the wok stirs the food ingredients inside it, and when the predetermined cooking process is completed, the cooking process can be completed.

[0003] For existing smart cooking machines, it is generally necessary to cook according to the formatted recipes pre-entered by the cooking machine company. However, due to the profound cooking culture in China and the various methods, many dishes do not have corresponding formatted recipes, so the smart cooking machine cannot cook the corresponding dishes, which limits the development of smart cooking machines. Summary of the invention

[0004] The present invention provides a recipe entry system for an intelligent cooking machine, which is used to enable the intelligent cooking machine to autonomously obtain recipe information and generate formatted recipes, thereby enriching the types of recipes that can be cooked by the intelligent cooking machine.

[0005] The present invention provides a recipe entry system for an intelligent cooking machine, comprising:

[0006] Recipe collection module, used to obtain recipe information;

[0007] The recipe analysis module is used to analyze the collected recipe information, extract various attribute information in the recipe and save it as a formatted recipe;

[0008] The collection management module is used to collect formatted recipes through user accounts and synchronize them to the cloud network;

[0009] The recipe execution module is used to import the formatted recipe selected by the user into the intelligent cooking machine for cooking.

[0010] Preferably, the recipe collection module includes:

[0011] A network search unit for searching for corresponding recipe information from the Internet through keywords set by the user;

[0012] A network collection unit for regularly obtaining updated recipe information from a preset website;

[0013] A recipe editing unit for sending an editing template of preset recipe information to the user and obtaining the edited template filled in by the user as recipe information.

[0014] Preferably, the recipe collection module further includes an identification collection unit, and the identification collection unit performs the following operations:

[0015] Obtain an operation video of the user during the process of manually stir-frying with a preset collection wok through a preset camera;

[0016] Perform action recognition and analysis on the operation video, determine multiple first action features included therein and classify them; wherein, the first action features include stir-frying, heat adjustment, addition of seasonings and ingredients, and taking the dish out of the pot;

[0017] Sort and label all the first action features in the entire stir-frying process corresponding to the operation video in time series to obtain a first stir-frying action sequence;

[0018] Determine the second action features of the user according to the pressure values collected by multiple pressure sensors preset on the collection wok and the deflection angle values of the wok surface collected by the gyroscope sensor, and determine the third action features of the user according to the gear signals collected by the fire power gear sensor preset on the collection wok; wherein, the second action features include flipping the wok, stir-frying, and taking the dish out of the pot;

[0019] Sort and label the second action features and the third action features analyzed by the sensors on the collection wok in time series to obtain a second stir-frying action sequence;

[0020] Align the first stir-frying action sequence with the second stir-frying action sequence, compare and determine the judgment nodes with different action features among them, send the operation video segment corresponding to the judgment node to the user and obtain the user's judgment result of the action features of the judgment node to correct the first stir-frying action sequence to obtain a third stir-frying action sequence;

[0021] Perform material recognition and analysis on the operation video, identify the types and quantities of the ingredients and seasonings added each time based on a pre-trained convolutional neural network, and use them as the material recognition result;

[0022] Use the material recognition result to expand the content corresponding to the seasoning and ingredient addition actions in the third stir-frying action sequence and generate recipe information.

[0023] Preferably, aligning the first stir-frying action sequence with the second stir-frying action sequence, comparing and determining the judgment nodes with different action characteristics therein, sending the operation video segment corresponding to the judgment node to the user and obtaining the user's judgment result of the action characteristics of the judgment node to correct the first stir-frying action sequence to obtain the third stir-frying action sequence includes:

[0024] Step S1: Align the first stir-frying action sequence with the second stir-frying action sequence, determine the action type judgment result and occurrence time point of each action characteristic, and determine each occurrence time point as a judgment node;

[0025] Step S2: Determine the action type judgment result of the action characteristic on any judgment node in the first stir-frying action sequence, determine the action type judgment result of the judgment node at the corresponding position in the second stir-frying action sequence, and compare whether the two action type judgment results are consistent;

[0026] Step S3: If they are consistent, do not change the action type judgment result on this judgment node in the first stir-frying action sequence;

[0027] Step S4: If they are inconsistent, intercept the operation video segment corresponding to this judgment node, send it to the user through the mobile terminal for manual recognition, and obtain the user's judgment result of the action characteristics of this judgment node to replace the action characteristics on this judgment node in the first stir-frying action sequence;

[0028] Step S5: Process all judgment nodes in the first stir-frying action sequence according to Steps 2 to 4, and use the processed first stir-frying action sequence as the third stir-frying action sequence.

[0029] Preferably, the recipe analysis module includes:

[0030] The first acquisition unit is used to acquire the first type of information in the recipe information; wherein, the first type of information includes the types of ingredients and the amounts of ingredients contained in the recipe information;

[0031] The second acquisition unit is used to acquire the second type of information in the recipe information: wherein, the second type of information includes the cooking method, total cooking duration, cooking duration of each cooking stage, heat intensity, and the ingredients, seasonings and water ratios that need to be added contained in the recipe information;

[0032] The third acquisition unit is used to acquire the third type of information in the recipe information; wherein, the third type of information includes the taste preference, nutritional content, and energy content of the dish announced in the recipe information;

[0033] A formatting output unit for filling a preset recipe template with the first type of information, the second type of information, and the third type of information to obtain a formatted recipe and output and save it.

[0034] Preferably, the collection management module includes:

[0035] A recipe recognition unit for receiving a formatted recipe and analyzing it to determine whether it lacks preset attribute information. If it lacks preset attribute information, it determines that the formatted recipe is invalid and abandons the formatted recipe.

[0036] A recipe processing unit for performing pre-processing before collection on a formatted recipe that does not lack preset attribute information and establishing an index path.

[0037] A recipe storage unit for collecting and storing the formatted recipe in the intelligent cooking machine, and synchronizing multiple formatted recipes stored therein to the account corresponding to the intelligent cooking machine in the cloud platform when the intelligent cooking machine is connected to the Internet.

[0038] Preferably, the recipe processing unit performs the following operations:

[0039] Extract the attribute information in multiple preset positions on the formatted recipe in sequence to obtain an attribute information sequence.

[0040] For any attribute information in the attribute information sequence, determine the position where the attribute information is arranged in the attribute information sequence, and extract the preset attribute keyword library for this position.

[0041] Use the attribute keyword library to match the attribute information to determine whether there are the same keywords.

[0042] If there are the same keywords, establish a corresponding mapping relationship between the formatted recipe and the keyword.

[0043] If there are no same keywords, create a new keyword in the attribute keyword library using the attribute information, and establish a mapping relationship between the formatted recipe and the new keyword.

[0044] Preferably, the recipe execution module includes:

[0045] A recipe selection unit for finding a corresponding formatted recipe according to the dish name selected by the user.

[0046] A recipe deconstruction unit for deconstructing the selected formatted recipe to determine the multiple cooking steps corresponding to the recipe.

[0047] A recipe execution unit is used to start the intelligent cooking machine to execute cooking work according to a formatted recipe, and remind the user to add ingredients and seasonings according to the content of the formatted recipe.

[0048] Preferably, it further includes a diet health management module, and the diet health management module performs the following operations:

[0049] Determine the dishes and the portion sizes of the dishes cooked by the user through the intelligent cooking machine within a preset time length recently;

[0050] Determine the formatted recipes corresponding to each dish, and determine the nutrient content and energy content in each dish according to the formatted recipes;

[0051] Calculate the total amount of nutrients and the total amount of energy ingested by the user based on the nutrient content and energy content in each dish according to the dishes and the portion sizes of the dishes cooked by the user through the intelligent cooking machine within a preset time length recently;

[0052] Compare and calculate the total amount of nutrients and the total amount of energy ingested by the user with a preset nutrient balance intake table to determine the nutrients and energy lacking by the user and the nutrients and energy excessively ingested by the user;

[0053] Select the first type of formatted recipes for supplementing the corresponding nutrients and energy according to the nutrients and energy lacking by the user;

[0054] Select the second type of formatted recipes for reducing the corresponding nutrients and energy according to the nutrients and energy excessively ingested by the user;

[0055] Comprehensively screen out the third type of formatted recipes according to the first type of formatted recipes and the second type of formatted recipes to participate in the user's balanced and healthy diet plan.

[0056] Preferably, it further includes a seasonal recipe recommendation module, and the seasonal recipe recommendation module is used to determine the production seasons of the main ingredients in all formatted recipes, and recommend the formatted recipes corresponding to various ingredients produced in the current season to the user.

[0057] Next, through the drawings and embodiments, the technical solutions of the present invention will be further described in detail. Description of the Drawings

[0058] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0059] Figure 1 It is a schematic structural diagram of a recipe input system of an intelligent cooking machine in an embodiment of the present invention;

[0060] Figure 2 It is a schematic structural diagram of the recipe collection module in the embodiment of the present invention;

[0061] Figure 3 It is a flowchart of the steps for obtaining the third stir-frying action sequence in the detection in the embodiment of the present invention. Detailed implementation manners

[0062] The following describes the preferred embodiments of the present invention with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only for the purpose of illustrating and explaining the present invention, and are not intended to limit the present invention.

[0063] The embodiment of the present invention provides a recipe entry system for an intelligent cooking machine provided by the present invention, including:

[0064] A recipe collection module 1, configured to obtain recipe information;

[0065] A recipe analysis module 2, configured to analyze the collected recipe information, extract various attribute information in the recipe and save it as a formatted recipe;

[0066] A collection management module 3, configured to collect the formatted recipe through a user account and synchronize it to the cloud network;

[0067] A recipe execution module 4, configured to import the formatted recipe selected by the user into the intelligent cooking machine for cooking work.

[0068] The working principle and beneficial effects of the above technical solution are as follows: The recipe collection module 1 searches for corresponding recipe information from the preset relevant websites on the Internet. After collecting the recipe information, the recipe analysis module 2 analyzes the recipe information, extracts the attribute information in the recipe information through keywords, including ingredient requirements, cooking duration, amount of seasoning added, cooking duration of each step, and adjustment of the fire power size of each cooking step, and saves the various attribute information as a formatted recipe; the collection management module 3 collects the formatted recipe through a user account and synchronizes it to the cloud network, so that users can make dishes that meet their own tastes by logging in to their accounts no matter where they go or what restaurant they are in; the recipe execution module 4 can read the collected formatted recipe and cook strictly according to the formatted recipe. Thus, the intelligent cooking machine can independently obtain and update the internal recipes, enriching the types of recipes that the intelligent cooking machine can cook.

[0069] In a preferred embodiment, the recipe collection module includes:

[0070] A network search unit 11, configured to search for corresponding recipe information from the Internet through the keyword vocabulary set by the user; the keyword vocabulary is content such as recipe name and main ingredients;

[0071] A network collection unit 12 is used to regularly obtain updated recipe information by fetching from a preset website address;

[0072] A recipe editing unit 13 is used to send an editing template of preset recipe information to a user and obtain the filled - in editing template by the user as recipe information.

[0073] The working principle and beneficial effects of the above - mentioned technical solution are as follows: Through the network search unit 11, a user can independently set keyword vocabulary to search for corresponding recipe information on the network for active collection. The content of the keyword vocabulary includes recipe names, main ingredients, main nutritional components, seasonal vegetables, etc., realizing the active search and collection of recipes according to the user's needs; Through the network collection unit 12, updated recipe information is regularly obtained and automatically collected according to a preset website address, enabling users to enjoy new types of dishes; When a user needs to write a recipe by themselves, the recipe editing unit 13 sends an editing template of preset recipe information to the user for filling, and obtains the filled - in editing template by the user as recipe information for collection, thus realizing the modification and creation of recipe information.

[0074] In a preferred embodiment, the recipe collection module further includes an identification collection unit, and the identification collection unit performs the following operations:

[0075] Obtain an operation video of the user during the process of manually stir - frying with a preset collection wok through a preset camera;

[0076] Determine the second action characteristics of the user according to the pressure values collected by multiple pressure sensors preset on the collection wok and the deflection angle value of the wok surface collected by a gyroscope sensor, and determine the third action characteristics of the user according to the gear signals collected by a fire power gear sensor preset on the collection wok; wherein, the second action characteristics include tossing the wok, stir - frying, and taking the food out of the pan;

[0077] Sort and label all the first action characteristics in the entire stir - frying process corresponding to the operation video in a time series to obtain a first stir - frying action sequence;

[0078] Determine the second action characteristics of the user according to the pressure values collected by multiple pressure sensors preset on the collection wok, and determine the third action characteristics of the user according to the gear signals collected by a fire power gear sensor preset on the collection wok; wherein, the second action characteristics include tossing the wok, stir - frying, and taking the food out of the pan;

[0079] Sort and label the second action characteristics and the third action characteristics analyzed by the sensors on the collection wok in a time series to obtain a second stir - frying action sequence;

[0080] Align the first stir-frying action sequence with the second stir-frying action sequence, compare and determine the judgment nodes with different action characteristics therein, send the operation video segment corresponding to the judgment node to the user and obtain the user's judgment result of the action characteristics of the judgment node to correct the first stir-frying action sequence to obtain a third stir-frying action sequence;

[0081] Perform material recognition and analysis on the operation video, identify the types and quantities of ingredients and seasonings added each time based on a pre-trained convolutional neural network, and use them as the material recognition results;

[0082] Use the material recognition results to expand the content corresponding to the seasoning and ingredient addition actions in the third stir-frying action sequence and generate recipe information.

[0083] The working principle and beneficial effects of the above technical solution are as follows: Through the recognition and collection unit, a preset camera is used to obtain the operation video of the user during manual stir-frying; the operation video is analyzed for action recognition to determine multiple first action features contained therein and classify them, thereby realizing the recognition of actions. Among them, the first action features include stir-frying, heat adjustment, adding seasonings and ingredients, and taking the cooked dish out of the pot. All the first action features in the entire stir-frying process corresponding to the operation video are sorted and marked in the time series to obtain the first stir-frying action sequence. Through the first stir-frying action sequence, the entire stir-frying action process of manual stir-frying in the video can be determined. Subsequently, the second action features of the user are determined according to the pressure values collected by multiple pressure sensors preset on the collection wok. For example, by combining the collection results of the pressure sensor preset on the handle of the collection wok and the gyroscope sensor, it is judged whether the user has an action of taking the cooked dish out of the pot. When the pressure value on the handle is greater than the preset value and the gyroscope detects that the deflection angle of the wok surface is greater than the preset value, it is judged that the user has an action of taking the cooked dish out of the pot. Or when the pressure value on the handle repeatedly exceeds the preset value for a preset number of times, it is judged that the user has a tossing action. Another example is to judge whether the user has been stir-frying with a spatula through the pressure vibration sensor preset on the collection wok, and determine the third action features of the user according to the gear signal collected by the fire power gear sensor preset on the collection wok. Among them, the second action features include tossing, stir-frying, and taking the cooked dish out of the pot. The second action features and the third action features analyzed by the sensors on the intelligent stir-frying machine are sorted and marked in the time series to obtain the second stir-frying action sequence. Through the second stir-frying action sequence, the entire stir-frying process that can be sensed on the intelligent stir-frying machine during manual stir-frying can be determined. The first stir-frying action sequence and the second stir-frying action sequence are aligned, and the judgment nodes with different action features are compared and determined. The operation video segment corresponding to the judgment node is sent to the user, and the user's judgment result of the action features at the judgment node is obtained to correct the first stir-frying action sequence, resulting in the third stir-frying action sequence. By comparing and determining the first stir-frying action sequence and the second stir-frying action sequence, to a certain extent, it makes up for the misjudgment of stir-frying actions in two monotonous recognition methods. When there are different judgment results for a certain stir-frying action in the determination of the first stir-frying action sequence and the second stir-frying action sequence, it can be corrected by the user's manual judgment result, so as to obtain a more correct third stir-frying action sequence. The operation video is analyzed for material recognition. Based on a pre-trained convolutional neural network, the type and quantity of each added ingredient and seasoning are recognized and used as the material recognition result, thereby realizing the recognition of the type and quantity of ingredients and seasonings during stir-frying. The material recognition result is used to expand the content corresponding to the seasoning and ingredient addition actions in the third stir-frying action sequence (it can also be manually added by the user to the content corresponding to the seasoning and ingredient addition actions in the third stir-frying action sequence), and recipe information is generated.The process of manual cooking by the user is recorded and the manual cooking process is recognized, so as to generate recipe information that can be included.

[0084] In a preferred embodiment, the first cooking action sequence is aligned with the second cooking action sequence, and a judgment node with different action characteristics is determined by comparison. The operation video segment corresponding to the judgment node is sent to the user, and the user's judgment result on the action characteristics of the judgment node is obtained to correct the first cooking action sequence. The obtained third cooking action sequence includes:

[0085] Step S1: Align the first cooking action sequence with the second cooking action sequence, determine the action type judgment result and occurrence time point of each action characteristic, and determine each occurrence time point as a judgment node;

[0086] Step S2: Determine the action type judgment result of the action characteristic on any judgment node in the first cooking action sequence, determine the action type judgment result of the judgment node at the corresponding position in the second cooking action sequence, and compare whether the two action type judgment results are consistent;

[0087] Step S3: If they are consistent, the action type judgment result on the judgment node in the first cooking action sequence is not changed;

[0088] Step S4: If they are inconsistent, the operation video segment corresponding to the judgment node is intercepted and sent to the user through a mobile terminal for manual recognition. The user's judgment result on the action characteristics of the judgment node is obtained to replace the action characteristics on the judgment node in the first cooking action sequence;

[0089] Step S5: All judgment nodes in the first cooking action sequence are processed according to Steps 2 to 4, and the processed first cooking action sequence is used as the third cooking action sequence.

[0090] The working principle and beneficial effects of the above technical solution are as follows: By aligning the first stir-frying action sequence with the second stir-frying action sequence, the action type judgment result and occurrence time point of each action feature are determined, and each occurrence time point is determined as a judgment node; By determining the action type judgment result of the action feature on any judgment node in the first stir-frying action sequence, the action type judgment result of the judgment node at the corresponding position in the second stir-frying action sequence is determined, and whether the two action type judgment results are consistent is compared. If they are consistent, the action type judgment result of the judgment node in the first stir-frying action sequence is not modified. If they are inconsistent, the operation video segment corresponding to the judgment node is intercepted and sent to the user through the mobile terminal for manual recognition, and the action feature judgment result of the user on the judgment node is obtained to replace the action feature of the judgment node in the first stir-frying action sequence. All judgment nodes in the first stir-frying action sequence are processed according to steps 2 to 4, and the processed first stir-frying action sequence is used as the third stir-frying action sequence. Thus, the comparison and recognition of the stir-frying actions on each corresponding judgment node in the stir-frying action sequences obtained by two different methods are realized, and there are no misjudged stir-frying actions in the adjusted third stir-frying action sequence.

[0091] In a preferred embodiment, the recipe analysis module includes:

[0092] The first collection unit is used to collect the first type of information in the recipe information; among them, the first type of information includes the types of ingredients and the amounts of ingredients contained in the recipe information;

[0093] The second collection unit is used to collect the second type of information in the recipe information: among them, the second type of information includes the cooking method, total cooking duration, cooking duration of each cooking stage, heat intensity, and the ingredients, seasonings and water ratios that need to be added in the recipe information;

[0094] The third collection unit is used to collect the third type of information in the recipe information; among them, the third type of information includes the taste preference, nutritional content, and energy content of the dish announced in the recipe information;

[0095] The formatting output unit is used to fill in the preset recipe template with the first type of information, the second type of information, and the third type of information to obtain a formatted recipe and output and save it.

[0096] The working principle and beneficial effects of the above technical solution are as follows: The recipe information is analyzed by means of keyword search, or the recipe information generated by using a preset recipe information editing template is directly obtained, and the recipe information is analyzed by collecting information according to the preset positions on the template. Among them, the first type of information in the recipe information is collected by the first collection unit; among them, the first type of information includes the types of ingredients and the amounts of ingredients included in the recipe information, which is convenient for users to search for corresponding recipes according to the types of ingredients, and is convenient for reminding users to prepare the corresponding ingredients for stir-frying; the second type of information in the recipe information is collected by the second collection unit: among them, the second type of information includes the cooking method, the total cooking time, the cooking time of each cooking stage, the size of the heat, and the ingredients, seasonings and water ratios that need to be added in the recipe information, which is convenient for the stir-frying machine to complete operations such as firepower adjustment, water addition, reminding to add corresponding ingredients, and stir-frying actions according to the corresponding stir-frying steps, cooking time, firepower size and other information; the third type of information in the recipe information is collected by the third collection unit; among them, the third type of information includes the taste preference, nutritional content, and energy content of the dishes announced in the recipe information, which is convenient for determining that users search for corresponding recipes according to taste preference, nutritional content, and energy content, or judging the nutritional content and energy content ingested by users according to the dishes recently cooked by the intelligent stir-frying machine; finally, the formatting output unit fills in a preset recipe template with the first type of information, the second type of information, and the third type of information to obtain a formatted recipe and outputs and saves it. Thus, the formatted analysis of the recipe information is realized, and the corresponding formatted recipe is obtained. The intelligent stir-frying machine can complete operations such as firepower adjustment, water addition, reminding to add corresponding ingredients, and stir-frying actions according to the corresponding stir-frying steps, cooking time, firepower size and other information in the formatted recipe.

[0097] In a preferred embodiment, the collection and management module includes:

[0098] A recipe recognition unit, which is used to receive the formatted recipe and analyze it to determine whether it lacks preset attribute information. If it lacks preset attribute information, it determines that the formatted recipe is invalid and abandons the formatted recipe;

[0099] A recipe processing unit, which is used to perform pre-processing before collection on the formatted recipe that does not lack preset attribute information and establish an index path;

[0100] A recipe storage unit, which is used to collect and store the formatted recipe in the intelligent stir-frying machine, and synchronize multiple formatted recipes stored in its internal memory to the corresponding account in the cloud platform when the intelligent stir-frying machine is connected to the Internet.

[0101] The working principle and beneficial effects of the above technical solution are as follows: a formatted recipe is received by a recipe recognition unit and analyzed to determine whether preset attribute information is missing. If the preset attribute information is missing, the formatted recipe is determined to be invalid and abandoned, wherein the necessary attribute information includes firepower size and cooking time, so as to prevent the formatted recipe from lacking necessary cooking steps and causing the intelligent cooking machine to be unable to work according to the corresponding steps; a formatted recipe that does not lack preset attribute information is pre-processed by a recipe processing unit, and an index path is established, so that the user can quickly search for the formatted recipe through different indexing methods; the formatted recipe is collected and stored in the intelligent cooking machine by a recipe storage unit, and when the intelligent cooking machine is connected to the Internet, a plurality of formatted recipes stored in the intelligent cooking machine are synchronized to the corresponding account of the intelligent cooking machine in the cloud platform, so as to prevent the loss of data stored in the intelligent cooking machine through cloud backup, so that the user can select the formatted recipe of his favorite dishes through cloud data anytime and anywhere and cook through other intelligent cooking machines, so that the intelligent cooking work is not limited to one intelligent cooking machine.

[0102] In a preferred embodiment, the recipe processing unit performs the following operations:

[0103] Extracting attribute information in a plurality of preset positions on the formatted recipe in sequence to obtain an attribute information sequence;

[0104] For any attribute information in the attribute information sequence, determine the position of the attribute information in the attribute information sequence, and extract the attribute keyword library preset for the position;

[0105] The attribute information is matched using the attribute keyword library to determine whether there is an identical keyword;

[0106] If the same keyword exists, a corresponding mapping relationship between the formatted recipe and the keyword is established;

[0107] If the same keyword does not exist, a new keyword is created in the attribute keyword library using the attribute information, and a mapping relationship between the formatted recipe and the new keyword is established.

[0108] The working principle and beneficial effects of the above technical solution are as follows: By sequentially extracting the attribute information in multiple preset positions on the formatted recipe, an attribute information sequence is obtained; for any attribute information in the attribute information sequence, determine the position where the attribute information is arranged in the attribute information sequence, and extract the attribute keyword library preset for this position; use this attribute keyword library to match the attribute information to determine whether there are the same keywords; if there are the same keywords, establish a corresponding mapping relationship between this formatted recipe and this keyword; if there are no same keywords, create a new keyword in this attribute keyword library using this attribute information, and establish a mapping relationship between this formatted recipe and the new keyword. Thus, by establishing a mapping relationship between the formatted recipe and the keyword, it enables users to quickly retrieve the formatted recipes that have been included through the keyword.

[0109] In a preferred embodiment, the recipe execution module includes:

[0110] A recipe selection unit for finding the corresponding formatted recipe according to the dish name selected by the user;

[0111] A recipe deconstruction unit for deconstructing the selected formatted recipe to determine the multiple cooking steps corresponding to this recipe;

[0112] A recipe execution unit for starting the intelligent cooking machine to execute the cooking work according to the formatted recipe, and reminding the user to add ingredients and seasonings according to the content of the formatted recipe.

[0113] The working principle and beneficial effects of the above technical solution are as follows: The recipe selection unit finds the corresponding formatted recipe according to the dish name selected by the user; the recipe deconstruction unit deconstructs the selected formatted recipe to determine the multiple cooking steps corresponding to this recipe; the recipe execution unit starts the intelligent cooking machine to execute the cooking work according to the formatted recipe, and reminds the user to add ingredients and seasonings according to the content of the formatted recipe. It realizes finding the corresponding formatted recipe according to the dish name, and after analyzing the multiple cooking steps in the formatted recipe, starting the intelligent cooking machine to execute the cooking work according to the formatted recipe, and reminding the user to add ingredients and seasonings according to the content of the formatted recipe, so as to complete a series of cooking work. It realizes the retrieval and execution work of the formatted recipe.

[0114] In a preferred embodiment, it further includes a diet health management module, and the diet health management module performs the following operations:

[0115] Determine the dishes and the amounts of the dishes cooked by the user through the intelligent cooking machine within a preset time length recently;

[0116] Determine the formatted recipe corresponding to each dish, and determine the nutrient content and energy content within each dish according to the formatted recipe;

[0117] Based on the dishes and the amounts of the dishes cooked by the user through the intelligent cooking machine within a recently preset time period, calculate the total amount of nutrients and the total amount of energy ingested by the user based on the nutrient content and energy content within each dish;

[0118] Compare and calculate the total amount of nutrients and the total amount of energy ingested by the user with the preset nutrient balance intake table to determine the nutrients and energy that the user lacks and the nutrients and energy that the user has ingested in excess;

[0119] Select the first type of formatted recipe for supplementing the corresponding nutrients and energy according to the nutrients and energy that the user lacks;

[0120] Select the second type of formatted recipe for reducing the corresponding nutrients and energy according to the nutrients and energy that the user has ingested in excess;

[0121] Comprehensively screen out the third type of formatted recipe according to the first type of formatted recipe and the second type of formatted recipe to participate in the user's balanced and healthy diet plan.

[0122] The working principle and beneficial effects of the above technical solution are as follows: By determining the dishes and the amounts of the dishes cooked by the user through the intelligent cooking machine within a recently preset time period; then determining the formatted recipe corresponding to each dish, and determining the nutrient content and energy content within each dish (the nutrient content and energy content contained in the unit weight of the dish); based on the dishes and the amounts of the dishes cooked by the user through the intelligent cooking machine within a recently preset time period (such as one day, one week, or one month, etc.), calculate the total amount of nutrients and the total amount of energy ingested by the user based on the nutrient content and energy content within each dish; compare and calculate the total amount of nutrients and the total amount of energy ingested by the user with the preset nutrient balance intake table to determine the nutrients and energy that the user lacks and the nutrients and energy that the user has ingested in excess; select the first type of formatted recipe for supplementing the corresponding nutrients and energy according to the nutrients and energy that the user lacks; select the second type of formatted recipe for reducing the corresponding nutrients and energy according to the nutrients and energy that the user has ingested in excess; comprehensively screen out the third type of formatted recipe according to the first type of formatted recipe and the second type of formatted recipe to participate in the user's balanced and healthy diet plan. Thus, the balanced management of the user's diet health is realized. On the premise of long-term use of this function, the nutrients and energy ingested by the user can be effectively adjusted dynamically, thereby improving the user experience.

[0123] In a preferred embodiment, comparing and calculating the total amount of nutrients and the total amount of energy ingested by the user with the preset nutrient balance intake table includes:

[0124] Represent the total amount of nutrients and total energy ingested by the user using the first data set: N = {N1, N2, N3,..., N i , N0};

[0125] Represent the preset nutritional balance intake table using the second data set: M = {M1, M2, M3,..., M i , M0};

[0126] Calculate the third data set Z using the first data set N and the second data set M, where:

[0127]

[0128] In the formula, N1 represents the total intake of the first nutrient component by the user, and so on, N i represents the total intake of the i-th nutrient component by the user, N0 represents the total intake of energy by the user, M1 represents the demand for the first nutrient component by the user in the nutritional balance intake table, and so on, M i represents the demand for the i-th nutrient component by the user in the nutritional balance intake table, M0 represents the demand for energy by the user, K1 represents the preset deviation correction coefficient for the first nutrient component, and so on, K i represents the preset deviation correction coefficient for the i-th nutrient component, and K0 represents the preset deviation correction coefficient for energy;

[0129] Determine that the nutrient components or energy with values greater than the preset first deviation threshold in the third data set are the nutrient components or energy with excessive intake;

[0130] Determine that the nutrient components or energy with values less than the preset second deviation threshold in the third data set are the nutrient components or energy with insufficient intake.

[0131] The working principle and beneficial effects of the above technical solution are: By setting a standard nutritional balance intake table and comparing and calculating it with the total amount of nutrients and energy already ingested by the user, the nutrient components and energy lacking by the user and the nutrient components and energy excessively ingested by the user can be determined.

[0132] In a preferred embodiment, it further includes a seasonal recipe recommendation module. The seasonal recipe recommendation module is used to determine the production seasons of the main ingredients in all formatted recipes, and recommend the formatted recipes corresponding to the various ingredients produced in the current season to the user.

[0133] The beneficial effect of the above technical solution is: It realizes the recommendation of formatted recipes corresponding to seasonal vegetables to the user, can reduce the user's vegetable purchase cost, and improve the user's experience.

[0134] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims

1. A recipe input system for an intelligent cooking machine, characterized in that, Including: A recipe collection module for obtaining recipe information; wherein, the recipe collection module includes: A network search unit for searching for corresponding recipe information from the Internet through keywords set by the user; A network collection unit for regularly obtaining updated recipe information from a preset website; A recipe editing unit for sending an editing template of preset recipe information to the user and obtaining the edited template filled by the user as recipe information; An identification and collection unit, and the identification and collection unit performs the following operations: Obtaining an operation video of the user during the process of manually stir-frying with a preset collection wok through a preset camera; Performing action recognition and analysis on the operation video, determining multiple first action features included therein and classifying them; wherein, the first action features include stir-frying, heat adjustment, addition of seasonings and ingredients, and taking the dish out of the pot; Sorting and labeling all the first action features in the entire stir-frying process corresponding to the operation video in time series to obtain a first stir-frying action sequence; Determining the second action features of the user according to the pressure values collected by multiple pressure sensors preset on the collection wok and the pan surface deflection angle values collected by a gyroscope sensor, and determining the third action features of the user according to the gear signals collected by a fire power gear sensor preset on the collection wok; wherein, the second action features include flipping the wok, stir-frying, and taking the dish out of the pot; Sorting and labeling the second action features and the third action features analyzed by the sensors on the collection wok in time series to obtain a second stir-frying action sequence; Aligning the first stir-frying action sequence with the second stir-frying action sequence, comparing and determining the judgment nodes with different action features therein, sending the operation video segment corresponding to the judgment node to the user and obtaining the user's judgment result of the action features of the judgment node to correct the first stir-frying action sequence to obtain a third stir-frying action sequence; Performing material recognition and analysis on the operation video, identifying the types and quantities of the ingredients and seasonings added each time based on a pre-trained convolutional neural network, and taking them as the material recognition results; Using the material recognition results to expand the content corresponding to the actions of adding seasonings and ingredients in the third stir-frying action sequence, and generating recipe information; A recipe analysis module for analyzing the collected recipe information, extracting various attribute information in the recipe and saving it as a formatted recipe; A collection management module for collecting the formatted recipes through the user account and synchronizing them to the cloud network; A recipe execution module for importing the formatted recipe selected by the user into an intelligent stir-frying machine for cooking work.

2. The recipe input system of an intelligent cooking machine according to claim 1, characterized in that, The aligning the first stir-frying action sequence with the second stir-frying action sequence, comparing and determining the judgment nodes with different action features therein, sending the operation video segment corresponding to the judgment node to the user and obtaining the user's judgment result of the action features of the judgment node to correct the first stir-frying action sequence to obtain a third stir-frying action sequence includes: Step S1: Align the first stir-fry action sequence with the second stir-fry action sequence, determine the action type judgment result and occurrence time point of each action feature, and determine each occurrence time point as a judgment node; Step S2: Determine the action type judgment result of the action feature at any judgment node in the first stir-fry action sequence, determine the action type judgment result of the action feature at the corresponding judgment node in the second stir-fry action sequence, and compare whether the two action type judgment results are consistent; Step S3: If they are consistent, do not modify the action type judgment result of the action feature at this judgment node in the first stir-fry action sequence; Step S4: If they are inconsistent, intercept the operation video segment corresponding to this judgment node, and send it to the user through the mobile terminal for manual recognition, and obtain the action feature judgment result of the user for this judgment node to replace the action feature at this judgment node in the first stir-fry action sequence; Step S5: Process all judgment nodes in the first stir-fry action sequence according to Steps 2 to 4, and use the processed first stir-fry action sequence as the third stir-fry action sequence.

3. The recipe input system of an intelligent cooking machine according to claim 1, characterized in that, The recipe analysis module includes: The first acquisition unit is used to acquire the first type of information in the recipe information; wherein, the first type of information includes the types of ingredients and the amounts of ingredients included in the recipe information; The second acquisition unit is used to acquire the second type of information in the recipe information: wherein, the second type of information includes the cooking method, total cooking duration, cooking duration of each cooking stage, heat intensity, and ingredients, seasonings and water ratio to be added included in the recipe information; The third acquisition unit is used to acquire the third type of information in the recipe information; wherein, the third type of information includes the taste preference, nutritional content, and energy content of the dish announced in the recipe information; The formatted output unit is used to fill in a preset recipe template with the first type of information, the second type of information, and the third type of information to obtain a formatted recipe and output and save it.

4. The recipe input system of an intelligent cooking machine according to claim 1, characterized in that, The collection management module includes: The recipe recognition unit is used to receive the formatted recipe and analyze it to determine whether it lacks preset attribute information. If it lacks preset attribute information, it determines that the formatted recipe is invalid and abandons the formatted recipe; The recipe processing unit is used to perform pre-processing before collection on the formatted recipe that does not lack preset attribute information and establish an index path; The recipe storage unit is used to collect and store the formatted recipe in the intelligent stir-fry machine, and synchronize multiple formatted recipes stored in it to the account corresponding to the intelligent stir-fry machine in the cloud platform when the intelligent stir-fry machine is connected to the Internet.

5. The recipe input system of an intelligent cooking machine according to claim 4, characterized in that, The recipe processing unit performs the following operations: Extract the attribute information in multiple preset positions on the formatted recipe in sequence to obtain an attribute information sequence; For any attribute information in the attribute information sequence, determine the position of the attribute information in the attribute information sequence, and extract the preset attribute keyword library for this position; Use the attribute keyword library to match the attribute information to determine whether there are the same keywords; If there are identical keywords, establish the corresponding mapping relationship between the formatted recipe and the keyword; If there are no identical keywords, use the attribute information to create a new keyword in the attribute keyword library and establish the mapping relationship between the formatted recipe and the new keyword.

6. The recipe input system of an intelligent cooking machine according to claim 1, characterized in that, The recipe execution module includes: A recipe selection unit for finding the corresponding formatted recipe according to the dish name selected by the user; A recipe deconstruction unit for deconstructing the selected formatted recipe to determine the multiple cooking steps corresponding to the recipe; A recipe execution unit for starting the intelligent cooking machine to execute the cooking work according to the formatted recipe and reminding the user to add ingredients and seasonings according to the content of the formatted recipe.

7. The recipe input system of an intelligent cooking machine according to claim 1, characterized in that It also includes a diet and health management module, and the diet and health management module performs the following operations: Determine the dishes and the portion of the dishes cooked by the user through the intelligent cooking machine within a preset time length recently; Determine the formatted recipes corresponding to each dish, and determine the nutrient content and energy content in each dish according to the formatted recipes; Calculate the total amount of nutrients and energy intake of the user based on the nutrient content and energy content in each dish according to the dishes and the portion of the dishes cooked by the user through the intelligent cooking machine within a preset time length recently; Compare and calculate the total amount of nutrients and energy intake of the user with the preset nutrient balance intake table to determine the nutrients and energy lacking by the user and the nutrients and energy excessively ingested by the user; Select the first type of formatted recipes for supplementing the corresponding nutrients and energy according to the nutrients and energy lacking by the user; Select the second type of formatted recipes for reducing the corresponding nutrients and energy according to the nutrients and energy excessively ingested by the user; Comprehensively screen out the third type of formatted recipes according to the first type of formatted recipes and the second type of formatted recipes to participate in the user's balanced and healthy diet plan.

8. The recipe input system of an intelligent cooking machine according to claim 1, wherein It also includes a seasonal recipe recommendation module, and the seasonal recipe recommendation module is used to determine the production seasons of the main ingredients in all formatted recipes and recommend the formatted recipes corresponding to the various ingredients produced in the current season to the user.

Images