CN111276147A - A method for recording diet based on voice input - Google Patents

Abstract

The invention discloses a diet recording method based on voice input. The method runs on a client and a server. Step 1. The user uses voice to input his own diet record through the client, uses voice recognition technology to recognize and translate the voice record It is text information, which is stored according to the specified data record format; step 2, the text information is uploaded to the server; step 3, use the fuzzy matching algorithm to search the diet records in the table one by one in the database, and re-list the search results back to the client for further processing Confirm, and call their corresponding nutrition data information to feed back to the client. In step 4, the unretrieved dishes suggest that the user convert to ingredient information input, and repeat steps 1 to 3 again to complete the input of diet detail records. The invention can improve the efficiency of the user's diet recording, and is convenient for follow-up diet analysis and consultation with a nutritionist.

Description

A method for recording diet based on voice input

technical field

The invention relates to the field of mobile medical information platforms, in particular to a method for recording dietary intake by using voice input.

Background technique

Nutrition is an important material basis for human life, growth and health. In recent years, the living standards of the Chinese people have been continuously improved, the nutritional supply capacity has been significantly enhanced, and the national nutrition and health status has been significantly improved. However, there are still problems such as coexistence of undernutrition and overnutrition among residents, frequent occurrence of nutrition-related diseases, and the lack of popularization of a nutritious and healthy lifestyle.

Paying attention to dietary balance is a science. The diagnosis and intervention of nutrition disciplines are closely related to the daily diet of users. Nutrition department clinics generally require users to record their own diet for a week. However, there are not many options for diet recording. Nowadays, with the rapid development of "Internet +", some people have proposed to use smart terminals to record the daily diet of users, and enable nutritionists to complete online consultations through smart terminals. However, the diet recording process of the current food nutrition analysis software is completed by manually inputting the name to retrieve food - selecting food intake - selecting intake quality, and the user's operation complexity is relatively high.

The problem to be solved by speech recognition technology (Auto Speech Recognize, ASR) is to enable the machine to "understand" human speech and "extract" the text information contained in the speech, which is equivalent to installing "ears" on the machine to make it possible. Possess the ability to "listen". In recent years, speech recognition technology has made great progress, and iFLYTEK is one of the leaders. It has the world's leading continuous speech recognition technology, with an accuracy rate of over 95%, a speech input speed of 180 words/min, and a corresponding time for the recognition result. Less than 500ms, and supports 5 languages of Chinese, English, Cantonese, Tibetan and Uyghur, as well as many regional dialects such as Sichuan, Henan, and Northeast China. The iFLYTEK speech recognition SDK is small and powerful, with no network burden and offline recognition.

SUMMARY OF THE INVENTION

In order to solve the above problems in the prior art, the present invention proposes a method for recording diet based on voice input. An Internet-based intelligent terminal utilizes iFLYTEK intelligent voice recognition technology to complete diet records, thereby improving the efficiency and accuracy of diet records.

A method for recording diet based on voice input of the present invention, the method runs on a client and a server, and specifically includes the following steps:

Step 1. Through the client, the user inputs his own dietary details record by voice, uses the voice recognition technology to recognize the voice record, translates it into text information, and stores it according to the specified data record format;

Step 2, the text information is uploaded to the server;

Step 3. Use the fuzzy matching algorithm to retrieve the dietary records in the table one by one in the database, re-list the retrieval results and feed them back to the client for confirmation, and call their corresponding nutritional component data information to feed back to the client,

Step 4. For the dishes that cannot be retrieved, the user is advised to switch to the input of ingredient information, and repeat steps 1 to 3 again to complete the input of the dietary details record.

The invention can improve the efficiency of the user's diet record, enable the user to complete the daily diet record in a timely manner, facilitate subsequent diet analysis and dietitian consultation, and ensure the user's diet health.

Description of drawings

1 is a block diagram of a method for recording diet based on voice input of the present invention;

Fig. 2 is the flow chart of a kind of diet recording method based on voice input of the present invention;

FIG. 3 is an example diagram of a user information confirmation interface.

Detailed ways

The technical solutions of the present invention will be described in detail below with reference to the accompanying drawings.

The general idea of the present invention is that. This method mainly includes the following steps:

As shown in Figure 1 and Figure 2, the voice input-based diet recording method of the present invention specifically includes the following procedures:

Step 1. Through the user-side client, the user uses voice to input his own dietary records, and stores them according to the specified data record format. The main input contents include the name of the ingredients and the intake quality, or the name and intake quality of common dishes, and the voice input The format is set to "Item N, food name, intake quality" (N represents the number of dietary records), for example: "Item 1, egg 100g; item 2, yogurt 150g; item 3 Kung Pao Chicken 100g; The fourth item, stir-fried pork with green peppers 100g." Utilize the offline voice recognition function of iFLYTEK's voice recognition plug-in, integrate it in the client, translate the voice diet records into text information in real time, and upload the text information to the server in real time ;

Table 2

Numbering food name intake quality 1 egg 100g 2 yogurt 200g 3 Kung Pao Chicken 100g 4 Stir-fried Pork with Green Peppers 100g

Step 2, according to the text information uploaded by the user, the server converts the text into a table form according to the format established during the voice input in step 1, as shown in Table 2, which is an example of converting text information into a table, which is convenient for subsequent step operations;

Step 3. Using the information in the diet record table, use the fuzzy matching algorithm to search through the database one by one. The fuzzy matching algorithm is embodied in that, according to the keywords entered by the user's voice, such as eggs, all the ingredients or dishes containing the word "egg" are searched in the database, and the five closest ingredients or dishes are selected, and their corresponding nutrients are called. The data information is fed back to the user-side client, and then the user can select the ingredients or dishes that are closest to what they want to input according to the five options provided, and click OK.

As shown in Figure 2, it is an example diagram of the user information confirmation interface, and the user is requested to carry out further information confirmation;

Step 4. If the dishes that cannot be retrieved in the database are found, the system recommends that the user switch to the input of ingredients. See the example of the user confirmation information form in Figure 3. The user uses voice to input the type and quality of ingredients again, and repeats the above steps to complete the diet record.

Claims (1)

1. a diet recording method based on voice input, is characterized in that, this method runs on client and server, specifically comprises the following steps: Step 1. Through the client, the user inputs his own dietary details record by voice, uses the voice recognition technology to recognize the voice record, translates it into text information, and stores it according to the specified data record format; Step 2, the text information is uploaded to the server; Step 3. Use the fuzzy matching algorithm to retrieve the dietary records in the table one by one in the database, re-list the retrieval results and feed them back to the client for confirmation, and call their corresponding nutritional component data information to feed back to the client, Step 4. For the dishes that cannot be retrieved, the user is advised to switch to the input of ingredient information, and repeat steps 1 to 3 again to complete the input of the dietary details record.

Images