Agricultural science popularization knowledge base interaction system and method based on intelligent terminal

Abstract

The invention relates to the field of agricultural science popularization, in particular to an agricultural science popularization knowledge base interaction system and method based on an intelligent terminal, and the system comprises a display module which is used for displaying science popularization content; the image acquisition module is used for acquiring facial expressions and shot images of a user; the expression recognition module is used for recognizing the satisfaction degree of the user for watching the popular science content through an expression recognition technology; the popular science content is a video set formed by splicing a plurality of video clips, the video clips are skipped through skipping pictures, the video clips are bound with preference classification data, and the skipping module is used for skipping the popular science content according to the satisfaction degree recognized by the expression recognition module when the user watches the popular science content. The number of preference classification video clips with high satisfaction is increased through the jumping picture, and the number of preference classification video clips with low satisfaction is reduced. According to the technical scheme, the science popularization knowledge content is dynamically designed based on the feedback of the user, so that the conflict emotion of the user on the agricultural science popularization process is reduced.

Description

Technical Field

[0001] The present invention relates to the field of agricultural science popularization, and in particular to an agricultural science popularization knowledge base interaction system and method based on an intelligent terminal. Background Art

[0002] Digital technology provides a wealth of information resources and display methods for agricultural science popularization. Through digital means, agricultural knowledge can be presented in the form of pictures, videos, animations, etc., making the popular science content more vivid and intuitive. These technologies can simulate agricultural production scenes, allowing users to understand the agricultural production process and technical points in an immersive way. Through interactive experience, users' learning interest and participation can be improved. Using intelligent interactive platforms, such as mobile APPs and websites, agricultural science knowledge can be easily obtained and online consultation and communication can be carried out. These platforms can also provide personalized learning resources and recommendations based on user needs and interests.

[0003] However, agricultural production is often accompanied by boring content, which may reduce the enthusiasm of users. Therefore, it is necessary to dynamically design the popular science knowledge content based on user feedback to reduce users' resistance to the agricultural science popularization process. Summary of the invention

[0004] In order to solve the above problems, the present invention provides an agricultural science knowledge base interactive system and method based on a smart terminal, which can recognize the user's expression changes and switch the science content.

[0005] In order to achieve the above-mentioned purpose, the technical solution of the present invention is as follows: an agricultural science knowledge base interactive system based on a smart terminal, comprising:

[0006] Display module: used to display popular science content;

[0007] Image acquisition module: used to collect user facial expressions and take images;

[0008] Expression recognition module: used to identify the user's satisfaction with watching popular science content through expression recognition technology;

[0009] Jump module: The popular science content is a video collection composed of several video clips. The video clips are jumped by jump screens. The video clips are bound with preference classification data. The jump module is used to increase the number of preference classification video clips with high satisfaction and reduce the number of preference classification video clips with low satisfaction through jump screens when the user watches the popular science content, based on the satisfaction identified by the expression recognition module.

[0010] The above scheme has the following beneficial effects:

[0011] 1. In this solution, users can watch popular science content through the display module. During the viewing, the user's facial expressions can be collected to judge the user's satisfaction with the popular science content, thereby analyzing whether the user is interested in the popular science content.

[0012] 2. In this solution, the popular science content is composed of several video clips, and the playback content can be switched by inserting a jump screen. When the user is watching the popular science content, the user's satisfaction is monitored. After the current video clip is played, the selection and playback of the video clip is switched according to the satisfaction.

[0013] Furthermore, it also includes an environment recognition module, which is used to recognize the scene in the image captured by the image acquisition module, so as to control the display module to display the popular science content related to the scene; the environment recognition module is also used to analyze the user's position in the scene based on the scene in the captured image, make detailed judgments on the popular science content according to the position, and control the display module to display the popular science content related to the position of the scene; it also includes a voice recognition module, which is used to recognize the user's voice commands and adjust the display of the popular science content according to the voice commands.

[0014] Furthermore, it also includes a mobile terminal, which is equipped with a touch screen, front and rear cameras, a microphone and a speaker; the touch screen is an anti-peeping screen, which is used to restrict the user within the facial expression collection range, and the front and rear cameras include a front camera and a rear camera.

[0015] Furthermore, the environment recognition module collects environment pictures through the front and rear cameras of the mobile terminal for scene recognition, and its specific logic is as follows: first, the collected pictures are preprocessed, and then the convolutional neural network in the deep learning algorithm is used to extract feature information of the preprocessed images. The convolutional neural network automatically learns feature information of different levels in the image through multiple convolutional layers and pooling layers, and matches and compares the extracted feature information with the pre-trained agricultural scene feature database. By calculating the similarity between the feature information, when the similarity exceeds the set threshold, the scene to which the current captured picture belongs can be determined.

[0016] Furthermore, it also includes a video clip repair module. The preference classification data includes the agricultural classification, location, work stage, video length, names of objects and organisms in the screen, and appearance time of the popular science content. The video clip repair module is used to collect the overall satisfaction rating of each video clip, extract the preference classification data of video clips with low satisfaction, and delete the appearance time of objects or organisms with common characteristics based on the common characteristics of the video clips with low satisfaction.

[0017] Furthermore, the mobile terminal is a smart phone, and the smart phone is also used to obtain networking mode information and positioning information when in charging state. If the networking mode information is WIFI networking and the positioning information remains unchanged for a preset period of time, the positioning information is defined as indoor residential scene information. If the positioning information is not in the indoor residential scene information, it is non-indoor residential scene information; the smart phone is also used to periodically obtain the positioning information and networking mode information of the GPS module or Beidou module of the smart phone when the user watches popular science content, and access the gyroscope of the smart phone to obtain the posture information of the smart phone; when the posture information of the smart phone changes from horizontal to vertical and lasts for more than a preset time, if the networking mode is data traffic and the positioning information is non-indoor residential scene information, the front and rear cameras are simultaneously started to capture environmental pictures and perform scene recognition; if the networking mode is WIFI and the positioning information is indoor residential scene information, only the rear camera is started; if the networking mode is data traffic and the positioning information is indoor residential scene information, a networking mode modification request is generated and only the rear camera is started.

[0018] Furthermore, the smartphone is also used to periodically obtain positioning information and networking information when the user is watching popular science content, access the gyroscope of the smartphone to obtain the posture information of the smartphone, and when the posture information of the smartphone is horizontal, collect light intensity information and light spectrum information of the ambient light through the front camera, and then determine if the light intensity information is higher than a preset threshold, and the spectral intensity of the light spectrum information that is biased towards ultraviolet rays is higher than the preset threshold, then define the positioning information as outdoor scene information; if the light intensity information is lower than the preset threshold, and the spectral intensity of the light spectrum information that is biased towards ultraviolet rays is lower than the preset threshold, then define the positioning information as indoor work scene information.

[0019] Furthermore, the outdoor scene information includes outdoor planting scenes and outdoor breeding scenes; the indoor work scene information includes indoor cultivation scenes and indoor planting scenes.

[0020] Furthermore, the environment recognition module is also used to identify landmarks in environment pictures captured by the front camera and / or rear camera based on the differences between indoor living scene information, outdoor planting scenes, outdoor breeding scenes, indoor cultivation scenes and indoor planting scenes, and play fixed video clips when the landmarks are identified.

[0021] Scene recognition is performed by fusing multiple information, which significantly improves the accuracy of recognition. The combination of networking information, positioning information, and gyroscope posture information obtained by smartphones in different states can more accurately determine the actual scene environment in which the user is located. For example, positioning information is used to determine indoor living scene information when charging, and light intensity and spectrum information collected by the front camera when the posture is horizontal are used to distinguish indoor and outdoor scenes. These multi-dimensional information comprehensive analyses effectively avoid possible misjudgments caused by a single information source, thereby greatly improving the accuracy of scene recognition, allowing the system to push highly relevant agricultural science content to users based on accurate scenes.

[0022] Based on the above scene recognition strategy, the system can effectively reduce the amount of calculation in the recognition process. In different scenarios, the camera is started for image acquisition in a targeted manner. For example, in an indoor living scene and when connected to the WIFI network, only the rear camera is started, avoiding unnecessary simultaneous activation of the front and rear cameras and related image processing operations. At the same time, the scene is preliminarily screened and classified in advance through preset information (such as indoor living scene information, light intensity threshold, etc.), reducing the subsequent complex image feature matching and analysis operations, improving the overall operation efficiency of the system, making the scene recognition process more efficient and faster, reducing user waiting time, and further improving the user experience.

[0023] From the perspective of user interaction experience, this scene recognition method has brought about improvements in many aspects. First of all, accurate scene recognition can ensure that users are pushed scientific knowledge that meets their current scene needs. Whether in indoor work scenes, outdoor planting and breeding scenes, or indoor living scenes, users can obtain targeted and practical agricultural science information in a timely manner, which enhances the practicality and attractiveness of science content. Secondly, the system automatically adjusts the camera startup strategy and networking method recommendations according to the scene, reducing the tediousness of user manual operations. For example, when there is high data traffic and indoor living scenes, a request to modify the networking method is automatically generated to remind users to switch networks to save traffic costs. This reflects the intelligence and humanity of the system, allowing users to feel convenient and considerate during use, thereby comprehensively improving the interactive experience between users and the agricultural science knowledge base interactive system.

[0024] An agricultural science knowledge base interaction method based on an intelligent terminal, comprising:

[0025] Step 1: The user carries the mobile terminal to the agricultural site;

[0026] Step 2: Users take photos of agricultural content they are interested in;

[0027] Step 3: The user adjusts the head position to watch the popular science content played on the touch screen;

[0028] Step 4: The mobile terminal captures the user's facial expression to analyze the user's viewing satisfaction, and adjusts the popular science content played through the jump module.

[0029] By shooting with mobile terminals, users can carry their mobile terminals to agricultural scenes of their interest, and the mobile terminals will interact with the agricultural scenes and users to customize the output of popular science content.

[0030] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a module schematic diagram of an embodiment of an agricultural science popularization knowledge base interactive system based on a smart terminal of the present invention;

[0032] Figure 2 The present invention is a schematic diagram of an embodiment of an agricultural science knowledge base interactive system and method based on a smart terminal. DETAILED DESCRIPTION

[0033] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0035] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0036] The following is further described in detail through specific implementation methods:

[0037] As attached Figure 1-Figure 2 As shown: The interactive system of agricultural science knowledge base based on intelligent terminal includes:

[0038] Display module: used to display popular science content.

[0039] Image acquisition module: used to collect user facial expressions and take images.

[0040] Expression recognition module: used to identify the user's satisfaction with watching popular science content through expression recognition technology.

[0041] Jump module: The popular science content is a video collection composed of several video clips. The video clips are jumped by jump screens. The video clips are bound with preference classification data. The jump module is used to increase the number of preference classification video clips with high satisfaction and reduce the number of preference classification video clips with low satisfaction through jump screens when the user watches the popular science content, based on the satisfaction identified by the expression recognition module.

[0042] Environment recognition module: used to recognize the scene in the image captured by the image acquisition module, so as to control the display module to display the popular science content related to the scene. The environment recognition module is also used to analyze the user's location in the scene based on the scene in the captured image, make a detailed judgment on the popular science content based on the location, and control the display module to display the popular science content related to the location of the scene. The environment recognition module is also used to identify the landmark site in the location, and play a fixed video clip after the landmark site is identified.

[0043] Voice recognition module: The voice recognition module is used to recognize the user's voice commands and adjust the display of popular science content according to the voice commands.

[0044] Mobile terminal: The mobile terminal is a mobile phone. The expression recognition module, jump module, environment recognition module, and voice recognition module are integrated in the mobile terminal in the form of an APP. The mobile terminal is equipped with a touch screen, front and rear cameras, a microphone, and a speaker.

[0045] The video clip repair module, the preference classification data includes the agricultural classification, location, work stage, video length, name of objects and organisms in the screen, and appearance time of popular science content. The video clip repair module is used to collect the overall satisfaction rating of each video clip, extract the preference classification data of video clips with low satisfaction, and delete the appearance time of objects or organisms with common characteristics according to the common characteristics of video clips with low satisfaction. The video clip repair module is integrated on the cloud server.

[0046] The user brings a mobile terminal into the agricultural scene, selects the agricultural facilities of interest, such as fish ponds, farmlands and fruit trees, and uses the mobile terminal to take images of the agricultural facilities. The jump module will identify the agricultural type through the images of the agricultural facilities, retrieve and play video clips of the relevant agricultural type in the video clips, and combine several video clips to piece together a complete video set. The video clips are jumped between by jumping screens.

[0047] In addition to the recognition of agricultural scenes, at some guided or special agricultural projects, such as signboards, rice-shrimp mixed agricultural models, photovoltaic agricultural models, etc., it is possible to identify and retrieve through landmark sites, and then play customized fixed explanatory video clips.

[0048] The voice recognition module supports the use of voice to control the popular science process. Voice commands can simplify the user's control operations and can control the popular science content without calling out the function menu when it is playing, thereby improving the user experience.

[0049] After the video clip is broadcast, the expression recognition module will obtain the user's expression by collecting the user's facial expression, and output the satisfaction through the positive and negative emotions of the expression. It will also judge the user's interest in the current video clip based on the satisfaction, thereby changing the subsequent output video clips to make the output video clips more in line with the user's preference classification, thereby reducing the user's boredom, optimizing the agricultural science popularization process, and improving the user experience.

[0050] For example, when a user is watching a science video on turning a fish pond, the entire process is divided into video clips such as drainage, dredging, drying, disinfection and water injection, and after the science video on turning a fish pond is finished, it will be connected to a science video on releasing fish fry. When the user is not dissatisfied with the current video clip, the entire science video is formed by splicing together multiple video clips of related fish pond turning, and the user will watch the entire fish pond turning steps in sequence; when the user is dissatisfied with any process of turning a fish pond, subsequent clips will be reselected, for example, the three steps of dredging, drying and disinfection will be simplified and merged into one video clip by removing details, or the steps of dredging and drying will be directly deleted to quickly advance to the science video of releasing fish fry.

[0051] After the user takes a photo of the agricultural facilities, the jump module will also determine the user's location based on the photographed image. Taking the fish pond as an example, the user will have different popular science content output when taking photos on the shore and taking a boat to the center of the fish pond. The corresponding agricultural production process will be output according to the user's own location. For example, shooting on the shore will output the embankment construction, fish fry release and shore bait release, while shooting in the center of the fish pond will output the processes of casting nets, salvaging floating objects and releasing bait from the hull, etc., thereby further refining the popular science content, making it more vivid and immersive.

[0052] The jump module is also used to speed up the playing speed of the video clip until the jump screen when the satisfaction level recognized by the expression recognition module is low.

[0053] When it is identified that the user's satisfaction is low, the playback speed of the video clip is accelerated so that the jump screen can be reached faster and the popular science content can be switched.

[0054] In addition to being able to switch the user experience of popular science content through satisfaction recognition, the content of each video set can also be improved through satisfaction. The low satisfaction of some videos may be due to the appearance of some objects or creatures, such as silt, peach oil, toads and insects, which may make some users feel disgusted when watching. The video clip repair module will extract the common characteristics of each video clip with low satisfaction, and delete the corresponding clips in the video clips, coded the content in the video clips, or replace the objects or creatures in the video clips with cartoons according to their common characteristics, so as to improve the satisfaction of the original video clips and repair the popular science process.

[0055] At the same time, there is a disgust tag library bound to user information, which is used to extract the common characteristics of video clips played by a single user when he or she has multiple disgust emotions. When the video clip watched by the user subsequently contains content in the disgust tag library, the video clip modified by the disgust tag library based on the user information is selected for playback.

[0056] The touch screen is an anti-peeping screen, which is used to restrict users within the range of facial expression collection.

[0057] The anti-peep screen reduces the display effect of the touch screen when viewed at a more oblique angle, thereby restricting the user's head position when viewing popular science content, making it easier to collect the user's facial expressions.

[0058] An agricultural science knowledge base interaction method based on an intelligent terminal, comprising:

[0059] Step 1: The user carries the mobile terminal to the agricultural site;

[0060] Step 2: Users take photos of agricultural content they are interested in;

[0061] Step 3: The user adjusts the head position to watch the popular science content played on the touch screen;

[0062] Step 4: The mobile terminal captures the user's facial expression to analyze the user's viewing satisfaction, and adjusts the popular science content played through the jump module.

[0063] In another embodiment, the mobile terminal is a smart phone, and the smart phone is also used to obtain networking mode information and positioning information when in charging state. If the networking mode information is WIFI networking and the positioning information remains unchanged for a preset period of time, the positioning information is defined as indoor residential scene information. If the positioning information is not in the indoor residential scene information, it is non-indoor residential scene information; the smart phone is also used to periodically obtain the positioning information and networking mode information of the GPS module or Beidou module of the smart phone when the user watches popular science content, and access the gyroscope of the smart phone to obtain the posture information of the smart phone; when the posture information of the smart phone changes from horizontal to vertical and lasts for more than a preset time, if the networking mode is data traffic and the positioning information is non-indoor residential scene information, the front and rear cameras are simultaneously started to capture environmental pictures and perform scene recognition; if the networking mode is WIFI and the positioning information is indoor residential scene information, only the rear camera is started; if the networking mode is data traffic and the positioning information is indoor residential scene information, a networking mode modification request is generated and only the rear camera is started.

[0064] The smartphone is also used to periodically obtain positioning information and networking information when the user watches popular science content, access the gyroscope of the smartphone to obtain the posture information of the smartphone, and when the posture information of the smartphone is horizontal, collect the light intensity information and light spectrum information of the ambient light through the front camera, and then determine if the light intensity information is higher than a preset threshold, and the spectrum intensity of the spectrum information biased towards ultraviolet rays is higher than the preset threshold, then define the positioning information as outdoor scene information; if the light intensity information is lower than the preset threshold, and the spectrum intensity of the spectrum information biased towards ultraviolet rays is lower than the preset threshold, then define the positioning information as indoor work scene information. The outdoor scene information includes outdoor planting scenes and outdoor breeding scenes; the indoor work scene information includes indoor cultivation scenes and indoor planting scenes. The environment recognition module is also used to identify markers in the environment picture according to the differences between indoor living scene information, outdoor planting scenes, outdoor breeding scenes, indoor cultivation scenes and indoor planting scenes, and play fixed video clips according to the environment pictures collected by the front camera and / or the rear camera, and play fixed video clips after identifying the markers.

[0065] Specific use: When the smartphone is in charging state, it continuously obtains information about the networking mode and positioning information. For example, set the preset duration to 30 minutes. If within these 30 minutes, the smartphone detects that the networking mode is WIFI, and the positioning information (such as longitude and latitude coordinates) always remains unchanged within a certain range, such as within a circular area with a radius of 50 meters, then the system defines this positioning information as indoor residential scene information. This is because in a home environment, when the mobile phone is in a charging state and connected to WIFI, the position is usually relatively fixed. Based on such a stable state and the common home WIFI coverage range, the above parameters are set, which can more accurately determine that the user is in an indoor living environment. If the positioning information is not within the area corresponding to the determined indoor residential scene information, it is determined to be non-indoor residential scene information.

[0066] When the user is watching popular science content, the smartphone periodically (for example, every 5 minutes) obtains the positioning information and networking information of the GPS module or Beidou module, and accesses the gyroscope to obtain the posture information. When the posture information of the smartphone is detected to be horizontal, the light intensity information and light spectrum information of the ambient light are collected through the front camera. The preset threshold of the light intensity information is set to 500 lux, and the preset threshold of the spectrum intensity biased towards ultraviolet rays in the light spectrum information is set to 10 microwatts / square centimeter nanometers. If the light intensity information is higher than 500 lux, and the spectrum intensity biased towards ultraviolet rays is higher than 10 microwatts / square centimeter nanometers, then the positioning information is defined as outdoor scene information. For example, in a sunny outdoor planting area, the light intensity generally exceeds 500 lux, and because the sunlight contains ultraviolet components, its ultraviolet spectrum intensity is also likely to be higher than the set threshold. On the contrary, if the light intensity information is lower than 500 lux, and the spectrum intensity biased towards ultraviolet rays is lower than 10 microwatts / square centimeter nanometers, it is determined to be indoor work scene information. For example, indoor cultivation rooms usually use artificial lighting, and the light intensity and ultraviolet spectrum intensity are relatively low, which meets this judgment condition.

[0067] When the posture information of the smartphone changes from horizontal to vertical and lasts for more than a preset time (such as 3 seconds), if the networking mode is data traffic and the positioning information is non-indoor residential scene information, the system will simultaneously start the front and rear cameras to collect environmental images and perform scene recognition. For example, when a user uses a mobile phone in an outdoor farmland, and the phone posture changes from a horizontal state of viewing information to a vertical state to prepare for shooting or viewing the surrounding environment, the front and rear cameras are turned on at the same time because it is in a non-indoor residential scene and uses data traffic. The front camera can capture environmental information from a wider perspective in front of the user, and the rear camera can capture details of specific targets or areas. The images collected by the two are used together by the environmental recognition module to accurately identify the scene, such as distinguishing between outdoor planting scenes and outdoor breeding scenes.

[0068] If the network connection mode is WIFI and the positioning information is indoor living scene information, only the rear camera is activated. For example, when a user is watching agricultural science content via WIFI in his living room and wants to identify and understand indoor flowers and plants, only the rear camera is activated to take pictures of the flowers. This can reduce the energy consumption and data processing burden caused by unnecessary activation of the front camera. At the same time, the rear camera is sufficient to meet the image acquisition needs of a limited range and specific objects indoors, so that the environmental recognition module can perform scene recognition and push scientific knowledge about related indoor planting scenes.

[0069] When the network connection mode is data traffic and the positioning information is indoor living scene information, the smartphone generates a network connection mode modification request and only activates the rear camera. For example, if the user is indoors but not connected to WIFI and uses data traffic to watch popular science content, the system will remind the user to switch to the WIFI network to save traffic costs, and only activate the rear camera for scene recognition, such as identifying indoor small agricultural cultivation equipment or indoor vegetables, and push corresponding indoor cultivation scene popular science knowledge to the user.

[0070] The environment recognition module uses the environment pictures collected by the front camera and / or the rear camera to identify the markers in the environment pictures according to the differences between the determined indoor living scene information, outdoor planting scenes, outdoor breeding scenes, indoor cultivation scenes and indoor planting scenes, and plays fixed video clips. For example, in the outdoor planting scene, if the irrigation facilities in the farmland are identified as a marker, the system will play relevant fixed video clips about the operation and maintenance of the irrigation system and the importance of irrigation technology in agricultural production; in the indoor cultivation scene, if the markers of specific plant varieties on the culture racks are identified, such as rare flower varieties, the system will play fixed video clips about the cultivation points, pest and disease control, and market value of the flower varieties, so as to provide users with more targeted and professional agricultural science knowledge.

[0071] In another embodiment, the user's interactive behaviors on different types of popular science content can be recorded in detail through a smartphone, including viewing time, number of repeated viewings, whether to collect, like or comment, etc. For example, if a user has watched and collected video content about fruit tree pruning for many times, the system determines that the user has a high preference for fruit tree planting and management.

[0072] Combined with the user's viewing history in different scenarios (such as indoor living scenarios, outdoor planting scenarios, outdoor breeding scenarios, indoor cultivation scenarios, etc.), analyze the user's inclination towards various popular science content in specific scenarios. For example, users often watch content about vegetable pest control in outdoor planting scenarios, while they are more concerned about flower breeding techniques in indoor cultivation scenarios.

[0073] The system comprehensively judges the user's immediate needs and possible concerns based on the current scene determined by the environment recognition module (such as the scene determined by the various information collection and analysis of the smartphone mentioned above), the user's current posture information (horizontal or vertical), whether the phone is charging, etc. For example, when the user is in an outdoor farming scene and the phone is in a vertical posture, the system will prioritize the popular science content that is closely related to the current farming scene and is operational, such as water quality monitoring and regulation methods for farming ponds, because the user may be more inclined to directly learn about farming practices in this scene and posture.

[0074] Considering the time factor, different types of popular science content are pushed at different time periods. For example, in the early morning, for users in outdoor planting scenes, content about crop morning dew management and suitable farming operation arrangements for the day are pushed; in the evening, for users in indoor living scenes, some cutting-edge agricultural technology information or agricultural culture popular science content is pushed to facilitate users to browse and learn during their rest time.

[0075] Integrate multi-source data from the image acquisition module (image information collected by the camera), the voice recognition module (user voice commands and related dialogue information), the environment recognition module (scene information), and user preferences and historical behavior records. For example, if a user asks about the fertilization techniques of a certain plant through voice in an indoor cultivation scene, the system will not only push relevant fertilization science content, but also push comprehensive maintenance science knowledge based on the characteristics of the plant's indoor cultivation, such as light, temperature, humidity requirements, etc.

[0076] Based on machine learning algorithms, multi-source data is deeply analyzed and modeled to achieve personalized popular science content recommendations. The system continuously trains the model based on various characteristic data of users, so that the recommended content can be dynamically adjusted as the user's usage and environmental changes. For example, for a user who has just entered the field of agricultural planting and often engages in outdoor planting activities, the system may first push some basic introductory knowledge on crop planting. As the user's attention to certain specific crops (such as tomato planting) increases, it will gradually push more in-depth and professional popular science content such as tomato planting technology, variety selection and breeding, and market sales.

[0077] After the popular science content is pushed, user feedback is collected in real time, such as whether the user skips the pushed content, satisfaction evaluation after watching (through the expression recognition module or manual evaluation by the user), etc. For example, if the user frequently skips a certain type of popular science content push, the system will re-evaluate the user's demand for this type of content and adjust the push strategy to reduce the push frequency of this type of content or optimize the content format.

[0078] Based on user feedback and new data accumulation, we regularly update and optimize the popular science content library and recommendation algorithm model. For example, when we find that users are paying more attention to a new agricultural technology (such as smart agricultural irrigation system), we will promptly update the popular science information about the technology in the content library and adjust the weight of the recommendation model to make it easier for relevant content to be pushed to interested users, thereby ensuring that the popular science content always meets user needs and agricultural technology development trends.

[0079] By combining the current development trends of smart phones, users can carry their smart phones to agricultural scenes of their interest, and the mobile terminals will interact with the agricultural scenes and users to customize the output of popular science content, greatly improving the user experience.

[0080] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the invention.

Claims

1. The agricultural science knowledge base interactive system based on intelligent terminals is characterized by: include: Display module: used to display popular science content; Image acquisition module: used to collect user facial expressions and take images; Expression recognition module: used to identify the user's satisfaction with watching popular science content through expression recognition technology; Jump module: The popular science content is a video collection composed of several video clips. The video clips are jumped through jump screens. The video clips are all bound with preference classification data. The jump module is used to increase the number of preference classification video clips with high satisfaction and reduce the number of preference classification video clips with low satisfaction through jump screens when users watch the popular science content according to the satisfaction identified by the expression recognition module; The jump module is also used to speed up the playing speed of the video clip until the jump screen when the satisfaction level recognized by the expression recognition module is low.

2. The agricultural science knowledge base interactive system based on intelligent terminal according to claim 1 is characterized in that: It also includes an environment recognition module, which is used to recognize the scene in the image captured by the image acquisition module, so as to control the display module to display the popular science content related to the scene; the environment recognition module is also used to analyze the user's position in the scene based on the scene in the captured image, make detailed judgments on the popular science content according to the position, and control the display module to display the popular science content related to the position of the scene; it also includes a voice recognition module, which is used to recognize the user's voice commands and adjust the display of the popular science content according to the voice commands.

3. The agricultural science knowledge base interactive system based on intelligent terminal according to claim 2 is characterized in that: It also includes a mobile terminal, which is equipped with a touch screen, front and rear cameras, a microphone and a speaker; the touch screen is an anti-peeping screen, which is used to restrict the user within the range of facial expression collection, and the front and rear cameras include a front camera and a rear camera.

4. The agricultural science knowledge base interactive system based on intelligent terminal according to claim 3 is characterized in that: The environment recognition module collects environment pictures through the front and rear cameras of the mobile terminal for scene recognition. The specific logic is as follows: first, the collected pictures are preprocessed, and then the convolutional neural network in the deep learning algorithm is used to extract feature information of the preprocessed images. The convolutional neural network automatically learns feature information of different levels in the image through multiple convolutional layers and pooling layers, and matches and compares the extracted feature information with the pre-trained agricultural scene feature database. By calculating the similarity between the feature information, when the similarity exceeds the set threshold, the scene to which the current captured picture belongs can be determined.

5. The agricultural science knowledge base interactive system based on intelligent terminal according to claim 4 is characterized in that: It also includes a video clip repair module. The preference classification data includes the agricultural classification, location, work stage, video length, names of objects and organisms in the screen, and appearance time of popular science content. The video clip repair module is used to collect the overall satisfaction rating of each video clip, extract the preference classification data of video clips with low satisfaction, and delete the appearance time of objects or organisms with common characteristics based on the common characteristics of video clips with low satisfaction.

6. The agricultural science knowledge base interactive system based on intelligent terminal according to claim 5 is characterized in that: The mobile terminal is a smart phone, which is also used to obtain networking mode information and positioning information when in charging state. If the networking mode information is WIFI networking and the positioning information remains unchanged for a preset period of time, the positioning information is defined as indoor residential scene information. If the positioning information is not in the indoor residential scene information, it is non-indoor residential scene information. The smart phone is also used to periodically obtain the positioning information and networking mode information of the GPS module or Beidou module of the smart phone when the user watches popular science content, and access the gyroscope of the smart phone to obtain the posture information of the smart phone. When the posture information of the smart phone changes from horizontal to vertical and lasts for more than a preset time, if the networking mode is data traffic and the positioning information is non-indoor residential scene information, the front and rear cameras are simultaneously started to capture environmental pictures and perform scene recognition; if the networking mode is WIFI and the positioning information is indoor residential scene information, only the rear camera is started; if the networking mode is data traffic and the positioning information is indoor residential scene information, a networking mode modification request is generated and only the rear camera is started.

7. The agricultural science knowledge base interactive system based on intelligent terminal according to claim 6 is characterized in that: The smartphone is also used to periodically obtain positioning information and networking information when the user is watching popular science content, access the gyroscope of the smartphone to obtain the posture information of the smartphone, and when the posture information of the smartphone is horizontal, collect the light intensity information and light spectrum information of the ambient light through the front camera, and then determine if the light intensity information is higher than a preset threshold, and the spectrum intensity of the spectrum information biased towards ultraviolet rays is higher than the preset threshold, then define the positioning information as outdoor scene information; If the light intensity information is lower than a preset threshold, and the spectrum intensity of the ultraviolet light in the light spectrum information is lower than the preset threshold, the positioning information is defined as indoor work scene information.

8. The agricultural science knowledge base interactive system based on intelligent terminal according to claim 7 is characterized in that: The outdoor scene information includes outdoor planting scenes and outdoor breeding scenes; the indoor work scene information includes indoor cultivation scenes and indoor planting scenes.

9. The agricultural science knowledge base interactive system based on intelligent terminal according to claim 8 is characterized in that: The environment recognition module is also used to identify landmarks in the environment pictures captured by the front camera and / or the rear camera based on the differences between indoor living scene information, outdoor planting scenes, outdoor breeding scenes, indoor cultivation scenes and indoor planting scenes, and play fixed video clips when the landmarks are identified.

10. A method for interacting with an agricultural science popularization knowledge base based on a smart terminal, a method for interacting with an agricultural science popularization knowledge base based on a smart terminal of claim 9, characterized in that: include: Step 1: The user carries the mobile terminal to the agricultural site; Step 2: Users take photos of agricultural content they are interested in; Step 3: The user adjusts the head position to watch the popular science content played on the touch screen; Step 4: The mobile terminal captures the user's facial expression to analyze the user's viewing satisfaction, and adjusts the popular science content played through the jump module.

Images