Tea leaf memory impurity removing system

Abstract

The present application relates to the technical field of tea impurity removal, and particularly relates to a tea memory impurity removal system. The technical scheme comprises a tea conveying system, an identification memory system and an intelligent picking system; the tea conveying system, the identification memory system and the intelligent picking system are sequentially connected to form a closed loop for transmission, identification and removal of tea and impurities; the identification memory system comprises an intelligent identification module, the intelligent identification module comprises a feature identification unit and an intelligent frame selection unit; the identification memory system further comprises a manual identification module, the manual identification module comprises a manual calibration unit and a summary screening unit, and the summary screening unit is used for summarizing all impurity characteristic values identified in the intelligent identification module and the manual identification module. The present application has the memory impurity removal function, reduces the tea impurity identification time, improves the impurity identification precision and efficiency, and is suitable for efficient nondestructive impurity removal treatment of tea.

Description

Technical Field

[0001] This invention relates to the field of tea impurity removal technology, and in particular to a tea memory impurity removal system. Background Technology

[0002] Impurity removal is a crucial step in tea production. During the harvesting process, tea leaves often become contaminated with impurities such as sand, grass, sticks, and metal particles. Impurity removal is essential to ensure the quality of the tea. Generally, impurities are removed through manual or machine screening. Manual screening is one of the most traditional and commonly used methods for removing impurities from tea leaves. Workers use sieves or their hands to pick out impurities, separating them from the tea leaves. This method requires a significant investment of manpower and time and is relatively inefficient. Machine screening can separate tea leaves from impurities based on characteristics such as size, weight, and shape. However, this method has the drawback of easily damaging the impurities themselves in practical applications.

[0003] Patent CN201910763762.X discloses a "Tea Impurity Removal Production Line and Method," which addresses the shortcomings of manual and machine impurity removal by utilizing a picking robot to accurately and quickly remove impurities from flat-lay conveyed tea leaves, achieving efficient impurity removal. However, this method requires photographing and manually marking each batch of tea leaves before the picking robot can remove impurities from their marked locations. It cannot memorize and store the characteristics of impurities, nor can it intelligently and automatically mark and remove impurities from continuously conveyed tea leaves. Therefore, the accuracy and efficiency of impurity identification need improvement. In view of this, this invention proposes a tea impurity removal system capable of intelligently removing impurities by accumulating and updating its memory. Summary of the Invention

[0004] The purpose of this invention is to address the problems existing in the background technology by proposing a tea impurity removal system that can intelligently remove impurities by accumulating and updating the memory of impurities.

[0005] The technical solution of the present invention: a tea memory and impurity removal system, comprising a tea conveying system, an identification and memory system, and an intelligent sorting system, wherein the tea conveying system is used for flattening and conveying tea leaves;

[0006] The tea conveying system, the identification and memory system, and the intelligent sorting system are connected in sequence to form a closed loop for the transmission, identification, and removal of tea and impurities.

[0007] The identification and memory system includes an intelligent identification module, which includes a feature identification unit and an intelligent box selection unit, used to intelligently identify and mark the box selection of impurities in tea leaves;

[0008] The identification and memory system also includes a manual identification module, which includes a manual calibration unit and a summary and filtering unit. The manual calibration unit is used for manual participation in the identification and calibration of impurities and the recording of feature values. The summary and filtering unit is used to summarize all impurity feature values ​​identified in the intelligent identification module and the manual identification module, and delete duplicate feature value data.

[0009] The recognition memory system also includes a memory storage module for storing feature values. The memory storage module includes a feature iteration unit and an update storage unit. The update storage unit is communicatively connected to the intelligent recognition module to enrich the intelligent recognition feature value comparison data.

[0010] An image compression unit is provided between the intelligent recognition module, the manual recognition module, and the memory storage module. The image compression unit is used to ensure the encryption and accuracy of image data transmission.

[0011] Preferably, the tea conveying system specifically includes a tea flattening module and an automatic conveying module. The tea flattening module is used to flatten and lay the tea leaves and impurities contained therein on the electric conveyor belt in the automatic conveying module for conveying.

[0012] The intelligent sorting system includes an intelligent control module and a sorting robot module. The intelligent control module is used to receive the identified feature value location information and start the sorting robot module to pick up and remove impurities.

[0013] Preferably, the intelligent recognition module further includes a feature definition unit, an image acquisition unit, and an image transmission unit;

[0014] The impurity feature values ​​defined in the feature definition unit include, but are not limited to, the grayscale, shape, and edge-reflection feature values ​​of the impurities;

[0015] The feature recognition unit is used to quickly identify the features of the tea leaves and impurities mixed in with them, which are transported by the automatic transmission module.

[0016] The intelligent selection unit is used to locate and frame the impurities identified by the feature recognition unit. The framing rule is the outer edge line of the identified feature value.

[0017] The image transmission unit is used to transmit the identified impurity feature value location and image data to the manual recognition module.

[0018] Preferably, the image acquisition unit further includes a camera element subunit and an illumination element subunit. The camera element subunit includes multiple cameras at different angles located above the conveyor belt of the automatic transmission module, used to fully acquire images of the shape of the transmitted tea leaves.

[0019] The lighting element subunit is used to compensate for the lighting emitted by the camera in the imaging element subunit.

[0020] Preferably, the intelligent selection unit includes a border processing subunit and a selection confirmation subunit. The border processing subunit is used to process the clarity of the border of the impurity marked by the intelligent selection unit, so that the edge between the impurity and the tea leaves is clear.

[0021] The box selection confirmation subunit is used to reconfirm the integrity of the border marked by the border processing subunit.

[0022] Preferably, the manual recognition module further includes an image display unit, a magnified display unit, and an image transmission unit;

[0023] The image display unit is used to receive encrypted image data transmitted by the image compression unit and display the image on the display screen;

[0024] The magnified display unit is used to magnify the content of the border marked on the display screen, which is a hand-controlled touch screen;

[0025] The image transmission unit 2 has the same function as the image transmission unit 1, both being used for image data transmission.

[0026] Preferably, the manual calibration unit is used in conjunction with the magnification display unit to manually operate and accurately calibrate the magnified impurity content, and also has the function of checking the content at any position on the image data and supplementing the calibration border.

[0027] Preferably, the aggregation and filtering unit is used to receive impurity location data determined by the intelligent recognition module and the manual recognition module, and to perform aggregation and filtering, deleting duplicate feature values.

[0028] Preferably, the feature iteration unit is used to summarize the feature value data of impurities, periodically iterate the feature value data, and improve the feature value data;

[0029] The update storage unit is used to delete duplicate feature value data that has been repeated more than once, and to update the stored data packet.

[0030] Compared with the prior art, the present invention has the following beneficial technical effects:

[0031] This invention uses a tea conveying system to flatten the tea leaves and then transport them via a conveyor belt, ensuring that the tea leaves are fully displayed under the image acquisition unit. It can accurately identify the features of the transported tea leaves and use an intelligent picking system to intelligently control a picking robot to remove the identified impurities, thereby achieving the purpose of precise picking and avoiding damage to the tea leaves, thus improving the accuracy of tea impurity removal and selection.

[0032] Furthermore, by combining the intelligent identification module and the manual identification module, the quality of tea can be controlled in two ways. It also has a memory storage module to store the feature value definition of each intelligently identified impurity and the feature value definition of manually identified impurities, which accelerates the efficiency of impurity identification in tea, avoids repeated feature value processing steps, and improves the efficiency of tea impurity removal. It is suitable for large-scale and efficient enterprises to remove impurities from tea.

[0033] In summary, this invention has a memory-based impurity removal function, which reduces the time required to identify tea impurities, improves the accuracy and efficiency of impurity identification, and is suitable for efficient non-destructive impurity removal processing of tea. Attached Figure Description

[0034] Figure 1 A schematic diagram of the principle of this invention is provided;

[0035] Figure 2 for Figure 1 Block diagram of the principle of the recognition and memory system;

[0036] Figure 3 This is a flowchart of the present invention.

[0037] Figure label:

[0038] 10. Tea conveying system; 101. Tea flattening module; 102. Automatic transfer module;

[0039] Recognition and memory system;

[0040] 201. Intelligent recognition module;

[0041] 2011, Feature Definition Unit;

[0042] 2012, Image Acquisition Unit; 20121, Camera Element Subunit; 20122, Illumination Element Subunit;

[0043] 2013, Feature Recognition Unit;

[0044] 2014, Intelligent Selection Unit; 20141, Border Processing Subunit; 20142, Selection Confirmation Subunit;

[0045] 2015, Image Transmission Unit 1;

[0046] 202. Manual recognition module; 2021. Image display unit; 2022. Magnified display unit;

[0047] 2023, Manual calibration unit;

[0048] 2024, Summary and Filtering Unit;

[0049] 2025, Image Transmission Unit 2;

[0050] 203. Memory storage module; 2031. Feature iteration unit; 2032. Update storage unit;

[0051] 204. Image compression unit;

[0052] 30. Intelligent picking system; 301. Intelligent control module; 302. Picking robot module. Detailed Implementation

[0053] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0054] Example

[0055] like Figure 1-3 As shown, the present invention proposes a tea memory and impurity removal system, including a tea conveying system 10, an identification and memory system 20, and an intelligent sorting system 30. The intelligent sorting system 30 includes an intelligent control module 301 and a sorting robot module 302. The intelligent control module 301 is used to receive the identified feature value position information and start the sorting robot module 302 to pick up and remove impurities. The sorting robot module 302 adopts an existing sorting robot and accurately executes the impurity position control command of the intelligent control module 301 to achieve accurate removal of impurities and avoid damage to the tea.

[0056] Specifically, the tea conveying system 10 is used for flattening and conveying tea leaves. The tea conveying system 10 specifically includes a tea flattening module 101 and an automatic transmission module 102. The tea flattening module 101 is used to flatten and lay the tea leaves and impurities contained therein on the electric conveyor belt in the automatic transmission module 102 for transmission.

[0057] The tea conveying system 10, the identification and memory system 20, and the intelligent sorting system 30 are connected in sequence to form a closed loop for the transmission, identification, and removal of tea and impurities, thereby realizing the closed-loop operation of the entire impurity removal system.

[0058] Furthermore, the recognition and memory system 20 includes an intelligent recognition module 201, which includes a feature recognition unit 2013 and an intelligent frame selection unit 2014, used to intelligently identify and calibrate frame selection of impurities in tea leaves; the intelligent recognition module 201 also includes a feature definition unit 2011, an image acquisition unit 2012, and an image transmission unit 2015; the impurity feature values ​​defined in the feature definition unit 2011 include, but are not limited to, the grayscale, shape, and angular-reflective feature values ​​of the impurities; the image acquisition unit 2012 also includes a camera element subunit 20121 and an illumination element subunit 20122, the camera element subunit 20121 includes multiple sets of cameras at different angles located above the electric conveyor belt of the automatic transmission module 102, used to fully acquire images of the shape of the transmitted tea leaves, ensuring that images of hidden impurities in the tea leaves are fully acquired, and the illumination element subunit 20122 is used to compensate for the impurities captured by the camera element subunit 20121. The camera captures light, improving shooting brightness and adjusting brightness to adapt to different working environments at different times; the feature recognition unit 2013 is used to quickly identify the features of the tea leaves and impurities transmitted by the automatic transmission module 102; the intelligent frame selection unit 2014 is used to locate and frame the impurities identified by the feature recognition unit 2013, and the frame selection rule is the outer edge line of the identified feature value, and a border is drawn for objects that meet any feature value; the intelligent frame selection unit 2014 includes a border processing subunit 20141 and a frame selection confirmation subunit 20142. The border processing subunit 20141 is used to process the clarity of the impurity border marked by the intelligent frame selection unit 2014, so that the edge between the impurity and the tea leaves is clear; the frame selection confirmation subunit 20142 is used to reconfirm the integrity of the border marked by the border processing subunit 20141; the image transmission unit 2015 is used to transmit the identified and confirmed impurity position and image data to the manual recognition module 202.

[0059] Furthermore, the recognition and memory system 20 also includes a manual recognition module 202, which includes a manual calibration unit 2023 and a summary filtering unit 2024. The manual calibration unit 2023 is used for manual identification and calibration of impurities and recording of feature values. The summary filtering unit 2024 is used to summarize all impurity feature values ​​identified in the intelligent recognition module 201 and the manual recognition module 202, and delete duplicate feature value data to form a feature value summary file for quick retrieval of image comparison data in the next step. The manual recognition module 202 also includes an image display unit 2021, a magnified display unit 2022, and an image transmission unit 2025. The image display unit 2021 is used to receive encrypted image data transmitted by the image compression unit 204 and display it. The screen displays images; the magnification display unit 2022 is used to magnify the content of the border marked on the display screen, which is a hand-controlled touch screen; the image transmission unit 2025 has the same function as the image transmission unit 2015, both used for image data transmission; the manual calibration unit 2023 is used in conjunction with the magnification display unit 2022 to manually operate and accurately calibrate the magnified impurity image content, making the impurity position calibration more accurate and avoiding damage caused by accidental contact of the tea leaves by the picking robot. It also has the function of checking the content at any position on the image data and supplementing the calibration border; the summary and filtering unit 2024 is used to receive the impurity position data judged by the intelligent recognition module 201 and the manual recognition module 202, and to summarize and filter it. For duplicate feature values, it saves one and deletes the redundant data.

[0060] In this embodiment, the identification memory system 20 further includes a memory storage module 203 for storing feature values. The memory storage module 203 includes a feature iteration unit 2031 and an update storage unit 2032. The update storage unit 2032 is communicatively connected to the intelligent identification module 201 and is used to enrich the comparison data of impurity feature values ​​for intelligent identification. The feature iteration unit 2031 is used to summarize the feature value data of impurities, periodically iterate the feature value data, and improve the feature value data. The update storage unit 2032 is used to delete duplicate feature value data more than twice and update the stored data packet.

[0061] An image compression unit 204 is provided between the intelligent recognition module 201, the manual recognition module 202, and the memory storage module 203. The image compression unit 204 is used to ensure the encryption and accuracy of image data transmission, and to ensure low-loss or even lossless transmission of images.

[0062] In this embodiment, the tea leaves and impurities are simultaneously transported by the tea leaf conveying system 10. The tea leaf flattening module 101 and the automatic conveying module 102 enable the tea leaves to be automatically conveyed after being flattened, so as to ensure that the tea leaves are laid on the electric conveyor belt in a single layer as much as possible, thereby ensuring the accuracy of subsequent image acquisition and meeting the requirements for obtaining sufficient images of impurities.

[0063] Secondly, the recognition and memory system 20 identifies and memorizes the collected images of tea leaves. The intelligent recognition module 201 performs initial identification and calibration of impurity features in the collected images of tea leaves and impurities. Combined with the image transmission unit 2015, the collected images containing impurities are transmitted to the manual recognition module 202 for secondary manual judgment of feature values. The image display unit 2021 and the magnification display unit 2022 magnify and display the images. The manual calibration unit 2023 is used to manually calibrate the position of impurities in the images and also has the function of adjusting the border, making the calibration of impurities more accurate and providing precise positioning for the subsequent impurity removal by the sorting robot. The summary and screening unit 2024 summarizes and filters all feature value data and calibrated impurity data. The memory storage module 203 stores the feature value data, which can be quickly retrieved during the collection of tea impurities later, improving the efficiency and accuracy of feature value calibration and achieving efficient impurity removal.

[0064] Finally, the intelligent picking system 30 is used to start the picking robot arm and accurately remove impurities from the image that have been confirmed by both intelligent and manual methods.

[0065] like Figure 3 As shown, the general process for removing impurities from tea leaves in this embodiment is as follows:

[0066] During tea transport, the tea image is acquired by the recognition and memory system 20. At the same time, the feature value data stored in the memory storage module 203 is compared. If one or more features that meet the impurity feature value data are detected in the acquired image, the intelligent picking system 30 is immediately executed to start the robotic arm to remove the impurities. If no impurities are detected in the acquired image or new, unsaved, blurry data appears, manual judgment is performed to achieve accurate manual selection.

[0067] The above specific embodiments are merely preferred embodiments of the present invention. Based on the technical solutions of the present invention and the relevant teachings of the above embodiments, those skilled in the art can make various alternative improvements and combinations to the above specific embodiments.

Claims

1. A tea memory and impurity removal system, comprising a tea conveying system (10), an identification and memory system (20), and an intelligent picking system (30), wherein the tea conveying system (10) is used for flattening and conveying tea leaves, characterized in that: The tea conveying system (10), the identification and memory system (20) and the intelligent sorting system (30) are connected in sequence to form a closed loop for the transmission, identification and removal of tea and impurities; The identification memory system (20) includes an intelligent identification module (201), which includes a feature identification unit (2013) and an intelligent box selection unit (2014) for intelligently identifying and calibrating the selection of impurities in tea leaves; The identification memory system (20) further includes a manual identification module (202), which includes a manual calibration unit (2023) and a summary filtering unit (2024). The manual calibration unit (2023) is used for manual participation in the identification and calibration of impurities and the recording of feature values. The summary filtering unit (2024) is used to summarize all impurity feature values ​​identified in the intelligent identification module (201) and the manual identification module (202) and delete duplicate feature value data. The recognition memory system (20) further includes a memory storage module (203) for storing feature values. The memory storage module (203) includes a feature iteration unit (2031) and an update storage unit (2032). The update storage unit (2032) is communicatively connected to the intelligent recognition module (201) to enrich the intelligent recognition feature value comparison data. An image compression unit (204) is provided between the intelligent recognition module (201), the manual recognition module (202), and the memory storage module (203). The image compression unit (204) is used to ensure the encryption and accuracy of image data transmission.

2. The tea memory and impurity removal system according to claim 1, characterized in that, The tea conveying system (10) specifically includes a tea flattening module (101) and an automatic transmission module (102). The tea flattening module (101) is used to flatten and lay the tea leaves and impurities contained therein on the electric conveyor belt in the automatic transmission module (102) for transmission. The intelligent picking system (30) includes an intelligent control module (301) and a picking robot module (302). The intelligent control module (301) is used to receive the identified feature value location information and start the picking robot module (302) to pick up and remove impurities.

3. The tea memory and impurity removal system according to claim 1, characterized in that, The intelligent recognition module (201) further includes a feature definition unit (2011), an image acquisition unit (2012), and an image transmission unit (2015); The impurity feature values ​​defined in the feature definition unit (2011) include, but are not limited to, the grayscale, shape, and edge-reflection feature values ​​of the impurities; The feature recognition unit (2013) is used to quickly identify the features of the tea leaves and impurities mixed in with them that are transported by the automatic transmission module (102). The intelligent selection unit (2014) is used to locate and frame the impurities identified by the feature recognition unit (2013), and the framing rule is the outer edge line of the identified feature value; The image transmission unit (2015) is used to transmit the identified impurity feature value location and image data to the manual recognition module (202).

4. The tea memory and impurity removal system according to claim 3, characterized in that, The image acquisition unit (2012) also includes a camera element subunit (20121) and an illumination element subunit (20122). The camera element subunit (20121) includes multiple cameras at different angles located above the conveyor belt of the automatic transmission module (102) for fully acquiring images of the shape of the tea leaves being transported. The lighting element subunit (20122) is used to compensate for the camera shooting light in the camera element subunit (20121).

5. The tea memory and impurity removal system according to claim 3, characterized in that, The intelligent selection unit (2014) includes a border processing subunit (20141) and a selection confirmation subunit (20142). The border processing subunit (20141) is used to process the clarity of the border of the impurities calibrated by the intelligent selection unit (2014), so that the edge between the impurities and the tea leaves is clear. The box selection confirmation subunit (20142) is used to reconfirm the integrity of the border marking by the border processing subunit (20141).

6. The tea memory and impurity removal system according to claim 5, characterized in that, The artificial recognition module (202) further includes an image display unit (2021), a magnified display unit (2022), and an image transmission unit (2025); The image display unit (2021) is used to receive encrypted image data transmitted by the image compression unit (204) and display the image on the display screen; The magnified display unit (2022) is used to magnify the content of the border marked on the display screen, which is a hand-controlled touch screen; The image transmission unit 2 (2025) has the same function as the image transmission unit 1 (2015), both of which are used for image data transmission.

7. The tea memory and impurity removal system according to claim 6, characterized in that, The manual calibration unit (2023) is used in conjunction with the magnified display unit (2022) to manually operate and accurately calibrate the magnified impurity content, and also has the function of checking the content at any position on the image data and supplementing the calibration border.

8. The tea memory and impurity removal system according to claim 1, characterized in that, The summary filtering unit (2024) is used to receive the impurity location data determined by the intelligent recognition module (201) and the manual recognition module (202), and to perform summary filtering, deleting duplicate feature values.

9. A tea memory and impurity removal system according to claim 1, characterized in that, The feature iteration unit (2031) is used to summarize the feature value data of impurities, periodically iterate the feature value data, and improve the feature value data. The update storage unit (2032) is used to delete duplicate feature value data that occurs more than twice and update the storage data packet.

Images