Tibetan medicine urine image acquisition device and quantitative identification method

Abstract

The invention discloses a Tibetan medicine urine image acquisition device and a quantitative recognition method, and relates to the technical field of medical image acquisition and processing.The Tibetan medicine urine image acquisition device is mainly structurally characterized in that an outer urine bottle is placed on a tray, the tray is rotatably arranged above a chassis, a rotary lifting rod is vertically and fixedly installed on the chassis, and an inner rod of the rotary lifting rod is slidably arranged in an outer rod in a sleeved mode; a horizontal rod is fixedly connected to the top of the inner rod, the inner urine bottle is rotatably connected to the horizontal rod through a lifting rope, an overlooking camera is arranged above the outer urine bottle, a side-looking camera is arranged on the side face of the outer urine bottle, a temperature sensor and a heating unit are fixedly installed on the top of the inner side of the upper cover, and the outer urine bottle and the inner urine bottle are both made of transparent materials. And a plurality of through holes are formed in the bottle body and the bottle bottom of the inner urine bottle. The temperature during urine image collection can be controlled and adjusted, floating skin, floating objects, suspended solids and sediments in urine can be filtered out, urine images at different stages are collected in multiple directions, the urine images are recognized and processed, and the effect of the Tibetan medicine three-time-nine diagnosis method can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image acquisition and processing, and in particular to a Tibetan medicine urine image acquisition device and a quantitative recognition method. Background Art

[0002] Tibetan medicine has a unique method of urinalysis called the Three Times and Nine Diagnosis Method, which diagnoses illnesses by observing urine's color, odor, foam, and sediment. The "Three Times" in the Three Times and Nine Diagnosis Method refer to the three stages of urine cooling: the hot period (hot urine), the warm period (heat dissipation period), and the cool period (cool urine). During the hot period, urine color is primarily observed, while during the warm period, the cracks, thickness, and location and morphology of urine substances are examined to diagnose illnesses.

[0003] This method, an intangible cultural heritage accumulated through long-term practice, holds significant significance for the development of modern medical diagnosis. However, because it relies on the physician's subjective sensory identification of urine characteristics, experience must be passed down orally, and accurate diagnosis requires extensive clinical experience, making it difficult to widely disseminate, learn, and use. In particular, the conventional Three-Time-Nine Diagnosis Method, which observes urine color during the warm urine phase and the cracks and thickness of the urine during the warm urine phase, as well as the location and morphology of urine substances, holds significant significance for the future of remote diagnosis and treatment in Tibetan medicine. Summary of the Invention

[0004] Based on the above problems, the first purpose of the present invention is to provide a Tibetan medicine urine image acquisition device that can control and adjust the temperature during urine image acquisition, can filter out floating skin, floating objects, suspended matter and sediment in urine, and can collect urine images at different stages from multiple angles.

[0005] The technical solution adopted by the present invention to achieve its first invention object is: a Tibetan medicine urine image acquisition device, including an outer urine bottle, an inner urine bottle, a DC motor, an image acquisition module, and a temperature control module, wherein:

[0006] The outer urine bottle is placed on a tray, which is rotatably arranged above the chassis. A DC motor is fixedly installed on the lower surface of the chassis, and the tray and the DC motor are in transmission connection. The DC motor can drive the tray, the outer urine bottle, and the inner urine bottle to rotate.

[0007] A rotating lifting rod is fixedly installed vertically on the chassis, the inner rod of the rotating lifting rod is slidably sleeved in the outer rod, the outer rod is installed with a fastening screw for locking the inner rod, the top of the inner rod is fixedly connected to the horizontal rod, and the urine inner bottle is rotatably connected to the horizontal rod through a hanging rope;

[0008] The outer diameter of the urine inner bottle is equal to the inner diameter of the urine outer bottle;

[0009] In this way, when the urine inner bottle is placed in the urine outer bottle, they can rotate together under the drive of the tray. By adjusting the rotating lifting rod, the urine inner bottle and the urine outer bottle can be separated to filter out the floating skin, floating objects, suspended matter and sediment in the urine.

[0010] The image acquisition module includes a top-view camera fixedly arranged above the urine outer bottle and a side-view camera fixedly arranged on the side of the urine outer bottle;

[0011] The temperature control module includes a temperature sensor and a heating unit fixedly mounted on the top inner side of the upper cover. The upper cover is placed on the chassis and houses the outer urine bottle, the inner urine bottle, and the image acquisition module. The temperature control module can heat the closed environment to ensure that the temperature of the urine is kept constant.

[0012] The outer urine bottle and the inner urine bottle are both made of transparent materials, and a plurality of through holes are provided on the body and the bottom of the inner urine bottle.

[0013] Furthermore, the specific transmission connection between the tray and the DC motor is as follows:

[0014] A vertical rotating shaft is fixedly connected to the bottom of the tray, and the rotating shaft is connected to a pair of first bearings and passes downward through the through hole on the chassis. The pair of first bearings are fixedly installed on the upper and lower sides of the chassis respectively. The lower end of the rotating shaft is fixedly connected to a first pulley, and the lower end of the vertical motor output shaft of the DC motor is fixedly connected to a second pulley, and a transmission belt is wrapped around the first pulley and the second pulley.

[0015] Furthermore, the specific manner in which the urine inner bottle is rotatably connected to the horizontal rod via the hanging rope is as follows:

[0016] A horizontally arranged second bearing is fixedly installed at the bottom of one end of the horizontal rod, and at least three hooks are fixedly installed on the inner ring wall of the second bearing. There are at least three hanging ropes, the top ends of the hanging ropes are hung on the hooks, and the bottom ends of the hanging ropes are fixedly connected to the top of the urine bottle.

[0017] Furthermore, the heating unit includes a heating control circuit and an electric heating device.

[0018] Furthermore, the device is also provided with a main control module, which includes:

[0019] Power supply unit, used to supply power to the main control module;

[0020] A core processor, configured to execute control logic for the image acquisition module, temperature control module, and DC motor;

[0021] A communication interface, comprising a first Bluetooth unit and a USB interface unit, wherein the first Bluetooth unit is used to communicate with the temperature control module, and the USB interface unit is used to connect to the top-view camera and the side-view camera;

[0022] a motor control unit, comprising a motor drive board for driving the DC motor;

[0023] Storage and expansion interface, including TF card unit and RJ45 unit. The TF card unit is used to realize local storage of calibration parameters, system logs and temporary image data, and the RJ45 unit is used to connect to the external network;

[0024] The heating unit in the temperature control module is further provided with a second Bluetooth unit connected to the heating control circuit, which is used to communicate with the first Bluetooth unit in the main control module.

[0025] In this way, the main control module can be used to centrally and uniformly control the heating unit for heating, control the overhead camera and the side view camera for shooting, and control the DC motor to drive the tray and the urine outer bottle to rotate, thereby improving the centralized control and automation working capabilities of the device and improving work efficiency.

[0026] Furthermore, the device is also provided with a host computer, which includes:

[0027] a storage unit, configured to store images captured by the top-view camera and the side-view camera and calculation results;

[0028] A computing unit, used for image recognition and calculation;

[0029] A display unit, used for displaying calculation results;

[0030] The host computer is connected to the USB interface unit of the communication interface in the main control module through a USB interface.

[0031] In this way, the urine images taken by the front view camera and the side view camera can be transferred to the host computer through the USB interface unit of the communication interface in the main control module, and the host computer can perform calculation and recognition processing and display the results.

[0032] The second purpose of the present invention is to provide a method for quantitative identification of urine in Tibetan medicine. The method can control and adjust the temperature during urine image collection, can filter out floating skin, floating objects, suspended matter and sediment in urine, can collect urine images at different stages from multiple angles, and can identify and process urine images, which is conducive to improving the effect of the Tibetan medicine three-time nine-diagnosis method.

[0033] The technical solution adopted by the present invention to achieve its second invention object is a method for quantitative identification of Tibetan medicine urine using the above-mentioned Tibetan medicine urine image acquisition device, comprising the following steps:

[0034] S1. Place the standard color module on a tray, then place the upper cover on the chassis. The main control module controls the top-view camera and the side-view camera to capture images of the standard color module and transmit them to the host computer. The host computer calculates the LAB measurement value of the standard color module and compares it with the true LAB value of the standard color module to obtain the color correction coefficient.

[0035] S2. Remove the upper cover, remove the standard color module, place the outer urine bottle on the tray, adjust the rotary lifting rod to place the inner urine bottle into the outer urine bottle, tighten the inner rod by tightening the screw, pour the urine into the inner urine bottle, and replace the upper cover on the bottom plate;

[0036] S3, the top-view camera captures the initial floating image of the urine, and the main control module drives the DC motor to rotate the tray, the outer urine bottle, and the inner urine bottle. The side-view camera captures the initial side view image of the urine and transmits it to the host computer;

[0037] S4. Adjust the urine temperature to a heat dissipation period by natural cooling, then use the top-view camera to capture a superficial image of the urine, and the side-view camera to capture a side view image of the urine, and transmit the image to the host computer;

[0038] S5. Calculate the edge features of the floating skin image taken in step S4 using the candy algorithm on the host computer, calculate the length value I of the floating skin crack edge excluding the outer edge, and then calculate the degree value P of the floating skin crack according to the formula I=-log2(1-P);

[0039] S6. Calculate the thickness edge features of the floating skin of the side view image taken in step S4 using the candy algorithm on the host computer. Calculate the thickness of any point on the floating skin edge according to the following formula:

[0040]

[0041] The average thickness of the floating skin is:

[0042]

[0043] Where n represents the total number of pixels from the left edge to the right edge of the floating skin in the side view image;

[0044] Under any horizontal coordinate on the image, the coordinates of the upper edge and the lower edge are (c, a) and (c, b) respectively;

[0045] The longitudinal field of view angle is β, the side view image resolution is M*N, where M is the horizontal resolution, N is the vertical resolution, and the distance between the side view camera and the center point of the urine outer bottle 1 is h;

[0046] S7, remove the upper cover, adjust the rotary lifting rod to separate the urine inner bottle and the urine outer bottle, re-place the upper cover on the chassis, heat the temperature inside the cover to the hot urine period through the temperature control module, capture the urine image through the side view camera and transmit it to the host computer, calculate the LAB value of the urine image on the host computer, and then calibrate the LAB value using the color correction coefficient obtained in step S1 to obtain the calibrated urine image LAB value, and finally calculate the distances S1, S2, and S3 between the calibrated urine image LAB value and the corresponding LAB standard values of urine green, urine yellow, and urine milky white, respectively;

[0047]

[0048] Where: (L1, A1, B1) is the LAB value of the calibrated urine image, (L2, A2, B2) is the LAB standard value of blue urine, (L3, A3, B3) is the LAB standard value of yellow urine, and (L4, A4, B4) is the LAB standard value of milky white urine.

[0049] Furthermore, after step S6 and before step S7, the quantitative index calculation of floating matter, suspended matter and sediment is performed, specifically:

[0050] The side-view camera records depth images of urine at different locations. Wavelet transform is used on the host computer to enhance and pre-process the depth images, calculate point cloud data, and reconstruct the entire urine after point cloud registration. The volume of the entire urine's three-dimensional features is calculated using voxel modeling as V1, the volume of floating matter as V2, the volume of suspended matter as V3, and the volume of sediment as V4. Quantitative indices are then calculated for each:

[0051]

[0052] Where: F, Z and C are the quantitative indices of floating matter, suspended matter and sediment respectively.

[0053] Furthermore, after the quantitative index calculation of the floating matter, suspended matter and sediment and before step S7, the recognition model of the floating matter, suspended matter and sediment and the urine image recognition are constructed on the host computer, specifically:

[0054] Floating matter, suspended matter, and sediment in multiple urine images are manually labeled, and urine substances are divided into three forms: floating matter, suspended matter, and sediment. These three forms are used as training and test sets, and a two-dimensional deep learning algorithm is used to construct a first recognition model based on deep CNN image classification for the three forms. The urine substances in multiple urine images are manually labeled and divided into six forms: slender like cow hair, shaped like a cotton ball, dense in the center and sparse around, scattered like gravel, floating like clouds, and like needle tips. These six forms are used as training and test sets, and a two-dimensional deep learning algorithm is used to construct a second recognition model based on deep CNN image classification for the six forms. Finally, the urine image to be identified is input, and the first recognition model is used to identify the urine substance as one of the floating matter, suspended matter, and sediment. Then, the second recognition model is used to identify the urine substance as one of the forms: slender like cow hair, shaped like a cotton ball, dense in the center and sparse around, scattered like gravel, floating like clouds, and like needle tips.

[0055] The beneficial effects of the present invention are:

[0056] The present invention can collect urine images at different stages, and can quantitatively calculate and identify urine-related parameter indicators, especially urine color, cracks of pee, thickness, and the morphology of substances in urine, which is beneficial to improving the effect of the Tibetan medicine three-time nine-diagnosis method. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a schematic structural diagram of a Tibetan medicine urine image acquisition device according to an embodiment of the present invention;

[0058] Figure 2 This is a schematic diagram of the device in use when photographing a standard color module according to an embodiment of the present invention;

[0059] Figure 3 This is a schematic diagram of the device in use before filtering urine according to an embodiment of the present invention;

[0060] Figure 4 This is a schematic diagram of the overall appearance of an embodiment of the present invention with the upper cover placed backward;

[0061] Figure 5 This is a top view schematic diagram of the connection structure between the urine inner bottle and the second bearing according to the present invention;

[0062] Figure 6 Schematic diagram of electrical connections according to an embodiment of the present invention.

[0063] To understand the internal structure, Figures 1 to 3 All figures show the upper cover raised. DETAILED DESCRIPTION

[0064] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0065] Figures 1 to 6 A specific embodiment of the Tibetan medicine urine image acquisition device of the present invention is shown, which includes a urine outer bottle 1, a urine inner bottle 2, a DC motor 3, an image acquisition module, a temperature control module, a main control module, and a host computer, wherein:

[0066] The urine outer bottle 1 is placed on the tray 4. The tray 4 is provided with a raised outer edge to confine the urine outer bottle 1 therein so that the urine outer bottle 1 does not slide out of the tray 4 when rotating. The tray 4 is rotatably arranged above the chassis 5. The DC motor 3 is fixedly mounted on the lower surface of the chassis 5. The tray 4 and the DC motor 3 are transmission-connected. The specific transmission connection method is as follows:

[0067] The bottom of the tray 2 is keyed to a vertical rotating shaft 12, which is connected to a pair of first bearings 25 and passes downward through the through-hole on the chassis 5. The pair of first bearings 25 are fixedly mounted on the upper and lower sides of the chassis 5 through flanged bearing seats. The lower end of the rotating shaft 12 is keyed to a first pulley 13, and the lower end of the vertical motor output shaft of the DC motor 3 is keyed to a second pulley 14. The transmission belt 15 is wrapped around the first pulley 13 and the second pulley 14.

[0068] The edge of the chassis 5 is fixedly mounted on a bracket to support the entire device;

[0069] A rotating lifting rod 6 is vertically fixedly installed on the right side of the urine outer bottle 1 on the chassis 5, and the inner rod 602 of the rotating lifting rod is slidably sleeved in the outer rod 601. A fastening screw 603 for locking the inner rod 602 is installed on the outer rod 601, and the fastening screw 603 adopts a handle screw. The top of the inner rod 602 is fixedly connected to the horizontal rod 604, and the urine inner bottle 2 is rotatably connected to the horizontal rod 604 through a hanging rope 26. The specific connection method is: a horizontally arranged second bearing 27 is fixedly installed on the bottom of one end of the horizontal rod 604 through a flanged bearing seat, and three hooks 28 are welded or bonded on the inner ring wall of the second bearing 27. There are three hanging ropes 26, and the top ends of the three hanging ropes 26 are hung on the three hooks 28 respectively, and the bottom ends of the three hanging ropes 26 are fixedly tied to three holes evenly distributed along the circumferential direction on the top of the bottle wall of the urine inner bottle 2;

[0070] The outer diameter of the urine inner bottle 2 is equal to the inner diameter of the urine outer bottle 1;

[0071] The image acquisition module includes a top-view camera 7 fixedly mounted above the urine outer bottle 1 and a side-view camera 8 fixedly mounted on the side of the urine outer bottle 1; the specific installation method is that an L-shaped rod 23 located on the left side of the urine outer bottle 1 is fixedly mounted on the chassis 5, the top-view camera 7 is fixedly mounted on the horizontal part of the L-shaped rod 23, and the side-view camera 8 is fixedly mounted on the vertical part of the L-shaped rod 23;

[0072] The temperature control module includes a temperature sensor 10 and a heating unit 11 fixedly bonded to the top inner side of the upper cover 9. The heating unit 11 includes a heating control circuit 1101 and a PTC heating device 1102. The heating control circuit 1101 uses a MOSFET switch to control the PTC heating device 1102, and cooperates with the temperature sensor 10 connected to the heating control circuit 1101 to achieve PID constant temperature regulation.

[0073] The upper cover 9 is placed on the bottom plate 5 and contains the urine outer bottle 1, the urine inner bottle 2, and the image acquisition module; a handle 24 is also fixedly connected to the top of the upper cover 9;

[0074] The outer urine bottle 1 and the inner urine bottle 2 are both made of glass, and a plurality of through holes 2a are provided on the body and bottom of the inner urine bottle 2.

[0075] In this embodiment, the upper surface of the chassis 5 and the inner side of the upper cover 9 are white;

[0076] The main control module includes:

[0077] The power supply unit 16 is used to power the main control module; the built-in lithium battery pack (backup power supply) and the external DC12V adapter support seamless switching;

[0078] The core processor 17 is used to execute the control logic of the image acquisition module, the temperature control module and the DC motor 3. The core processor 17 adopts a high-performance embedded single-board computer (Jetson Nano), supports multi-threaded processing and peripheral expansion, runs an embedded operating system, executes control logic, communication protocol analysis, task scheduling and data transfer;

[0079] The communication interface 18 includes a first Bluetooth unit 1801 and a USB interface unit 1802, wherein the first Bluetooth unit 1801 uses a low-power Bluetooth 5.0 chip (HC-05) for communicating with the temperature control module, and the USB interface unit 1802 is used to connect to the top-view camera 7 and the side-view camera 8;

[0080] The motor control unit 19 includes a motor drive board 1901 for driving the DC motor 3. The motor drive board 1901 controls the DC motor 3 via an H-bridge circuit (PWM speed regulation). This embodiment also integrates a photoelectric encoder 1902 for feedback to achieve closed-loop speed control.

[0081] Storage and expansion interface 20, including TF card unit 2001 and RJ452002 unit, wherein TF card unit 2001 is used to realize local storage of calibration parameters, system logs and temporary image data, and RJ45 unit 2002 is used to connect to the external network to realize optional network transmission to the host computer or cloud;

[0082] Each unit of the main control module of this embodiment is integrated and installed in the main control module box 22, and the main control module box 22 is fixed to the lower surface of the chassis 5 by screws;

[0083] The heating unit 11 in the temperature control module is further provided with a second Bluetooth unit 1103 connected to the heating control circuit 1101 for communicating with the first Bluetooth unit 1801 in the main control module;

[0084] The host computer 21 includes:

[0085] A storage unit 2101 is used to store images captured by the top-view camera 7 and the side-view camera 8 and calculation results;

[0086] A computing unit 2102, configured to recognize and compute images;

[0087] Display unit 2103, used to display calculation results;

[0088] The host computer 21 is connected to the USB interface unit 1802 of the communication interface 18 in the main control module via a USB interface.

[0089] The method for quantitative identification of Tibetan medicine urine using the above-mentioned collection device comprises the following steps:

[0090] S1, such as Figure 2 As shown, a standard color module 29 is placed on the tray 4. The shape and size of the standard color module 29 are the same as those of the urine outer bottle 1. Then, the upper cover 9 is placed on the chassis 5. The main control module controls the top view camera 7 and the side view camera 8 to capture the image of the standard color module 29 and transmit it to the host computer 21. The LAB measurement value of the standard color module 29 is calculated on the host computer 21 and compared with the LAB true value of the standard color module 29 to obtain the color correction coefficient.

[0091] S2, such as Figure 3 As shown, remove the upper cover 9, remove the standard color module 29, place the urine outer bottle 1 on the tray 4, adjust the rotary lifting rod 6 to place the urine inner bottle 2 into the urine outer bottle 1, tighten the fastening screw 603 to lock the inner rod 602, pour the urine into the urine inner bottle 2, and re-place the upper cover 9 on the bottom plate 5;

[0092] S3, the top-view camera 7 captures the initial floating image of the urine, and the main control module drives the DC motor 3 to rotate the tray 4, the urine outer bottle 1 and the urine inner bottle 2, and the side-view camera 8 captures the initial side-view image of the urine and transmits it to the host computer 21;

[0093] S4, adjust the urine temperature to the heat dissipation period by natural cooling method, then take the floating skin image of the urine by the top-view camera 7, take the side view image of the urine by the side-view camera 8, and transmit it to the host computer 21;

[0094] S5. Calculate the edge features of the floating skin image taken in step S4 using the candy algorithm on the host computer 21, calculate the length value I of the floating skin crack edge excluding the outer edge, and then calculate the degree value P of the floating skin crack according to the formula I=-log2(1-P);

[0095] The degree value P of the superficial cracks can be used to assist in the diagnosis of diseases. For example, if the P value is greater than a given threshold, it is determined that there is a mass (such as a nodule);

[0096] S6. On the host computer 21, the thickness edge feature of the floating skin of the side view image taken in step S4 is calculated using the candy algorithm. The thickness of any point on the floating skin edge is calculated according to the following formula:

[0097]

[0098] The average thickness of the floating skin is:

[0099]

[0100] Where n represents the total number of pixels from the left edge to the right edge of the floating skin in the side view image;

[0101] Under any horizontal coordinate on the image, the coordinates of the upper edge and the lower edge are (c, a) and (c, b) respectively;

[0102] The longitudinal field of view angle is β, the side view image resolution is M*N, where M is the horizontal resolution, N is the vertical resolution, and the distance between the side view camera 8 and the center point of the urine outer bottle (1) is h;

[0103] The average thickness D of the floating skin can be used to assist in the diagnosis of diseases. For example, if D exceeds a given threshold, it is judged as a heat syndrome; if D is less than a given threshold, it is judged as a cold syndrome.

[0104] Then the quantitative index of floating matter, suspended matter and sediment is calculated, specifically:

[0105] The side-view camera 8 records depth images of different positions of the urine. Specifically, when the urine outer bottle 1 and the urine inner bottle 2 rotate, the side-view camera 8 continuously captures multiple images, which are then synthesized into a depth image on the host computer 21. The depth image is then enhanced and pre-processed using wavelet transform on the host computer 21, and point cloud data is calculated. After point cloud registration, the entire urine is reconstructed, and the volume of the three-dimensional features of the entire urine is calculated using a voxel modeling method as V1, the volume of floating matter as V2, the volume of suspended matter as V3, and the volume of sediment as V4. Quantitative indices are then calculated for each of them:

[0106]

[0107] Where: F, Z and C are the quantitative indices of floating matter, suspended matter and sediment respectively;

[0108] Quantitative indices of floating, suspended, and sediment can be used to assist in diagnosing diseases, such as:

[0109] The urine is mainly composed of floating objects, that is, F is the largest among F, Z and C, indicating that the lesion is in the heart and lungs above the chest and abdomen;

[0110] The urine is mainly composed of suspended matter, that is, Z is the largest among F, Z, and C, indicating that the lesion is in the upper abdomen, namely the liver, gallbladder, pleural layer, etc.

[0111] The urine is mainly composed of sediment, that is, C is the largest among F, Z and C, indicating that the lesion is in the lower abdomen, that is, the large intestine, small intestine, bladder, genitals, etc.

[0112] Then, we build the recognition model of floating matter, suspended matter and sediment and perform urine image recognition, specifically:

[0113] On the host computer 21, the floating objects, suspended objects and sediments in multiple urine images are manually annotated, and the substances in the urine are divided into three forms: floating objects, suspended objects and sediments. The three forms are used as training sets and test sets, and a two-dimensional deep learning algorithm is used to construct a first recognition model based on deep CNN image classification for the three forms; the substances in the urine in the multiple urine images are manually annotated and divided into six forms: slender like cow hair, shaped like cotton balls, dense in the center and sparse around, scattered like gravel, floating like clouds, and thin like needle tips. The method comprises the following steps: first, a urine image to be identified is input, and the first recognition model is used to identify the substance in the urine as one of floating matter, suspended matter, and sediment; second, a urine image to be identified is input, and the first recognition model is used to identify the substance in the urine as one of floating matter, suspended matter, and sediment; and third, the method comprises the following steps: first, a urine image to be identified is input, and the first recognition model is used to identify the substance in the urine as one of floating matter, suspended matter, and sediment; and fourth, a urine image to be identified is input, and the first recognition model is used to identify the substance in the urine as one of slender matter like cow hair, shaped like a cotton ball, dense in the center and sparse around, dispersed like gravel, floating like clouds, and shaped like a needle tip;

[0114] The appearance of substances in urine can be used to help diagnose diseases, such as:

[0115] If it is as thin and long as cow hair, it is a "Long" disease; if it is shaped like a cotton ball, dense in the center and sparse around it, it is a "Chiba" disease; if it is scattered like sand, it is a kidney disease; if it is floating like clouds, it is a lung heat disease; if it is like a needle tip, it is a cold "Bacon" disease.

[0116] S7, such as Figure 1 As shown, the upper cover 9 is removed and the rotary lifting rod 6 is adjusted, specifically: loosen the fastening screw 603, manually pull up the inner rod 602 of the rotary lifting rod 6 to separate the urine inner bottle 2 and the urine outer bottle 1, filter out the floating skin, floating objects, suspended matter and sediment in the urine, and rotate the inner rod 602 to rotate the urine inner bottle 2 to the side, re-tighten the fastening screw 603 to lock the inner rod 602, and re-place the upper cover 9 on the chassis 5, and heat the temperature inside the cover to the hot urine period through the temperature control module, and capture the urine image in the urine outer bottle 1 through the side-view camera 8 and transmit it to the host computer 21, calculate the LAB value of the urine image on the host computer 21, and then calibrate the LAB value by the color correction coefficient obtained in step S1 to obtain the calibrated urine image LAB value, and finally calculate the distances S1, S2, and S3 between the calibrated urine image LAB value and the LAB standard values corresponding to urine green, urine yellow, and urine milky white, respectively;

[0117]

[0118] Where: (L1, A1, B1) is the LAB value of the calibrated urine image, (L2, A2, B2) is the LAB standard value of blue urine, (L3, A3, B3) is the LAB standard value of yellow urine, and (L4, A4, B4) is the LAB standard value of milky white urine.

[0119] The above S1, S2, and S3 values can be used to assist in diagnosing diseases, for example:

[0120] Among S1, S2, and S3, if S1 is the smallest, the urine is blue, which is Long-type disease; if S2 is the smallest, the urine is yellow, which is Chiba disease; if S3 is the smallest, the urine is milky white, which is Bacon disease.

[0121] The display unit 2103 of the host computer 21 displays the original state of the hot urine period, the calibrated urine color state, and the two-dimensional and three-dimensional states of urine skin, floating matter, suspended matter and sediment in the heat dissipation period; and displays the calculation and recognition results such as the skin crack degree value, skin thickness value, quantitative index of floating matter, suspended matter and sediment, calibrated urine color LAB value, determination of urine thickness, position and morphological characteristics of objects in urine, etc.

[0122] The above embodiments of the present invention are merely examples for illustrating the present invention and are not intended to limit the embodiments of the present invention. Those skilled in the art will appreciate that other variations and modifications can be made based on the above description. It is not possible to enumerate all embodiments here. Any obvious variations or modifications arising from the technical solutions of the present invention remain within the scope of protection of the present invention.

Claims

1. A Tibetan medicine urine image acquisition device, characterized in that: It comprises an outer urine bottle (1), an inner urine bottle (2), a DC motor (3), an image acquisition module, and a temperature control module, wherein: The urine outer bottle (1) is placed on a tray (4), the tray (4) is rotatably arranged above the chassis (5), the DC motor (3) is fixedly mounted on the lower surface of the chassis (5), and the tray (4) and the DC motor (3) are transmission-connected; A rotating lifting rod (6) is vertically fixedly mounted on the chassis (5); an inner rod (602) of the rotating lifting rod is slidably sleeved inside an outer rod (601); a fastening screw (603) for locking the inner rod (602) is mounted on the outer rod (601); a horizontal rod (604) is fixedly connected to the top of the inner rod (602); and the urine inner bottle (2) is rotatably connected to the horizontal rod (604) via a hanging rope (26); The outer diameter of the urine inner bottle (2) is equal to the inner diameter of the urine outer bottle (1); The image acquisition module comprises a top-view camera (7) fixedly arranged above the urine outer bottle (1) and a side-view camera (8) fixedly arranged on the side of the urine outer bottle (1); The temperature control module comprises a temperature sensor (10) and a heating unit (11) fixedly mounted on the inner top of the upper cover (9); the upper cover (9) is placed on the chassis (5) and accommodates the urine outer bottle (1), the urine inner bottle (2), and the image acquisition module therein; The outer urine bottle (1) and the inner urine bottle (2) are both made of transparent materials, and a plurality of through holes (2a) are provided on the body and bottom of the inner urine bottle (2).

2. The Tibetan medicine urine image acquisition device according to claim 1, characterized in that: The specific transmission connection between the tray (4) and the DC motor (3) is as follows: A vertical rotating shaft (12) is fixedly connected to the bottom of the tray (2), the rotating shaft (12) is connected to a pair of first bearings (25) and passes downward through a through hole on the chassis (5), the pair of first bearings (25) are fixedly installed on the upper and lower sides of the chassis (5), the lower end of the rotating shaft (12) is fixedly connected to a first pulley (13), the lower end of the vertical motor output shaft of the DC motor (3) is fixedly connected to a second pulley (14), and a transmission belt (15) is wrapped around the first pulley (13) and the second pulley (14).

3. The Tibetan medicine urine image acquisition device according to claim 1, characterized in that: The specific manner in which the urine inner bottle (2) is rotatably connected to the horizontal rod (604) via the hanging rope (26) is as follows: A horizontally arranged second bearing (27) is fixedly mounted on the bottom of one end of the horizontal rod (604), and at least three hooks (28) are fixedly mounted on the inner ring wall of the second bearing (27). There are at least three hanging ropes (26), the top ends of the hanging ropes (26) are hung on the hooks (28), and the bottom ends of the hanging ropes (26) are fixedly connected to the top of the urine inner bottle (2).

4. The Tibetan medicine urine image acquisition device according to claim 1, characterized in that: The heating unit (11) comprises a heating control circuit (1101) and an electric heating device (1102).

5. The Tibetan medicine urine image acquisition device according to claim 4, characterized in that: A main control module is also provided, and the main control module includes: A power supply unit (2), used for supplying power to the main control module; A core processor (17) for executing control logic for the image acquisition module, the temperature control module and the DC motor (3); a communication interface (18), comprising a first Bluetooth unit (1801) and a USB interface unit (1802), wherein the first Bluetooth unit (1801) is used to communicate with the temperature control module, and the USB interface unit (1802) is used to connect to the top-view camera (7) and the side-view camera (8); a motor control unit (19), comprising a motor drive board (1901) for driving the DC motor (3); A storage and expansion interface (20), comprising a TF card unit (2001) and an RJ45 (2002) unit, wherein the TF card unit (2001) is used to implement local storage of calibration parameters, system logs, and temporary image data, and the RJ45 unit (2002) is used to connect to an external network; The heating unit (11) in the temperature control module is further provided with a second Bluetooth unit (1103) connected to the heating control circuit (1101) for communicating with the first Bluetooth unit (1801) in the main control module.

6. The Tibetan medicine urine image acquisition device according to claim 5, characterized in that: A host computer (21) is also provided, and the host computer (21) includes: A storage unit (2101) for storing images captured by the top-view camera (7) and the side-view camera (8) and calculation results; A computing unit (2102), configured to recognize and compute images; A display unit (2103), used for displaying calculation results; The host computer (21) is connected to the USB interface unit (1802) of the communication interface (18) in the main control module via a USB interface.

7. A method for quantitative identification of Tibetan medicine urine using the collection device according to claim 6, characterized in that: The steps include: S1, placing the standard color module (29) on the tray (4), then placing the upper cover (9) on the chassis (5), controlling the top view camera (7) and the side view camera (8) through the main control module to shoot the image of the standard color module (29) and transmit it to the host computer (21), calculating the LAB measurement value of the standard color module (29) on the host computer (21), and comparing it with the LAB true value of the standard color module (29) to obtain the color correction coefficient; S2. Remove the upper cover (9), remove the standard color module (29), place the urine outer bottle (1) on the tray (4), adjust the rotary lifting rod (6) to place the urine inner bottle (2) into the urine outer bottle (1), tighten the fastening screw (603) to lock the inner rod (602), pour urine into the urine inner bottle (2), and re-place the upper cover (9) on the bottom plate (5); S3, the top-view camera (7) captures the initial floating image of the urine, and the main control module drives the DC motor (3) to drive the tray (4), the urine outer bottle (1) and the urine inner bottle (2) to rotate, and the side-view camera (8) captures the initial side-view image of the urine and transmits it to the host computer (21); S4, adjusting the urine temperature to the heat dissipation period by a natural cooling method, then taking a floating image of the urine by a top-view camera (7), and taking a side-view image of the urine by a side-view camera (8), and transmitting the images to the host computer; S5, using the candy algorithm on the host computer (21) to calculate the edge features of the floating skin image taken in step S4, calculate the length value I of the floating skin crack edge excluding the outer edge, and then calculate the degree value P of the floating skin crack according to the formula I=-log2(1-P); S6. Using the candy algorithm on the host computer (21), the thickness edge feature of the floating skin of the side view image taken in step S4 is calculated, and the thickness of any point on the floating skin edge is calculated according to the following formula: The average thickness of the floating skin is: Where n represents the total number of pixels from the left edge to the right edge of the floating skin in the side view image; Under any horizontal coordinate on the image, the coordinates of the upper edge and the lower edge are (c, a) and (c, b) respectively; The longitudinal field of view angle is β, the side view image resolution is M*N, where M is the horizontal resolution, N is the vertical resolution, and the distance between the side view camera (8) and the center point of the urine outer bottle (1) is h; S7, remove the upper cover (9), adjust the rotary lifting rod (6) to separate the urine inner bottle (2) and the urine outer bottle (1), re-place the upper cover (9) on the bottom plate (5), heat the temperature inside the cover to the hot urine period through the temperature control module, shoot the urine image through the side view camera (8) and transmit it to the host computer (21), calculate the LAB value of the urine image on the host computer (21), and then calibrate the LAB value by the color correction coefficient obtained in step S1 to obtain the calibrated urine image LAB value, and finally calculate the distances S1, S2, and S3 between the calibrated urine image LAB value and the corresponding LAB standard values of urine green, urine yellow, and urine milky white, respectively; Where: (L1, A1, B1) is the LAB value of the calibrated urine image, (L2, A2, B2) is the LAB standard value of blue urine, (L3, A3, B3) is the LAB standard value of yellow urine, and (L4, A4, B4) is the LAB standard value of milky white urine.

8. A Tibetan medicine urine quantitative identification method according to claim 7, characterized in that: After step S6 and before step S7, the quantitative index calculation of floating matter, suspended matter and sediment is further performed, specifically: The depth images of different positions of urine are recorded by a side-view camera (8), and the depth images are enhanced and pre-processed by wavelet transform on the host computer (21), and the point cloud data is calculated. After the point cloud is aligned, the whole urine is reconstructed, and the volume of the three-dimensional features of the whole urine is calculated by voxel modeling as V1, the volume of floating matter as V2, the volume of suspended matter as V3, and the volume of sediment as V4, and then the quantitative index is calculated respectively: Where: F, Z and C are the quantitative indices of floating matter, suspended matter and sediment respectively.

9. A Tibetan medicine urine quantitative identification method according to claim 8, characterized in that: After the quantitative index calculation of the floating matter, suspended matter and sediment and before step S7, the recognition model construction of the floating matter, suspended matter and sediment and the urine image recognition are performed on the host computer (21), specifically: Manually labeling multiple urine images for floating matter, suspended matter, and sediment, classifying urine materials into three forms: floating matter, suspended matter, and sediment. Using these three forms as training and test sets, a first recognition model based on deep CNN image classification for these three forms was constructed using a two-dimensional deep learning algorithm. The urine substances in multiple urine images were manually labeled and divided into six forms: as slender as cow hair, shaped like a cotton ball, dense in the center and sparse around, scattered like gravel, floating like clouds, and like needle tips. The six forms were used as training sets and test sets, and a two-dimensional deep learning algorithm was used to construct a second recognition model based on deep CNN image classification for the six forms. Finally, the urine image to be identified was input, and the first recognition model was used to identify the substance in the urine as one of floating matter, suspended matter, and sediment. Then, the second recognition model was used to identify the form of the substance in the urine as one of slender as cow hair, shaped like a cotton ball, dense in the center and sparse around, scattered like gravel, floating like clouds, and like needle tips.

Images