Non-contact cattle rumen filling degree detection device and detection method

Abstract

The invention discloses a non-contact bovine rumen filling degree detection device and method. The device comprises a one-way channel, a first light source module, a second light source module, an image acquisition module, a synchronous triggering unit and an analysis and calculation module, the one-way channel controls cattle to advance, the analysis and calculation module recognizes a cattle rumen area picture, judges whether the cattle rumen area is located in the horizontal center of an imaging image or not, sends a control signal to the synchronous triggering unit if the cattle rumen area is located in the center, and receives a first picture, a second picture and a natural light picture which are respectively projected to the cattle rumen area. A fullness index is calculated based on the first and second pictures and the natural light picture. The first light source module and the second light source module respectively project a cattle rumen area, and the image acquisition module acquires a cattle rumen area picture and an exposed picture; and the synchronous triggering unit executes projection and acquisition instructions. According to the invention, rapid, automatic and accurate detection of the filling degree index can be realized.

Description

Technical Field

[0001] The invention relates to the field of bovine rumen fullness detection, and in particular to a non-contact bovine rumen fullness detection device and a detection method. Background Art

[0002] The fullness of the cow's rumen can reflect the cow's rumination and digestion status in real time, help accurately control the feeding amount and nutritional ratio, prevent food accumulation or malnutrition, ensure digestive health, improve milk production efficiency and reduce feed waste. It is one of the key indicators for evaluating cattle feeding and digestive health.

[0003] At present, the detection of cattle rumen fullness mainly relies on manual visual inspection, lacks a quantitative measurement mechanism, and the detection results are greatly affected by the operator's experience and have subjective errors. The existing patent "A cattle rumen fullness measurement device and cattle rumen fullness measurement system" (patent number: CN 112545492 B) changes the size of the light channel by physically pressing to drive the compression part to deform, and measures the cattle rumen fullness by detecting the light intensity. This method requires contact with individual cattle and external force intervention may change the actual filling state of the rumen, resulting in inaccurate measurement. The existing patent "A cattle rumen fullness detection device and detection method" (application number 202411627435.9) uses ultrasonic ranging technology, but still needs to contact the cattle's body surface through a fixed device. In actual application, the operation is difficult and the efficiency is low. Summary of the invention

[0004] The purpose of the present invention is to provide a non-contact bovine rumen fullness detection device and detection method, aiming to solve the problem of rapid detection and accuracy of bovine rumen fullness.

[0005] The present invention provides a non-contact bovine rumen filling degree detection device, comprising: One-way channel, light source module, image acquisition module, synchronous trigger unit and analysis and calculation module; The light source module includes a first light source and a second light source, the image acquisition module includes a first image acquisition unit and a second image acquisition unit, and the analysis and calculation module includes a recognition unit and a calculation unit; The one-way channel is used to guide the cattle to move forward in a preset direction; The first light source and the second light source are installed above and below the one-way channel, respectively, and are used to project the cow rumen area at different angles; The first image acquisition unit is used to acquire a picture of the cow's abdomen after the cow enters the one-way channel under natural light, and send it to the recognition unit; The identification unit is used to identify the cow rumen area in the cow abdomen picture, and when it is determined that the cow rumen area is located in the horizontal center of the cow abdomen picture, send a control signal to the synchronization trigger unit; The synchronous trigger unit is used to receive the control signal sent by the identification unit, control the first light source to turn on, and then control the second image acquisition unit to acquire a first picture of the cattle rumen area projected by the first light source, then control the first light source to turn off, the second light source to turn on, control the second image acquisition unit to acquire a second picture of the cattle rumen area projected by the second light source, and finally control the second light source to turn off, and control the second image acquisition unit to acquire a natural light picture of the cattle rumen area under natural light only; The second image acquisition unit sends the first picture, the second picture and the natural light picture of the cow rumen area to the computing unit; The calculation unit is used to receive the first picture, the second picture and the natural light picture of the cow rumen area, and calculate the cow rumen fullness index.

[0006] The present invention also provides a non-contact method for detecting the fullness of the cattle rumen, which guides the cattle to move forward in a preset direction through a one-way channel; The first image acquisition unit acquires a picture of the cow's abdomen after the cow enters the one-way channel under natural light, sends the picture of the cow's abdomen to the recognition unit, recognizes the cow's rumen area in the cow's abdomen picture, and sends a control signal to the synchronization trigger unit when it is determined that the cow's rumen area is located at the horizontal center of the cow's abdomen picture; The synchronous trigger unit controls the first light source to turn on, and then controls the second image acquisition unit to acquire the first image of the cattle rumen area projected by the first light source, and then controls the first light source to turn off, and the second light source to turn on, and controls the second image acquisition unit to acquire the second image of the cattle rumen area projected by the second light source, and finally controls the second light source to turn off, and controls the second image acquisition unit to acquire the natural light image of the cattle rumen area under natural light only; The first picture, the second picture and the natural light picture of the cow rumen area are sent to a calculation unit to calculate the cow rumen fullness index.

[0007] By adopting the embodiments of the present invention, physical contact can be completely avoided, animal stress reactions can be eliminated, and no human intervention is required from cattle guidance and positioning to result output, which can greatly improve detection efficiency.

[0008] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it is implemented in accordance with the contents of the specification, and in order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0010] Figure 1 is a schematic diagram of a non-contact bovine rumen fullness detection device according to an embodiment of the present invention; Figure 2 It is a circuit structure diagram of a non-contact bovine rumen fullness detection device according to an embodiment of the present invention.

[0011] Description of reference numerals: 1: Equipment body; 2: One-way channel; 3: Light source No. 1; 4: Light source No. 2; 5: Image acquisition module; 6: Synchronous trigger unit; 7: Analysis and calculation module. DETAILED DESCRIPTION

[0012] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are part of the embodiments of the present invention, rather than 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.

[0013] Device Embodiment According to an embodiment of the present invention, a non-contact bovine rumen fullness detection device is provided. Figure 1 Schematic diagram of a non-contact bovine rumen fullness detection device according to an embodiment of the present invention. Figure 1 As shown, specifically including: A one-way channel 2, a light source module, an image acquisition module 5, a synchronous trigger unit 6 and an analysis and calculation module 7 are arranged on the device body 1, wherein the light source module includes a first light source 3 and a second light source 4, the image acquisition module 5 includes a first image acquisition unit and a second image acquisition unit, and the analysis and calculation module 7 includes a recognition unit and a calculation unit; The one-way channel 2 is used to guide the cattle to move forward in a preset direction; The first light source 3 and the second light source 4 are installed above and below the one-way channel 2, respectively, and are used to project the rumen area of ​​the cow at different angles; The first image acquisition unit is used to acquire a picture of the cow's abdomen after the cow enters the one-way channel 2 under natural light, and send it to the recognition unit; The identification unit is used to identify the cow rumen area in the cow abdomen picture, and when it is determined that the cow rumen area is located in the horizontal center of the cow abdomen picture, send a control signal to the synchronization trigger unit; The synchronous trigger unit 6 is used to receive the control signal sent by the identification unit, control the first light source 3 to turn on, and then control the second image acquisition unit to acquire the first picture of the cattle rumen area projected by the first light source 3, and then control the first light source 3 to turn off, the second light source 4 to turn on, and control the second image acquisition unit to acquire the second picture of the cattle rumen area projected by the second light source 4, and finally control the second light source 4 to turn off, and control the second image acquisition unit to acquire the natural light picture of the cattle rumen area under natural light only; The second image acquisition unit sends the first picture, the second picture and the natural light picture of the cow rumen area to the computing unit; The calculation unit is used to receive the first picture, the second picture and the natural light picture of the cow rumen area, and calculate the cow rumen fullness index.

[0014] The specific design of the device is as follows: The equipment body 1 adopts a high-strength aluminum alloy frame with an anti-corrosion coating sprayed on the surface. The protection level reaches IP68. It can withstand humidity and ammonia corrosion in the pasture environment. It has an integrated height adjustment mechanism. The height adjustment mechanism realizes vertical movement through a guide rail and is fixed to the equipment body 1 through a fixed knob and a one-way channel 2. The adjustment range is 1.2-1.8m, which is suitable for the rumen position of cattle of different sizes.

[0015] The one-way channel 2 is installed in the production area as the overall support of the equipment. The channels on both sides are vertically fixed on the ground with a width of 700mm. Adjustable limit baffles are set on both sides (adjustment range ±50mm). The inner side of the baffle is covered with a flexible rubber layer (thickness 10mm). The channel floor is embedded with anti-slip striped steel plates (stripes depth 2mm) with an inclination angle of 2° to guide the cattle to move forward naturally.

[0016] The first light source 3 and the second light source 4 are used to project the rumen area of ​​the cow respectively; the first light source 3 and the second light source 4 use an 850nm near-infrared light source array and pulse modulation technology to ensure short-term high-brightness projection. The projection range covers the left rumen fossa area.

[0017] The light source No. 1 3 is installed at the upper left rear of the one-way channel 2, 1.5m above the ground, with a projection angle of 30°, using a 16×16 array 850nm near-infrared LED, a pulse modulation peak power of 200mW, a duty cycle of 1:10, and a spot diameter coverage range of 800-1000mm.

[0018] The second light source 4 is installed at the lower left rear of the one-way channel 2, 0.8m above the ground, with a projection angle of 45°. The second light source 4 and the first light source 3 are in an asymmetric geometric distribution to enhance the difference in projection contours and compensate for the projection deformation error caused by body position deviation. A near-infrared light source of 850nm in the same band is used, and the light intensity distribution gradient design (central light intensity> edge light intensity 20%) is used.

[0019] The image acquisition module 5 is fixed on the equipment body 1 in a manner perpendicular to the unidirectional channel 2, with a vertical distance close to the height of the cow rumen. The height is adjustable and installed in the center of the right side wall of the unidirectional channel 2 (opposite to the rumen projection surface). It includes a high frame rate global shutter camera and a matching narrow-band filter, specifically a global shutter CMOS camera with a frame rate ≥120fps, equipped with an 850nm narrow-band filter (bandwidth ±5nm, transmittance >90%), installed in the center of the right side wall of the channel, 1.2m away from the central axis of the channel, and the optical axis coincides with the normal of the rumen projection surface.

[0020] The synchronous trigger unit 6 uses an FPGA hardware trigger circuit to control the light source pulse and the camera exposure timing. The delay time from light source on to exposure is less than 100μs, and the time-sharing trigger accuracy is ±10μs, ensuring that there is no light pollution in the dual-light source image acquisition interval. The synchronous trigger unit 6 is used to control the time-sharing trigger timing of the first light source 3 and the second light source 4 and the second image acquisition unit to ensure that the light source irradiation is synchronized with the image acquisition while ensuring that there is no crosstalk between the dual light source images; the exposure time of the second image acquisition unit is strictly synchronized with the light source pulse (exposure time 1ms) to suppress ambient light noise.

[0021] The analysis and calculation module 7 has a built-in embedded GPU (computing power ≥ 4 TFLOPS), which automatically identifies the rumen area of ​​the cattle and determines the location of the rumen area of ​​the cattle in combination with a cattle rumen area detection model based on deep vision. The deep vision is implemented using YOACT and Cascade Mask R-CNN algorithms.

[0022] The device also includes a network module, which uses an RFID antenna to identify the individual identity of the cow and upload the fullness index to the cloud, and connect the business to the application system.

[0023] Method Embodiment According to an embodiment of the present invention, a non-contact method for detecting the fullness of the bovine rumen is provided, and a non-contact device for detecting the fullness of the bovine rumen is provided based on the device embodiment, and specifically includes: 1. Guide the cow to move forward in a preset direction through the one-way channel 2, collect a picture of the cow's abdomen after the cow enters the one-way channel under natural light through the first image acquisition unit, send the picture of the cow's abdomen to the recognition unit, recognize the cow's rumen area in the cow's abdomen picture, and send a control signal to the synchronization trigger unit 6 when it is determined that the cow's rumen area is located in the horizontal center of the cow's abdomen picture; Rumen area recognition and positioning uses instance segmentation models such as YOACT and Cascade Mask R-CNN for real-time recognition, and high-accuracy recognition is achieved through confidence adjustment and area secondary verification.

[0024] When the centroid of the recognition result area is in the horizontal center of the image, the high-speed time-sharing projection imaging is automatically triggered. When the rumen area of ​​the cow is in the horizontal center of the imaging image, the time-sharing trigger timing of the first light source 3 and the second light source 4 and the image acquisition module is controlled by the synchronous trigger unit 6, and the synchronous trigger unit 6 performs the following steps in the following timing: The synchronous trigger unit 6 triggers the first light source 3 to turn on. The light source duration is 10ms. The exposure time of the image acquisition module is strictly synchronized with the light source pulse (exposure time 1ms) to suppress ambient light noise. The second image acquisition unit is synchronously triggered to capture the image ; Turn off light source 3 No. 1: wait for 2ms to eliminate the afterglow; the synchronous trigger unit 6 switches to trigger the second light source 4 to turn on, and synchronously triggers the second image acquisition unit to capture the image ; Turn off light source No. 2 4.

[0025] Collect natural light pictures of the rumen area of ​​cattle when there is no light source No. 1 3 and light source No. 2 4 but only natural light, and send them to the calculation unit; 2. The calculation unit obtains the first image, the second image, and the natural light image of the cattle rumen area, and calculates the effective projection areas of the first image and the second image based on the first image, the second image, and the natural light image of the cattle rumen area; Since the same-band light source is used, the projection signals of the two light sources need to be separated by high-speed time-sharing triggering. In a very short time interval, the change of ambient light can be ignored, and differential processing can effectively extract the effective projection area. Through differential processing, the ambient light and sensor noise are eliminated, and the rumen shadow area independently projected by the two light sources is accurately extracted, and the effective projection area is finally calculated. , The experiment shows that the algorithm can still maintain a signal-to-noise ratio of ≥ 28dB under 2000lux ambient light.

[0026] Calculating the effective projection areas of the first image and the second image based on the first image and the second image and the natural light image of the cow rumen area specifically includes: using formula 1 to calculate the effective projection areas A1 and A2, and formula 1 is as follows: , Formula 1; in Natural light picture of the rumen area of ​​cattle at coordinates The pixel value at For the first image at coordinates The pixel value at For the second image at coordinates The pixel value at ; , is the standard deviation of the natural light image of the cattle rumen area, and the threshold is set to 3 times the standard deviation of the natural light image of the cattle rumen area, covering 99.7% of the noise fluctuation (based on the 3σ principle of normal distribution); is a step function used to binarize the projection area. , otherwise it is 0, and v is the function parameter.

[0027] 3. Calculate the deformation difference index between the first image and the second image; The asymmetric layout of the dual light sources leads to differences in projection contours. The gradient difference is used to quantify the deformation caused by body position deviation, solving the problem that a single light source is insensitive to contour deformation. Calculate the deformation difference index When the cow tilts sideways, the projection area of ​​light source 1 3 (upper back) decreases, while the projection area of ​​light source 2 4 (lower front) increases. The experimental data showed that when the body position was shifted by 5°, The value increased by 3.2 times, and the sensitivity increased by 4 times compared with the single light source area method.

[0028] Calculating the deformation difference index between the first image and the second image specifically includes: calculating the deformation difference index D based on Formula 2, where Formula 2 is as follows: Formula 2; in, The coordinates in the first picture are The gradient operator of the effective pixel point i is used to calculate the edge strength of the pixel point in the first picture; The coordinates in the second picture are The gradient operator of the effective pixel point i is used to calculate the edge strength of the pixel point of the second picture; is the number of effective pixels in the overlapping area of ​​the dual light sources, The larger the value, the more significant the difference in the projection contours of the two light sources, which indicates that the body position deviation or local rumen deformation is more severe.

[0029] 4. Calculate the overlapping area of ​​the first image and the second image, and calculate the dynamic weight; First, calculating the overlapping area of ​​the first image and the second image specifically includes: calculating the overlapping area of ​​the first image and the second image based on Formula 3, where Formula 3 is as follows: Formula 3; is the threshold for determining the overlapping area.

[0030] Second, calculating the dynamic weight specifically includes calculating the dynamic weight based on Formula 4, which is as follows: When the body position is stable, , , the area mean is preferred; When the body position shifts, , , enhance deformation difference The weight of .

[0031] The overlapping area of ​​the dual light source projections reflects the reliable measurement area when the body position is stable. The higher the proportion, the higher the measurement confidence. Adaptively balance the contributions of area and deformation indices.

[0032] 5. By integrating the effective projection area of ​​dual light sources ( , )、Overlapping area( ), deformation difference index ( ) and dynamic weights ( ) four core variables to achieve non-contact accurate detection of rumen filling degree. Among them, the area mean Reflects the overall filling state of the rumen and the proportion of overlapping areas Characterizes body position stability, deformation difference index Dynamic offset and dynamic weight of quantized projection profile The contribution weights of the area and deformation characteristics are adaptively adjusted according to the proportion of the overlapping area. Based on the above variables, the filling index is calculated .

[0033] The filling index is calculated using Formula 5 based on the effective projected area, deformation difference index, overlapping area and dynamic weight. Formula 5 is as follows: ; in, is the mean projection area under standard filling conditions, is the maximum allowable deformation difference; is a dynamic weight, with a value range of 0 or 1; It is the ratio of the mean projection area of ​​dual light sources to the standard area; is the normalized value of the deformation difference, which is used to compensate for the area distortion caused by body position deviation. The lower it is, the stronger the deformation compensation is, which is suitable for activity interference scenarios.

[0034] 6. Identify the individual identity of the cow through the network module and upload the fullness index to the cloud.

[0035] The present invention scores the fullness index, specifically including: Based on clinical palpation verification and pasture management needs, a nonlinear segmented scoring system was established, with K1, K2, K3, and K4 as scoring thresholds: when the quantitative value is greater than K1, the score is 1 point; when the quantitative value is greater than K2 and less than or equal to K1, the score is 2 points; when the quantitative value is greater than K3 and less than or equal to K2, the score is 3 points; when the quantitative value is greater than K4 and less than or equal to K3, the score is 4 points; when the quantitative value is less than or equal to K4, the score is 5 points.

[0036] The present invention discloses a non-contact bovine rumen fullness detection device and detection method, which can quickly and accurately calculate the fullness of the rumen. Compared with the prior art, the present invention has the following advantages.

[0037] Non-contact and stress-free: completely avoid physical contact, eliminate animal stress response, and comply with animal welfare requirements; Strong anti-interference ability: non-visible light band time-division differential imaging technology effectively reduces ambient light interference; High-precision analysis: Through multi-dimensional fusion of dual projection areas, overlapping areas and morphological similarity, the accuracy of single light source solutions is effectively improved; Fully automated operation: no human intervention is required from cattle guidance and positioning to result output, which can greatly improve detection efficiency; Low cost and easy to deploy: No 3D reconstruction hardware is required, only 2D image processing algorithm is needed, which effectively reduces the equipment cost; Assess the degree of fullness to quantify the quality.

[0038] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements of the technical solutions of the embodiments of the present invention do not cause the essence of the corresponding technical solutions to deviate from the scope of this solution.

Claims

1. A non-contact bovine rumen fullness detection device, characterized in that: It includes a one-way channel, a light source module, an image acquisition module, a synchronous trigger unit and an analysis and calculation module; The light source module includes a first light source and a second light source, the image acquisition module includes a first image acquisition unit and a second image acquisition unit, and the analysis and calculation module includes a recognition unit and a calculation unit; The one-way channel is used to guide the cattle to move forward in a preset direction; The first light source and the second light source are installed above and below the one-way channel, respectively, and are used to project the rumen area of ​​the cow at different angles; The first image acquisition unit is used to acquire a picture of the cow's abdomen after the cow enters the one-way channel under natural light, and send it to the recognition unit; The identification unit is used to identify the cow rumen area in the cow abdomen picture, and when it is determined that the cow rumen area is located in the horizontal center of the cow abdomen picture, send a control signal to the synchronization trigger unit; The synchronous trigger unit is used to receive the control signal sent by the identification unit, control the first light source to turn on, and then control the second image acquisition unit to acquire a first picture of the cattle rumen area projected by the first light source, and then control the first light source to turn off, the second light source to turn on, and control the second image acquisition unit to acquire a second picture of the cattle rumen area projected by the second light source, and finally control the second light source to turn off, and control the second image acquisition unit to acquire a natural light picture of the cattle rumen area under natural light only; The second image acquisition unit sends the first picture, the second picture and the natural light picture of the cow rumen area to the computing unit; The calculation unit is used to receive the first picture, the second picture and the natural light picture of the cow rumen area, and calculate the cow rumen fullness index.

2. The device according to claim 1, characterized in that The calculation unit is specifically used to calculate the effective projection area of ​​the first image and the second image based on the first image and the second image and the natural light image of the bovine rumen area, calculate the deformation difference index of the first image and the second image, calculate the overlapping area of ​​the first image and the second image, calculate the dynamic weight, and calculate the fullness index based on the effective projection area, the deformation difference index, the overlapping area and the dynamic weight.

3. A non-contact method for detecting bovine rumen fullness, based on the device according to claim 1 or 2, characterized in that: Guide cattle in a preset direction through one-way channels; The first image acquisition unit acquires a picture of the cow's abdomen after the cow enters the one-way channel under natural light, sends the picture of the cow's abdomen to the recognition unit, recognizes the cow's rumen area in the cow's abdomen picture, and sends a control signal to the synchronization trigger unit when it is determined that the cow's rumen area is located at the horizontal center of the cow's abdomen picture; The synchronous trigger unit controls the first light source to turn on, and then controls the second image acquisition unit to acquire the first image of the cattle rumen area projected by the first light source, and then controls the first light source to turn off, and the second light source to turn on, and controls the second image acquisition unit to acquire the second image of the cattle rumen area projected by the second light source, and finally controls the second light source to turn off, and controls the second image acquisition unit to acquire the natural light image of the cattle rumen area under natural light only; The first picture, the second picture and the natural light picture of the cow rumen area are sent to a calculation unit to calculate the cow rumen fullness index.

4. The method according to claim 3, characterized in that: The calculation of the cattle rumen filling index specifically includes: Based on the first image and the second image and the natural light image of the bovine rumen area, the effective projection areas of the first image and the second image are calculated, the deformation difference index of the first image and the second image is calculated, the overlapping area of ​​the first image and the second image is calculated, the dynamic weight is calculated, and the fullness index is calculated based on the effective projection area, the deformation difference index, the overlapping area and the dynamic weight.

5. The method according to claim 4, characterized in that The method of calculating the effective projection areas of the first image and the second image based on the first image and the second image and the natural light image of the cow rumen region specifically includes: using formula 1 to calculate the effective projection areas A1 and A2, and formula 1 is as follows: ， Formula 1; in Natural light picture of the rumen area of ​​cattle at coordinates The pixel value at For the first image at coordinates The pixel value at For the second image at coordinates The pixel value at ; , is the standard deviation of the natural light image of the cow rumen area, the threshold It is set to 3 times the standard deviation of the natural light image of the cow rumen area; is a step function used to binarize the projection area. , otherwise it is 0, and v is the function parameter.

6. The method according to claim 4, characterized in that The calculating of the deformation difference index between the first image and the second image specifically includes: calculating the deformation difference index D based on Formula 2, where Formula 2 is as follows: Formula 2: in, The coordinates in the first picture are The gradient operator of the effective pixel point i is used to calculate the edge strength of the pixel point in the first picture; The coordinates in the second picture are The gradient operator of the effective pixel point i is used to calculate the edge strength of the pixel point of the second picture; is the number of effective pixels in the overlapping area of ​​light source No. 1 and light source No.

2.

7. The method according to claim 5, characterized in that Calculating the overlapping area of ​​the first image and the second image specifically includes: calculating the overlapping area of ​​the first image and the second image based on Formula 3, where Formula 3 is as follows: Formula 3: is the threshold for determining the overlapping area.

8. The method according to claim 7, characterized in that Calculating the dynamic weight specifically includes calculating the dynamic weight based on Formula 4, which is as follows: When the body position is stable, , ; When the body position shifts, , , Formula 4.

9. The method according to claim 8, characterized in that The calculation of the fullness index based on the effective projection area, the deformation difference index, the overlapping area and the dynamic weight specifically includes: calculating the fullness index based on formula 5, which is as follows: , Formula 5; in, is the mean projection area under standard filling conditions, is the maximum allowable deformation difference; is a dynamic weight, with a value range of 0 or 1; It is the ratio of the mean projection area of ​​dual light sources to the standard area; It is the normalized value of deformation difference, which is used to compensate for the area distortion caused by body position deviation.

10. The method according to claim 3, characterized in that: The method also includes scoring a filling index.

Images