Device and method for automatically measuring maturing rate of rice grains

Abstract

The invention discloses an automatic measuring device for the maturing rate of rice grains. The whole device is installed on a rack. The device provided by the embodiment of the invention mainly comprises two measuring assembly lines, the first assembly line is used for counting the total number of grains to be measured, and the first assembly line mainly comprises a first feeding device, a first conveying device and a first image acquisition device; and the second assembly line is used for counting and weighing the winnowed full grains and mainly comprises a second feeding device, a second conveying device, a second image acquisition device and an electronic balance. The key component for connecting the two assembly lines is the winnowing device. Automatic control and data processing of the whole system are completed by the single-chip microcomputer and the processor together. The invention provides an automatic measuring device and a measuring method for the maturing rate of rice grains, which realize full-process automation from sample input to result output and ensure high precision and high efficiency of a measuring result.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of agricultural automation measurement technology, and in particular to a device and method for automatically measuring the grain setting rate of rice. BACKGROUND

[0002] In the field of agricultural research, the grain setting rate of rice, i.e. the percentage of full grains in the total number of grains per panicle, is one of the key indicators for measuring the yield composition and breeding results of rice. In the fields of rice breeding, high-yield cultivation technology research, physiological and ecological analysis, and yield prediction, it is necessary to quickly and accurately measure the grain setting rate of rice.

[0003] Currently, the traditional method for measuring the grain setting rate of rice mainly relies on manual operation. The general process includes: manual threshing, separating full grains and empty grains by air or water selection, then manually counting the number of each type of grain, and finally calculating the setting rate. This traditional method has obvious drawbacks: first, the labor intensity is high, and a large amount of manpower and time is consumed, especially when a large number of samples need to be processed, the efficiency is extremely low; second, the accuracy of the measurement results is heavily dependent on the experience and responsibility of the operator, and is highly subjective, which can easily introduce human error, resulting in poor consistency and reliability of the data; third, the entire process cannot be automated, and the data cannot be easily recorded and managed electronically.

[0004] With the development of computer vision technology and image processing technology, some semi-automatic methods have emerged that attempt to use image analysis for grain counting. For example, the grains are laid flat on a scanner or a specific background plate, an image is obtained by taking a photo or scanning, and then software algorithms are used for counting. However, these methods still have some technical difficulties: 1) it is difficult to effectively distinguish between full grains and empty grains, as their projected outlines on a two-dimensional image can be very similar, and relying solely on image feature recognition has a high error rate; 2) grains are prone to overlap and adhesion, posing a great challenge to accurate segmentation and counting; 3) manual separation and placement of the grains are usually required, and the degree of automation is still not high, and the efficiency problem has not been fundamentally solved; 4) there is a lack of integrated weighing function, and other important agronomic trait data such as thousand-grain weight cannot be obtained simultaneously. SUMMARY

[0005] In view of the above defects of the prior art, the technical problem to be solved by the present application is that the prior art measuring method of the rice seed setting rate mainly relies on manual work, which has the problems of high labor intensity, low efficiency, strong subjectivity and easy introduction of errors; and the existing part of the automatic image analysis method is difficult to effectively separate the full grains and the empty grains in physics, resulting in low recognition accuracy, and usually needs manual assistance for feeding and spreading, and fails to realize the full-process automation and data integration. Therefore, the present application provides a rice grain seed setting rate automatic measuring device and measuring method, which realizes full-process automation from sample input to result output, and ensures high precision and high efficiency of the measuring result.

[0006] To achieve the above object, the present application provides a rice grain seed setting rate automatic measuring device, comprising

[0007] a rack, a first feeding device, a first conveying device, a first image acquisition device, an air separation device, a second feeding device, a second conveying device, a second image acquisition device, a weighing device and a control device;

[0008] Among them, for supporting the first feeding device, the first conveying device, the first image acquisition device, the air separation device, the second feeding device, the second conveying device, the second image acquisition device, the weighing device and the single-chip microcomputer;

[0009] The first feeding device is configured to uniformly feed the rice grains to be measured to the first conveying device, the first conveying device is arranged below the first feeding device, the air separation device is arranged at the end of the first conveying device, and the air separation device is arranged at the end of the first conveying device. The air separation device is used for air separation of the rice grains to be measured on the first conveying device to separate full grains and empty grains; the first image acquisition device is arranged above the first conveying device;

[0010] The second feeding device is arranged below the air separation device, and is used for collecting full grains and conveying the full grains to the second conveying device; the second conveying device is arranged below the second feeding device; the second image acquisition device is arranged above the second conveying device; and the load bearing device is arranged at the end of the second conveying device;

[0011] The control device is connected with the first feeding device, the first conveying device, the first image acquisition device, the air separation device, the second feeding device, the second conveying device, the second image acquisition device and the weighing device.

[0012] Further, the control device comprises a single-chip microcomputer, a processor and a display module, the single-chip microcomputer is connected with the first feeding device, the second feeding device, the first conveying device, the air separation device, the second conveying device and the weighing device, and is used for controlling the operation of the first feeder, the second feeder, the first conveying device, the air separation device and the second conveying device, and receiving the weighing data of the weighing device;

[0013] The processor is connected with the first image acquisition device, the second image acquisition device and the single-chip microcomputer respectively, and is used for receiving the weighing data transmitted by the first image acquisition device, the second image acquisition device and the single-chip microcomputer, performing grain identification and grain number counting on the received image data by using image processing technology, and calculating the setting rate of the rice grains based on the statistical result.

[0014] The display module is connected with the processor, and is used for displaying the setting rate of the rice grains and the weighing result.

[0015] Further, the first feeding device comprises a feeding hopper and a first feeder, the feeding hopper is arranged above the first feeder, and the feeding hopper is provided with a square opening.

[0016] Further, the first image acquisition device comprises a first camera and a first camera support, the first camera support is fixed on the rack, and the first camera is installed on the first camera support and arranged above the first conveying device.

[0017] Further, the first conveying device comprises a first direct-current speed reduction motor, a first synchronous belt transmission device, a first conveying belt, a first driving shaft, a first driven shaft and a first shaft support block.

[0018] The first conveying belt is sleeved on the first driving shaft and the first driven shaft.

[0019] The first direct-current speed reduction motor is in transmission connection with the first driving shaft through the first synchronous belt transmission device, so as to drive the first conveying belt to move.

[0020] The first shaft support block is fixed on the rack, and is used for supporting the first driving shaft and the first driven shaft.

[0021] Further, the second feeding device comprises a collecting hopper and a second feeder, the collecting hopper is arranged below the winnowing device, the second feeder is arranged below the collecting hopper, and the collecting hopper has a square opening.

[0022] Further, the second conveying device comprises a second direct-current speed reduction motor, a second synchronous belt transmission device, a second conveying belt, a second driving shaft, a second driven shaft and a second shaft support block.

[0023] The second conveying belt is sleeved on the second driving shaft and the second driven shaft.

[0024] The second direct-current speed reduction motor is in transmission connection with the second driving shaft through the second synchronous belt transmission device, so as to drive the second conveying belt to move.

[0025] The second shaft support block is fixed on the rack, and is used for supporting the second driving shaft and the second driven shaft.

[0026] Further, the winnowing device comprises a centrifugal fan and a wind channel device, the centrifugal fan is arranged on the frame, the wind channel device is arranged at the air outlet of the centrifugal fan, the air outlet of the wind channel device is arranged below the end of the first conveying device, and the air outlet of the wind channel device is arranged to be inclined along the free fall direction of the grains, and the air outlet is inclined upward by 15 degrees relative to the horizontal direction.

[0027] A further preferred embodiment of the present application provides an automatic measurement method for the setting rate of rice grains, using the above-mentioned automatic measurement device for the setting rate of rice grains, specifically comprising the following steps:

[0028] The processor is connected with the first image acquisition device, the second image acquisition device and the single-chip microcomputer for receiving the first video image from the first image acquisition device, the second video image from the second image acquisition device and the weight data sent by the single-chip microcomputer;

[0029] The image frames in the first video image and the second video image are extracted at a set interval, and the extracted image frames are preprocessed, edge detected and dynamically image spliced, so as to reconstruct and store the complete grain contour information in the memory;

[0030] The reconstructed and stored complete grain contour is profile screened and classified and counted;

[0031] The processor receives the weight data sent by the weighing device, and calculates the setting rate R = n1 / n0 based on the above-mentioned statistical result; the processor displays the calculated setting rate and weight information through the webpage of the display module, and provides the download function of the detection data of this time.

[0032] Further, the image frames in the first video image and the second video image are extracted at a set interval, and the extracted image frames are preprocessed, edge detected and dynamically image spliced, so as to reconstruct and store the complete grain contour information in the memory, specifically comprising the following steps:

[0033] A frame of image is extracted as a key frame every preset frame number from the video image, and the key frame is preprocessed, including grayscale and Gaussian filter denoising;

[0034] The Canny edge detection algorithm is used for the processed key frame image, and the edge contour information of the grain is accurately extracted to form an edge map;

[0035] The edge maps of two key frames extracted continuously are compared, the theoretical displacement of the grains in the image is calculated based on the conveying belt speed and the frame rate, and matching recognition is performed by calculating the intersection-over-union between the edge contours in the prediction area; for the contour fragments that match successfully, geometric alignment and contour splicing are performed, and the overlapping contours are de-duplicated, so that a complete grain closed contour image is reconstructed in the memory, and each reconstructed contour is assigned a unique label.

[0036] Technical effects

[0037] The device and method for automatically measuring the rice grain setting rate provided by the application integrate various functional modules such as image acquisition, air separation, weighing measurement, conveying control and data processing, and have the advantages of compact structure, high automation degree, high recognition accuracy and high measurement efficiency.

[0038] The device adopts a structure design of "two-stage conveying + two-stage image acquisition + air separation + weighing", and realizes a pipeline operation through modular design. The first image acquisition system dynamically photographs and counts all the grains to be measured, the air separation device realizes physical separation of full and empty grains, the second image acquisition system counts the separated full grains again, and the electronic balance simultaneously completes the weight measurement, and finally the processor automatically calculates the setting rate.

[0039] In terms of image acquisition and processing, the system adopts video frame extraction, key frame splicing and edge detection algorithm, combined with contour matching and de-duplication strategy, effectively deals with complex situations such as grain occlusion, adhesion and overlap, and ensures the accuracy and robustness of image recognition and counting. The conveying structure cooperates with the guide structure design to keep the grains stable in the conveying process, which is convenient for image processing.

[0040] The air separation device optimizes the airflow direction and stability through the inclined angle air duct and the honeycomb flow regulation structure, realizes efficient and accurate separation of grains, and solves the problem that the image features are difficult to directly distinguish full and empty grains. The weighing module and the image module are cooperatively linked to provide complete data support for the setting rate measurement.

[0041] In addition, the device cooperatively controls the single-chip microcomputer and the processor to realize the collaborative operation and data acquisition and processing of multiple modules, and the measurement results can be displayed and downloaded in real time through the web page, which is convenient for management and traceability.

[0042] In summary, the application realizes the full-process automation of the rice grain setting rate from image acquisition, physical separation, data calculation to result output, and is especially suitable for application scenarios such as rice breeding, quality detection and high-throughput intelligent screening, and has strong practicality and popularization value.

[0043] The concept, specific structure and generated technical effects of the present application will be further described below in combination with the drawings to fully understand the purposes, features and effects of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 is a whole structure schematic diagram of a rice grain setting rate automatic measuring device of a preferred embodiment of the present application;

[0045] Figure 2 is a flowchart of a rice grain setting rate automatic measuring method of a preferred embodiment of the present application;

[0046] Figure 3 is a funnel structure schematic diagram of a rice grain setting rate automatic measuring device of a preferred embodiment of the present application;

[0047] Figure 4 is a wind sifting device structure schematic diagram of a rice grain setting rate automatic measuring device of a preferred embodiment of the present application;

[0048] Figure 5 is a flowchart of a rice grain setting rate automatic measuring method of a preferred embodiment of the present application;

[0049] wherein, 1, feeding hopper; 2, first feeder; 3, first conveying belt; 4, first camera; 5, first camera support; 6, first guide protrusion; 7, first shaft support block; 8, first driven shaft; 9, air duct device; 10, centrifugal fan; 11, collecting hopper; 12, second feeder; 13, second driven shaft; 14, second conveying belt; 15, second camera; 16, second camera support; 17, second DC motor; 18, second shaft support block; 19, second guide protrusion; 20, electronic balance; 21, second driving shaft; 22, second synchronous belt transmission device; 23, first DC motor; 24, first synchronous belt transmission device; 25, first driving shaft; 26, single-chip microcomputer; 91, first air duct section; 92, transition air duct section; 93, rectifier section. DETAILED DESCRIPTION

[0050] In order to make the technical problems, technical solutions and beneficial effects of the present application more clear, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0051] In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular procedures, techniques, etc., in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.

[0052] As shown in the drawings, Figure 1 The present application provides a device for automatically measuring the grain setting rate of rice, which is installed on a frame. The device mainly comprises two measuring lines. The first line is used for counting the total number of the rice grains to be measured, and mainly comprises a first feeding device, a first conveying device and a first image acquisition device. The second line is used for counting and weighing the full grains after air separation, and mainly comprises a second feeding device, a second conveying device, a second image acquisition device and an electronic balance. The key component connecting the two lines is an air separation device. The automatic control and data processing of the whole system are completed by a single-chip microcomputer and a processor. The device for automatically measuring the grain setting rate of rice comprises:

[0053] a frame, a first feeding device, a first conveying device, a first image acquisition device, an air separation device, a second feeding device, a second conveying device, a second image acquisition device, a weighing device and a control device.

[0054] The first feeding device, the first conveying device, the first image acquisition device, the air separation device, the second feeding device, the second conveying device, the second image acquisition device, the weighing device and the single-chip microcomputer are supported by the frame.

[0055] The first feeding device is configured to uniformly feed the rice grains to be measured to the first conveying device. The first conveying device is arranged below the first feeding device. The air separation device is arranged at the end of the first conveying device to separate the full grains and the empty grains by air separation. The first image acquisition device is arranged above the first conveying device.

[0056] The second feeding device is arranged below the air separation device to collect the full grains and convey them to the second conveying device. The second conveying device is arranged below the second feeding device. The second image acquisition device is arranged above the second conveying device. The weighing device is arranged at the end of the second conveying device.

[0057] The control device is connected with the first feeding device, the first conveying device, the first image acquisition device, the air separation device, the second feeding device, the second conveying device, the second image acquisition device and the weighing device.

[0058] Specifically, the control device comprises a single-chip microcomputer 26, a processor and a display module. The single-chip microcomputer 26 is connected with the first feeding device, the second feeding device, the first conveying device, the winnowing device, the second conveying device and the weighing device, and is used for controlling the operation of the first feeder, the second feeder, the first conveying device, the winnowing device and the second conveying device, and receiving the weighing data of the weighing device.

[0059] The processor is connected with the first image acquisition device, the second image acquisition device and the single-chip microcomputer 26 respectively, and is used for receiving the weighing data transmitted by the first image acquisition device, the second image acquisition device and the single-chip microcomputer 26, performing grain identification and grain number counting on the received image data by using image processing technology, and calculating the seed setting rate of the rice grain based on the statistical result.

[0060] The display module is connected with the processor, and is used for displaying the seed setting rate of the rice grain and the weighing result.

[0061] The first feeding device comprises a feeding hopper 1 and a first feeder 2. The feeding hopper 1 is arranged above the first feeder 2, and the feeding hopper 1 is provided with a square opening.

[0062] The first image acquisition device comprises a first camera 4 and a first camera support 5. The first camera support 5 is fixed on the rack, and the first camera 4 is installed on the first camera support 5 and arranged above the first conveying device.

[0063] The first conveying device comprises a first direct-current speed reduction motor, a first synchronous belt transmission device, a first conveying belt 3, a first driving shaft, a first driven shaft 8 and a first shaft support block 7.

[0064] The first conveying belt 3 is arranged on the first driving shaft and the first driven shaft 8. The first direct-current speed reduction motor is in transmission connection with the first driving shaft through the first synchronous belt transmission device, so as to drive the first conveying belt to move. The first shaft support block 7 is fixed on the rack, and is used for supporting the first driving shaft and the first driven shaft 8. The first shaft support block 7 is provided with a first guide protrusion 6, which is located at the side edge of the first conveying belt 3, so as to limit the transverse displacement of the first conveying belt 3 when the first conveying belt 3 moves, so as to prevent the grains from deviating to one side on the first conveying belt 3.

[0065] The second feeding device comprises a collecting hopper 11 and a second feeder 12. The collecting hopper 11 is arranged below the winnowing device, and the second feeder 12 is arranged below the collecting hopper. The collecting hopper is provided with a square opening.

[0066] The second conveying device comprises a second direct-current speed reduction motor 17, a second synchronous belt transmission device 22, a second conveying belt 14, a second driving shaft 21, a second driven shaft 13 and a second shaft support block 18.

[0067] The second conveying belt 14 is arranged on the second driving shaft and the second driven shaft 13;

[0068] The second direct current deceleration motor 17 is in driving connection with the second driving shaft through the second synchronous belt transmission device 22, so as to drive the second conveying belt 14 to move.

[0069] The second shaft support block 18 is fixed on the frame and is used for supporting the second driving shaft and the second driven shaft. The second guiding protrusion 19 is arranged on the second shaft support block 18 and is located at the side edge of the second conveying belt 14, so as to limit the transverse displacement of the second conveying belt 14 during movement, so as to prevent the grains from being deviated to one side on the second conveying belt 14.

[0070] The air selection device comprises the centrifugal fan 10 and the air duct device 9. The centrifugal fan 10 is arranged on the frame, and the air duct device 9 is arranged at the air outlet of the centrifugal fan. The air outlet of the air duct device 9 is arranged below the end of the first conveying device, and the air outlet of the air duct device is arranged to be inclined along the free fall direction of the grains, and the air outlet is inclined upward by 15 degrees relative to the horizontal direction.

[0071] The centrifugal fan 10 is used for generating air flow.

[0072] The air duct device 9 is used for collecting the air flow and guiding the to-be-tested paddy grains, so as to realize the separation of the empty grains and the full grains. The air duct device is arranged to be inclined along the free fall direction of the grains, and the air outlet of the air duct device is inclined upward by 15 degrees relative to the horizontal direction, so as to reduce the air flow turbulence and optimize the deflection effect on the grains.

[0073] The air duct device 9 comprises a first air duct section 91, a transition air duct section 92 and a rectifier section 93.

[0074] The first air duct section 91 is connected with the air outlet of the centrifugal fan 10.

[0075] The transition air duct section 92 connects the first air duct section 91 and the rectifier section 93 and has a length of 145 mm, and is used for smoothly transitioning the air flow from the size of 60 mm (width) x 70 mm (height) to the size of 110 mm (width) x 8 mm (height).

[0076] The rectifier section 93 is located at the air outlet end of the transition air duct section 92, has a cross-sectional size of 110 mm (width) x 8 mm (height) and a depth of 25 mm, and has a honeycomb-shaped rectifier grid structure. The honeycomb-shaped rectifier grid structure has a hexagonal diameter of 3 mm and a wall thickness of 0.8 mm, and is used for reducing the air flow turbulence and improving the air flow uniformity.

[0077] Another preferable embodiment of the present application provides a kind of paddy grain soundness automatic measurement method, using the above-mentioned paddy grain soundness automatic measurement device, specifically comprising the following steps:

[0078] The processor is connected with the first image acquisition device, the second image acquisition device and the single-chip microcomputer respectively, and is used for receiving the first video image from the first image acquisition device, receiving the second video image from the second image acquisition device and receiving the weighing data sent by the single-chip microcomputer;

[0079] The continuous frames in the first video image and the second video image are extracted at a set interval, and the extracted image frames are preprocessed, edge detected and dynamically image spliced, so that the complete grain contour information is reconstructed and stored in the memory;

[0080] The contour screening and classification counting are performed on the complete grain contour which has been reconstructed and stored.

[0081] The processor receives the weight data sent by the weighing device, and calculates the kernel rate R = n1 / n0 based on the above statistical result; the processor displays the calculated kernel rate and weight information through the webpage end of the display module, and provides the download function of the detection data.

[0082] Further, the continuous frames in the first video image and the second video image are extracted at a set interval, and the extracted image frames are preprocessed, edge detected and dynamically image spliced, so that the complete grain contour information is reconstructed and stored in the memory, and the method specifically comprises the following steps:

[0083] A frame of image is extracted as a key frame every interval of a preset frame number from the video image, and the key frame is preprocessed, including gray scale and Gaussian filter denoising;

[0084] The Canny edge detection algorithm is used for the processed key frame image, and the edge contour information of the grain is accurately extracted to form an edge graph.

[0085] The edge graphs in the two continuously extracted key frames are compared, the theoretical displacement of the grain in the image is calculated based on the conveyor belt speed and the frame rate, and the matching recognition is performed in the prediction area by calculating the intersection and union ratio between the edge contours; for the matched contour fragments, the geometric alignment and contour splicing are performed, and the overlapping contours are de-duplicated, so that the complete grain closed contour image is reconstructed in the memory, and each reconstructed contour is assigned a unique label.

[0086] The specific use method of the device provided by the embodiment of the application is as follows:

[0087] The operator pours the to-be-detected rice grains into the feeding hopper 1, and the first feeder 2 uniformly and stably conveys the grains to the first conveying device below. The first feeder 2 is preferably an electromagnetic vibration feeder, which generates high-frequency vibration through electromagnetic driving, so that the grains are uniformly moved along the feeding groove and fall on the first conveying belt 3 below one by one, and the grains are uniformly distributed and have no overlap before image acquisition, which is beneficial to the improvement of the subsequent counting accuracy.

[0088] Preferably, the feed hopper 1 adopts an overall square structure (such as Figure 3 As shown), its upper inlet and lower outlet are both rectangular openings, but the upper port is larger than the lower port, forming a typical trapezoidal convergence structure. The four inner walls of the funnel are arranged at an angle, forming four relatively symmetrical inclined ramps for guiding the grains to flow downward. The structure can essentially be regarded as a hollow cuboid, with a trapezoidal ramp volume dug out from its interior to form an internal contraction channel, which helps the grains to slide down smoothly. In order to prevent the grains from being retained or accumulated inside, the feeding funnel 1 is preferably made of metal material with good surface smoothness and structural strength, thereby improving the fluidity of the falling grains and their anti-static dust accumulation ability. This structure not only facilitates manual feeding, but also ensures that the grains can smoothly enter the first feeder 2, avoiding problems such as piling and blocking, and improving the stability and automation level of the overall operation of the system.

[0089] The first conveyor device includes a drive system consisting of a first DC reduction motor 23, a first synchronous belt drive 24, a first driving shaft 25, and a first driven shaft 8, which is used to drive the first conveyor belt 3 to move at a constant speed. The first driving shaft 25 and the first driven shaft 8 are supported by a first shaft support block 7 fixed to the frame. The first shaft support block 7 is provided with a first guide protrusion 6 along the direction of the conveyor belt's travel. This guide protrusion 6 is closely attached to the side edge of the first conveyor belt 3 and can effectively limit any lateral displacement that may occur during the conveyor belt's operation, thereby ensuring that the grains always move along the conveyor belt's predetermined trajectory and stably pass through the center of the field of view of the camera above.

[0090] The first DC reduction motor 23 provides a stable driving force, driving the first drive shaft 25 through the first synchronous belt drive 24, thereby achieving uniform motion of the first conveyor belt 3. The motor's built-in reduction mechanism effectively reduces speed and increases torque, ensuring smoother conveyor system operation. This prevents image blur or grain position shifts caused by conveyor speed fluctuations, thereby ensuring image recognition accuracy. Preferably, the motor is a 12V DC reduction motor, facilitating low-voltage power supply control. It offers fast response, low noise, and smooth startup and shutdown, making it suitable for the continuous dynamic conveying requirements of this device.

[0091] When the grains move along the first conveyor belt 3, they pass through a first image acquisition device located above the first conveyor belt 3. The first image acquisition device is used to acquire first video images of all grains to be measured that pass below the first conveyor belt 3.

[0092] In this embodiment, the first camera 4 is a USB3.0 interface color high frame rate industrial camera with driverless function, which has high image acquisition efficiency and stability. In practical application, the frame rate is set to 60 frames per second, and the resolution is set to 640x480, which can meet the clear imaging requirements of grain dynamic state, while avoiding image blur, ghosting and other problems. The camera is equipped with a lens with a focal length of 3.5mm, which can realize complete coverage and appropriate field of view of the grain on the conveying belt, ensuring that the grain profile in the image acquisition area is clear and the edge features are obvious, which is convenient for subsequent image processing and counting.

[0093] As shown in Figure 4 When the grain freely falls from the end of the first conveying belt 3, it will pass through the effective action area of the air separation device. The air separation device is used to separate full grains and empty grains, which includes a centrifugal fan 10 and an air duct device 9 connected thereto. The centrifugal fan 10 generates a high-speed airflow. The airflow first enters the first air duct section 91 of the air duct device 9, and then passes through the transition air duct section 92. The transition air duct section 92 smoothly converges the cross section of the airflow from a larger square (e.g. 60mm wide x 70mm high) into a flat rectangle (e.g. 110mm wide x 8mm high), which can increase the airflow speed. At the air outlet, a rectification section 93 is provided, which has a honeycomb rectification grid structure inside. This structure can eliminate turbulence in the airflow and shape it into uniform and stable laminar flow. The outlet of the entire air duct device 9 is inclined upward by 15 degrees relative to the horizontal direction, so that this stable airflow can be blown to the falling grain at the best angle, so as to accurately blow the lighter empty grains away from the original falling trajectory, while the heavier full grains remain vertically falling due to their larger inertia.

[0094] Preferably, the centrifugal fan 10 is a small and medium-sized high-speed centrifugal fan with speed regulation capability, with an air volume of about 1.8-2.0m 3 / min, a wind pressure range of 1.2-1.4inH2O, and a rated speed of 2400-3000RPM, which can be adjusted according to actual operation requirements. The fan outputs stable airflow with a certain angle and speed through the air duct structure, which can blow the lighter empty grains away from the original vertical falling trajectory, while the full grains are basically not affected due to their larger inertia, thereby realizing physical separation of the grains. The speed regulation capability of the fan helps to adapt to the density difference of different batches of grains, further improving the air separation precision.

[0095] The full grains are collected by the collecting funnel 11 of the second feeding device after falling vertically. The collecting funnel 11 and the feeding funnel 1 in the first feeding device have the same structural form and material, that is, a square converging structure with a large upper part and a small lower part, and the inner wall is composed of four inclined surfaces, which are preferably made of metal to improve the falling smoothness and anti-stagnation performance of the grains. Then, the grains are uniformly sent to the second conveying belt 14 of the second conveying device by the second feeder 12. The structure of the second conveying device is the same as that of the first conveying device shown in the figure, including a second direct-current speed reducer 17, a second synchronous belt transmission device 22, a second driving shaft 21, a second driven shaft 13 and a shaft support block 18 with a second guide protrusion 19. Figure 1

[0096] When the full grains move along the second conveying belt 14, they will pass through the second image acquisition device located above. The second image acquisition device includes a second camera 15 mounted on a second camera bracket 16, which is used to acquire video images of all full grains. Finally, the full grains fall into the electronic balance 20 from the end of the second conveying belt 14 and are accumulated and weighed. In this embodiment, the second camera 15 has the same setting parameters as the first camera 4, including frame rate, resolution and focal length, to ensure the consistency of image acquisition and the universality of subsequent image processing algorithms.

[0097] The control system of the device is composed of a single-chip microcomputer 26 and a processor to realize automatic control and data processing of the whole grain sorting process. The single-chip microcomputer 26 serves as a lower control unit and has rich input and output interface resources, including multiple digital and analog pins, PWM control capability and multiple serial communication ports, which can connect and control motors, feeders, image acquisition devices, electronic balances and other external devices. The working voltage of the single-chip microcomputer is 5V, the main frequency reaches 16MHz, and it has large program storage space and data caching capability, which is suitable for multi-channel concurrent control and real-time data acquisition tasks.

[0098] The processor serves as an upper computer and is mainly responsible for image recognition, data calculation and human-computer interaction. It communicates with the single-chip microcomputer through a serial port, receives image recognition results, weight data and running state information, and runs image processing and statistical analysis algorithms to output the recognition and weighing results of the full grains. The processor is also connected to a display module, which can be a web interface, a local screen or other visual terminal, used to display the analysis results and device running status to the user in real time.

[0099] ​The electronic balance is used for cumulative weighing of full grains falling from the end of the second conveying belt. The balance has a weighing range of not less than 600 g and a minimum graduation value of not more than 0.01 g, and can meet the high-precision weighing requirement in the grain sorting process. The balance is provided with a serial communication interface (such as RS232 or RS485), supports common communication protocols (such as MODBUS-RTU or ASCII code), and can stably and reliably interact with the single-chip microcomputer. The scale pan has moderate size and stable structure, and is suitable for automatic falling of grains and accurate weighing.

[0100] The above control structure realizes integrated collaborative work of multiple function modules such as image acquisition, motion control, weight detection and data display, and ensures the automation, precision and stability of system operation. The specific models of the single-chip microcomputer, the processor, the electronic balance and the display module are not limited, and devices with corresponding function and performance parameters can be used instead.

[0101] In combination with the above rice grain setting rate automatic measurement device, the embodiment of the present application further provides a rice grain setting rate automatic measurement method, as shown in Figure 5 The specific steps are as follows:

[0102] Step one, video image acquisition and key frame extraction; specifically, under the control of the single-chip microcomputer 26, the first feeding device and the second feeding device uniformly send the to-be-measured grains and the full grains to the first conveying device and the second conveying device respectively. The first conveying device and the second conveying device stably run at a preset speed (for example, 10 cm / s) under the accurate control of the single-chip microcomputer 26, and drive the grains to pass through the field of view of the first image acquisition device and the second image acquisition device.

[0103] The processor continuously receives the video image data transmitted from the first image acquisition device and the second image acquisition device. In order to quickly process the video image and ensure the integrity of the grain outline, the processor does not process every frame in the video, but sets a preset frame interval to extract key frames. For example, the processor extracts one frame as a key frame every 25 frames.

[0104] Step two, pre-processing of the extracted key frame image, including grayscale and Gaussian filter denoising; specifically including:

[0105] (1) Grayscale

[0106] The processor performs grayscale processing on the extracted color key frame image to remove color redundancy information, reduce computational complexity, and preserve the brightness features of the grain edges. The weighted average method is used to convert the RGB three channels to grayscale values, and the calculation formula is as follows:

[0107] I(x, y) = 0.299 · R(x, y) + 0.587 · G(x, y) + 0.114 · B(x, y)

[0108] where I(x, y) represents the brightness value of the pixel point (x, y) in the grayscale image, R, G, and B are the red, green, and blue channel values of the corresponding pixels in the original image. The converted image is a single-channel grayscale image, and the pixel value range is [0, 255].

[0109] (2) Gaussian filter denoising

[0110] To weaken the high-frequency noise in the image and smooth the grayscale image while preserving the edge information, the processor applies a two-dimensional Gaussian filter to the grayscale image. The Gaussian kernel function expression is:

[0111]

[0112] where σ is the standard deviation, which determines the filter strength; x and y are the relative coordinates of the kernel center; the filter process is performed by convolving the image with the kernel, and each pixel in the output image is the weighted average value of its neighborhood. The commonly used kernel size is 3x3 or 5x5, and the parameter σ can be set according to the image noise level.

[0113] Step three, the Canny edge detection algorithm is used on the preprocessed key frame image to accurately extract the edge contour information of the grain and form an edge map; the algorithm includes the following steps:

[0114] (1) Use the Sobel operator to calculate the gradient of each pixel point in the image in the horizontal and vertical directions, and then calculate its gradient amplitude and gradient direction. The gradient amplitude reflects the intensity of the pixel brightness change, and the gradient direction represents the direction of the fastest brightness change. The Sobel convolution kernel is as follows:

[0115]

[0116] Convolve the image grayscale matrix with G x and G y respectively to get the horizontal gradient value I x and the vertical gradient value I y of each pixel point. Then calculate the gradient amplitude M(x, y) and direction θ(x, y) of each pixel:

[0117]

[0118] (2) Non-maximum suppression is performed on the gradient image. Specifically, traverse each pixel point and check its neighborhood pixels in the gradient direction. If the gradient amplitude of the point is not the local maximum value, suppress it (set it to 0) to refine the edge and remove the false edge response.

[0119] (3) Double threshold processing. Set a high threshold and a low threshold, mark the pixels with gradient amplitude higher than the high threshold as "strong edge", mark the pixels between the high and low thresholds as "weak edge", and directly suppress the pixels lower than the low threshold.

[0120] (4) Perform edge connection (hysteresis threshold processing). Traverse all "weak edge" pixels, if there is any "strong edge" pixel in its 8-neighborhood, keep the "weak edge" pixel, otherwise suppress it.

[0121] Finally, the processor converts each key frame image into a binary edge map, in which the grain edges are represented by pixel value 1 and the background area is 0, which can be used for subsequent contour extraction and splicing recognition.

[0122] Step four, in order to accurately identify the image contour of each complete grain, the processor analyzes the continuously extracted key frame edge map, and completes the reconstruction of the grain contour through matching, splicing and deduplication. The process includes the following steps:

[0123] (1) Edge contour extraction

[0124] The processor performs contour extraction operation on each edge map. For each contour, record its center coordinates (x i ,y i ), pixel area A i , and bounding box B i =[x min ,y min ,x max ,y max ].

[0125] (2) Grain motion prediction

[0126] Let the conveyor belt speed be v (unit: pixel / s), the camera frame rate be f (unit: frame / s), and the key frame extraction interval be N frames, then the theoretical displacement between two frames is:

[0127]

[0128] The processor predicts the approximate position of a contour in the k+1 frame based on this displacement, and sets a search area centered on this position for matching.

[0129] For example: if the center coordinates of a contour in the k frame are (x = 120, y = 300), the conveyor belt speed is 200 pixels / s, the camera frame rate is 30 frames / s, and the key frame interval is 25 frames, then the predicted position in the k+1 frame is:

[0130]

[0131] The expected position is y = 300 + 167 = 467, and the processor searches for a matching contour around this point.

[0132] (3) Contour stitching and deduplication

[0133] For two contours in two consecutive frames that are matched to belong to the same grain, the processor merges them into a new contour, as follows:

[0134] First, the contours are merged:

[0135] Let the contour of a grain in the kth frame be Its bounding box is Its pixel area is The matched contour in the k+1th frame is Its bounding box is Its pixel area is The pixel sets of the two contours are P i (k) and P j (k+1) .

[0136] The pixel set of the merged contour is:

[0137]

[0138] The pixel area of the merged contour is:

[0139]

[0140] The bounding box of the merged contour is the minimum bounding rectangle of the two bounding boxes:

[0141]

[0142] The processor merges the contours and updates their attributes in the above manner.

[0143] Next, deduplication is performed:

[0144] When the same grain is repeatedly identified and merged in multiple key frames, multiple overlapping contour candidates may be formed The processor deduplicates these contour candidates as follows:

[0145] (i) Determine whether the intersection over union (IoU) of the bounding boxes of multiple contours is greater than a threshold T IoU (such as 0.5);

[0146] (ii) If there is significant overlap between multiple contours, keep the contour with the largest pixel area:

[0147]

[0148] (iii) delete the rest of the overlapping contours, avoid duplicate counting.

[0149] (5) Complete contour reconstruction and labeling

[0150] The processor assigns a unique number ID to each successfully reconstructed grain contour i , and stores its parameters (such as center point, area, bounding box, contour coordinate set) in the memory structure for subsequent screening and counting.

[0151] Step five, for the complete grain contour that has been reconstructed and stored in step four, contour screening and classification counting is performed; specifically including the following steps:

[0152] (1) Contour screening

[0153] The processor iterates through each closed grain contour in memory, calculates the area of the pixels it encloses, denoted as:

[0154] A i = the number of pixels contained in contour i

[0155] Compare this area with the pre-set area threshold A min (eg 200 pixels):

[0156] If A i <A min , the contour is determined to be invalid (such as produced by dust, debris or image noise), discarded;

[0157] If A i ≥ A min , it is retained as a valid contour.

[0158] (2) Contour classification and counting

[0159] For the valid contours that pass the screening, the processor classifies and counts them according to the image source:

[0160] (i) If from the first image acquisition device (such as a video stream that identifies all rice grain contours), count the total number of valid contours as the total number of rice grains to be measured, denoted as:

[0161] n0 = the number of all valid contours (ii) If from the second image acquisition device (such as a video stream that identifies full grain), count the total number of valid contours as the number of full rice grains, denoted as:

[0162] n1 = the number of full grain contours

[0163] Step six, the processor receives the weight data sent by the weighing device, and calculates the setting rate R based on the statistical results above; the processor displays the setting rate and weight information through the webpage of the display module, and provides a download function of the detection data this time; specifically including:

[0164] (1) Weight data receiving: the processor receives the real-time weight data sent by the electronic balance 20 forwarded by the single-chip microcomputer 26.

[0165] (2) Setting rate calculation: based on the total grain number n0 and the full grain number n1 obtained by the statistical results in step five above, the processor performs the calculation of the setting rate R. The calculation formula of the setting rate R is:

[0166]

[0167] (3) Result display and download: the processor displays the calculated setting rate R (usually expressed in percentage) and the weight information received from the electronic balance 20 in real time or history through the webpage of the display module. The webpage interface design is intuitive, which is convenient for users to view and understand. At the same time, the system provides a download function of the detection data this time, allowing users to download all related data including n0, n1, R and weight information to the local storage device in a standard format (such as CSV, Excel, etc.), so as to further analyze or archive the data.

[0168] The automatic measurement method of the rice grain setting rate provided by the application effectively replaces the traditional setting rate measurement method relying on manual sorting and counting, significantly reduces the labor cost, and improves the detection efficiency and accuracy. The whole measurement process is completed by the image acquisition device, the air separation device and the weighing module. The processor automatically completes the grain recognition, classification counting and setting rate calculation through image processing and calculation logic. Users only need to put the rice grains to be measured into the equipment, without manual screening and placement, so as to quickly obtain the key data such as setting rate, total grain number, full grain number and weight. The method is simple in operation, high in automation and strong in applicability, and is especially suitable for rice breeding selection, high-throughput measurement and quality evaluation, etc. application scenarios, which provides an efficient and reliable technical means for scientific researchers and agricultural workers.

[0169] The above detailed the preferred embodiments of the application. It should be understood that those skilled in the art can make many modifications and changes without creative labor based on the concept of the application. Therefore, any technical solution obtained by logical analysis, reasoning or limited experiment based on the prior art according to the concept of the application shall be within the protection scope determined by the claims.

Claims

1. An automatic measuring device for rice grain setting rate, characterized in that: include Frame, first feeding device, first conveying device, first image acquisition device, air separation device, second feeding device, second conveying device, second image acquisition device, weighing device, control device; Among them, it is used to support the first feeding device, the first conveying device, the first image acquisition device, the air separation device, the second feeding device, the second conveying device, the second image acquisition device, the weighing device and the single chip computer; The first feeding device is configured to uniformly feed the rice grains to be tested to the first conveying device, the first conveying device being disposed below the first feeding device, the air separation device being disposed at the end of the first conveying device to perform air separation on the rice grains to be tested on the first conveying device to separate full grains from empty grains; the first image acquisition device is disposed above the first conveying device; The second feeding device is arranged below the winnowing device, and is used to collect full grains and convey them to the second conveying device, which is arranged below the second feeding device; the second image acquisition device is arranged above the second conveying device; and the load-bearing device is arranged at the end of the second conveying device. The control device is connected to the first feeding device, the first conveying device, the first image acquisition device, the air separation device, the second feeding device, the second conveying device, the second image acquisition device and the weighing device.

2. The automatic measuring device for rice grain setting rate according to claim 1, characterized in that: The control device includes a single-chip microcomputer, a processor, and a display module. The single-chip microcomputer is connected to the first feeding device, the second feeding device, the first conveying device, the air separation device, the second conveying device, and the weighing device, and is used to control the operation of the first feeder, the second feeder, the first conveying device, the air separation device, and the second conveying device, and receive weighing data from the weighing device; The processor is connected to the first image acquisition device, the second image acquisition device, and the single-chip microcomputer, respectively, and is used to receive weighing data transmitted by the first image acquisition device, the second image acquisition device, and the single-chip microcomputer, perform grain recognition and grain number counting on the received image data using image processing technology, and calculate the rice grain setting rate based on the statistical results; The display module is connected to the processor and is used for displaying the rice grain setting rate and weighing results.

3. The automatic measuring device for rice grain setting rate according to claim 1, wherein: The first feeding device includes a feeding funnel and a first feeder. The feeding funnel is arranged above the first feeder and is provided with a square opening.

4. The automatic measuring device for rice grain setting rate according to claim 1, wherein: The first image acquisition device includes a first camera and a first camera bracket, the first camera bracket is fixed on the frame, and the first camera is installed on the first camera bracket and is arranged above the first conveying device.

5. The automatic measuring device for rice grain setting rate according to claim 1, wherein: The first conveying device includes a first DC reduction motor, a first synchronous belt transmission device, a first conveyor belt, a first driving shaft, a first driven shaft and a first shaft support block; Wherein, the first conveyor belt is sleeved on the first driving shaft and the first driven shaft; The first DC reduction motor is connected to the first driving shaft through the first synchronous belt transmission device to drive the first conveyor belt to move; The first shaft support block is fixed on the frame and is used to support the first driving shaft and the first driven shaft.

6. The automatic measuring device for rice grain setting rate according to claim 1, characterized in that: The second feeding device includes a collecting funnel and a second feeder. The collecting funnel is arranged below the air separation device. The second feeder is arranged below the collecting funnel. The collecting funnel has a square opening.

7. The automatic measuring device for rice grain setting rate according to claim 1, characterized in that: The second conveying device includes a second DC reduction motor, a second synchronous belt transmission device, a second conveyor belt, a second driving shaft, a second driven shaft and a second shaft support block; Wherein, the second conveyor belt is sleeved on the second driving shaft and the second driven shaft; The second DC reduction motor is connected to the second driving shaft through the second synchronous belt transmission device to drive the second conveyor belt to move; The second shaft support block is fixed on the frame and is used to support the second driving shaft and the second driven shaft.

8. The automatic measuring device for rice grain setting rate according to claim 1, wherein: The air separation device includes a centrifugal fan and an air duct device, the centrifugal fan is arranged on a frame, the air duct device is arranged at the air outlet of the centrifugal fan, the air outlet of the air duct device is arranged below the end of the first conveying device, and the air outlet of the air duct device is inclined along the free fall direction of the grains, and the air outlet is inclined upward by 15 degrees relative to the horizontal direction.

9. A method for automatically measuring rice grain setting rate, characterized in that: Using the automatic measuring device for rice grain setting rate according to any one of claims 1 to 8 specifically comprises the following steps: The processor is respectively connected to the first image acquisition device, the second image acquisition device and the single chip microcomputer for receiving the first video image from the first image acquisition device, the second video image from the second image acquisition device, and the weighing data sent by the single chip microcomputer; Extracting image frames from consecutive frames in the first video image and the second video image at set intervals, and performing preprocessing, edge detection, and dynamic image splicing on the extracted image frames, thereby reconstructing and storing complete grain contour information in a memory; For the reconstructed and stored complete grain outlines, outline screening and classification counting are performed; The processor receives the weight data sent by the weighing device and calculates the fruiting rate R=n1 / n0 based on the above statistical results; the processor displays the calculated fruiting rate and weight information through the web page of the display module and provides a download function for the test data.

10. The method for automatically measuring rice grain setting rate according to claim 9, wherein: Extracting image frames from continuous frames in the first video image and the second video image at set intervals, and performing preprocessing, edge detection, and dynamic image splicing on the extracted image frames, thereby reconstructing and storing complete grain contour information in a memory, specifically comprising the following steps: Extracting a frame of image from the video image at intervals of a preset number of frames as a key frame, and performing preprocessing on the key frame, including grayscale conversion and Gaussian filtering denoising; The Canny edge detection algorithm is used on the processed key frame images to accurately extract the edge contour information of the grains and form an edge map; The edge maps of two continuously extracted key frames are compared, and the theoretical displacement of the grain in the image is calculated based on the conveyor belt speed and frame rate. Matching and identification are performed by calculating the intersection-over-union ratio between edge contours within the prediction area. For successfully matched contour segments, geometric alignment and contour splicing are performed, and overlapping contours are deduplicated to reconstruct a complete closed contour image of the grain in memory, and each reconstructed contour is assigned a unique label.

Images