A method and device for real-time online detection of foreign matter in a round bale silage corn harvester

Abstract

The present application relates to a kind of disc cutting platform silage corn harvester foreign matter real-time online detection method and device, the device is by X Ray transmitter, X Ray receiving device, 5G intelligent signal transmitter, image processing and foreign matter identification system, real-time display and alarm system composition; X Ray receiving device includes photoelectric converter, circuit board, photosensitive rubber plate, realizes X Ray signal conversion and transmission;Image processing and foreign matter identification system includes filtering denoising module, image enhancement module, foreign matter identification module, realizes fast, high-precision identification foreign matter;Real-time display and alarm system includes real-time display screen, manual brake button, detection alarm, real-time transmission detection information, and send alarm.The present application solves the problem that non-metallic foreign matter such as stone cannot be detected, improves the detection rate and detection stability, avoids the damage of blade inside silage machine, prevents metal, stone and other foreign matter from entering the body of livestock with silage feed.

Description

Technical Field

[0001] This invention relates to the field of agricultural machinery technology, specifically to a method and device for real-time online detection of foreign objects in a disc header silage corn harvester. Background Technology

[0002] The field working environment of silage corn harvesters is complex. The harvested corn plants contain foreign objects such as metal and stones. These foreign objects enter the shredding device through the disc header. The blades of the shredding device are damaged when they collide with the foreign objects. Once the blades are damaged, the repair is complicated and expensive, which delays the harvest and causes economic losses.

[0003] Foreign objects such as metal and stones mixed in with corn plants can enter the digestive organs of livestock, such as the esophagus and stomach, when fed to them after harvest. This can damage the digestive system and even endanger the animal's life.

[0004] Currently, most foreign object detection methods for silage corn harvesters rely on electromagnetic and eddy current effects. However, these methods can only detect ferromagnetic foreign objects and cannot detect non-metallic foreign objects such as stones. Compared to current foreign object detection methods for silage corn harvesters, the use of… X To detect foreign objects entering the disc header of a silage harvester during harvesting, a real-time foreign object detection device for disc header silage corn harvesters was developed, utilizing high stability... X The different penetrating properties of X-rays to metal, stones, and corn plants allow for the real-time online detection of foreign objects by setting local grayscale values ​​for metal, stones, and corn plants, which has good engineering application value. Summary of the Invention

[0005] The purpose of this invention is to provide an accurate, efficient, and rapid detection method and real-time online detection device for identifying foreign objects such as metal and stones that enter a disc header silage corn harvester.

[0006] The first technical problem to be solved by this invention is that most current methods for detecting foreign objects in silage corn harvesters are based on electromagnetic effects and eddy current effects. These methods have great limitations on the types of foreign objects that can be detected, making it impossible to accurately detect non-metallic foreign objects such as stones. This device can detect foreign objects such as metals and stones mixed in with corn plants.

[0007] The second technical problem to be solved by this invention is that the disc header of the silage corn harvester is made of metal, and the detection of metal foreign objects using methods such as electromagnetic effect and eddy current effect is greatly interfered with. This detection device is placed 30-50mm behind the disc header to ensure that the detection area is not interfered with by the metal material of the silage corn harvester itself, and can ensure comprehensive detection of metal, stones and other foreign objects in the corn plants.

[0008] The third technical problem to be solved by this invention is that the current silage corn harvester has low efficiency and poor accuracy in identifying foreign objects. This application provides a real-time online image processing system for foreign object detection in a disc header silage corn harvester that integrates a filtering and noise reduction module, an image enhancement module, and a foreign object identification module, so as to achieve fast and high-precision identification of foreign objects.

[0009] To achieve the above technical objectives, the present invention utilizes X Since rays have different penetrating properties to different objects, a method and device for real-time online detection of foreign objects in a disc header corn harvester are provided, including:

[0010] A real-time online foreign object detection device for a disc header corn harvester, characterized in that the real-time online detection device includes... X X-ray emission device 1 X 4. X-ray receiving device; 5. 5G intelligent signal transmitter; 7. Real-time display and alarm system; 2. Rectangular support frame; 3. Clamping pressure plate; 19. Silicone gasket.

[0011] The X X-ray emitter 1 includes high stability X 12. X-ray source, 13. X-ray tube, 14. X-ray detector.

[0012] The X The X-ray receiving device 4 includes a photoelectric converter 17, a circuit board 16, and a photosensitive adhesive plate 15.

[0013] The 5G intelligent signal transmitter 5 is equipped with a high-efficiency ISM band radio frequency LoRa spread spectrum chip, which can modify various parameters such as serial port rate, transmission power, and transmission rate online, store digital signals and transmit them to a 315MHz radio wave channel. The on-board power supply of the silage corn harvester provides power to the 5G intelligent signal transmitter device.

[0014] The real-time display and alarm system 7 includes a real-time display screen 8, a manual braking button 9, a reset button 10, and a detection alarm 11.

[0015] As a preferred option, the rectangular support 2 is equipped with... X The X-ray emitter 1 and the rectangular support frame 2 are welded to the middle of the rear support beam of the disc cutter frame of the silage harvester. The length is 1000~1200mm, the width is 100~120mm, and the height is 50~60mm.

[0016] Preferably, two clamping pressure plates 3 are provided, each made of transparent polycarbonate sheet, for fixing... X X-ray emitter 1 and XThe X-ray receiving device 4 has a clamping plate 3 connected to the rectangular support frame 2 by bolts. The clamping plate 3 has a length of 1000~1200mm, a width of 100~120mm, and a thickness of 8~10mm.

[0017] Preferably, there are two silicone pads 19, made of nitrile rubber with a thickness of 10~12mm, which have good wear resistance, aging resistance and shock absorption.

[0018] The X X-ray emitter 1, including:

[0019] High stability X X-ray source 12 is used to emit high-stability radiation at energy levels of 70~90kV. X ray.

[0020] X-ray tube 13, X The channel through which the radiation source reaches the designated location.

[0021] X-ray detector 14, used for detection X If the radiation is emitted normally, and there is an abnormality, the radiation detector 14 transmits the information to the real-time display screen 8 by emitting radio waves and issues an alarm.

[0022] X X-ray receiving device 4 includes:

[0023] Photosensitive film 15 sensing X The image generated by the X-ray is transmitted to the photoelectric converter 17.

[0024] The upper part of the photoelectric converter 17 fits into the groove on the lower surface of the photosensitive adhesive 15, and the lower part fits into the groove on the upper surface of the circuit board 16, and is used to convert the light signal transmitted by the photosensitive adhesive 15 into an electrical signal.

[0025] The circuit board 16 transmits the collected electrical signals to the 5G smart signal transmitter 5 mounted on the circuit board 16.

[0026] A rectangular slot 18, located on the circuit board 16, is used to fix the 5G smart signal transmitter 5.

[0027] The 5G intelligent signal transmitter 5 is equipped with a high-efficiency ISM band radio frequency LoRa spread spectrum chip, which can modify various parameters such as serial port rate, transmission power, and radio frequency rate online, store electrical signals, and transmit them to a 315MHz radio wave channel. The image processing and foreign object recognition system 6 in the cab receives the radio waves of this channel and obtains the image to be detected. The entire signal transmission and foreign object recognition process takes less than 0.02 seconds. The maximum movement of the corn plant within 0.02 seconds is 200~300mm, which is more than enough to successfully identify the foreign object before it enters the chopping device.

[0028] Image processing and foreign object recognition system 6, including:

[0029] The filtering and denoising module is responsible for filtering and denoising the image to be detected. It compares the mean square error and peak signal-to-noise ratio (PSNR), objective evaluation metrics for image denoising, and employs a recursive filtering and denoising algorithm based on Formula 2. The output signal is obtained. Y , X Indicates input, Y 0 =X , Y Indicates the output, sets the number of iterations. n It is 10. γ Weighting coefficients are set during detection. γ The calculation formula is 20% of the denoised result is re-entered into the input, reducing noise in dynamic images and eliminating random noise in dynamic images.

[0030] The image enhancement module performs tonal enhancement processing on the image to improve image quality. By calculating the objective evaluation indicators of image information entropy and contrast, this invention adopts the multi-scale Retinex algorithm with adaptive weights. It uses four Gaussian functions with different scale parameter values ​​to perform convolution operations with the image to be detected, calculates the weighted average value of pixels in the Gaussian neighborhood of each pixel and outputs it, until the traversal is completed.

[0031] Calculated according to the formula for convolution average , n The number of images to be detected. σ 1 2 This is a scale parameter, with values ​​of 0.2, 0.4, 0.6, and 0.8. S The coordinates of all points in the data image. p Given the original image within the Gaussian template, sum the convolution outputs of each of the four channels, then divide by the total number of channels to obtain the average convolution value for each channel.

[0032] The Softmax function is used to weight the convolutional average of the four channels' outputs, with the weight allocation coefficients... ω i The calculation formula is The channel with more incident components is assigned a greater weight.

[0033] The final image data is It retains useful information that reflects the original characteristics of objects, thus improving the quality of image recognition.

[0034] The foreign object detection module detects data images and uses the grayscale values ​​of foreign objects such as metal and stones. It employs a grayscale-based template matching method, with a local matching area of ​​7-8 mm.2 The similarity threshold is set to 90%, and the similarity condition is that the gray value of the image to be detected is less than the gray value of the template image.

[0035] according to X The X-rays penetrate foreign objects such as metal and stones to obtain different grayscale values ​​of the image, and identify the image to be detected. When the similarity is greater than a set threshold, foreign objects such as metal and stones entering the disc header of the silage corn harvester are detected. The detection method has a response time of less than 0.02s.

[0036] The real-time display and alarm system 7 includes: a real-time display screen 8, a manual braking button 9, and a detection alarm 11.

[0037] Real-time display screen 8 is used to display in real time. X Dynamic images of foreign object detection under X-ray penetration.

[0038] Manual brake button 9 allows for manual braking to stop the vehicle.

[0039] Reset button 10, the detection device returns to its initial state.

[0040] The detection alarm 11 is used to remind the driver that foreign objects such as metal or stones have entered the disc cutter table, and at the same time cut off the power input to brake and stop the operation.

[0041] Local grayscale values ​​are the intensity information of local pixels in a grayscale image, that is, the color depth of a point in a black and white image. The range of grayscale values ​​is from 0 to 255, where 0 represents black and 255 represents white.

[0042] The beneficial effects of this invention are mainly reflected in the following aspects:

[0043] 1. When the disc header silage corn harvester is operating, all corn plants and foreign objects such as metal and stones will pass through it. X In the X-ray detection area, the cutting table will not interfere with the foreign object detection area, thus avoiding the influence of the harvester's own metal on the detection.

[0044] 2. X X-ray emitter and X The X-ray receiving device improves its anti-interference, shock resistance, and stability under the pressure of the clamping plate and the buffering effect of the silicone gasket.

[0045] 3. Utilizing high stability XThe penetration properties of X-rays differ for metals, stones, and corn plants. By setting local grayscale values ​​for metals, stones, and corn plants, foreign objects can be detected online in real time. This technology can detect foreign objects other than metals, such as stones. The development of foreign object detection technology for disc header silage corn harvesters can solve problems such as damage to cutting blades and other working parts caused by foreign objects entering the machine body during harvesting, and damage to the digestive system and even endangering life of livestock caused by foreign objects such as metals and stones mixed in with green fodder.

[0046] 4. The 5G intelligent signal transmitter transmits wireless signals in real time to the image processing and foreign object recognition system in the driver's cab. The system performs filtering and noise reduction, image enhancement, and foreign object recognition on the detected images. The detection response time is within 0.02 seconds, and the control from X-ray emission to braking and stopping is within 0.1 seconds. It can quickly and accurately identify foreign objects. Attached Figure Description

[0047] Figure 1 This is an isometric view of the testing device for a disc header silage corn harvester.

[0048] Figure 2 This is a schematic diagram illustrating real-time online detection, image processing, and recognition of foreign objects.

[0049] Figure 3 for X Schematic diagram of a radiation emitter.

[0050] Figure 4 for X Schematic diagram of a radiation receiving device.

[0051] Figure 5 This is a flowchart of an online foreign object detection image processing system.

[0052] Figure 6 This is a flowchart of a foreign object detection method.

[0053] Figure 7 for X Schematic diagram of X-ray detection of foreign objects.

[0054] Figure 8 This is a schematic diagram of a 5G smart signal transmitter.

[0055] Figure 9 This is a screenshot of the real-time online detection and display interface.

[0056] Figure 10 for X Schematic diagram of X-ray emission device installation.

[0057] Figure 11 This is a flowchart of the template matching algorithm.

[0058] Explanation of reference numerals in the attached figures:

[0059] 1. X 1. X-ray emission device; 2. Rectangular support frame; 3. Clamping pressure plate; 4. X 5. X-ray receiving device; 6. 5G intelligent signal transmitter; 7. Image processing and foreign object recognition system; 8. Real-time display and alarm system; 9. Real-time display screen; 10. Manual braking button; 11. Reset button; 12. Detection alarm; 13. High stability X 13. X-ray source; 14. X-ray tube; 15. X-ray detector; 16. Photosensitive film; 17. Circuit board; 18. Photoelectric converter; 19. Rectangular card slot; 10. Silicone pad. Detailed Implementation

[0060] The specific embodiments of the present invention will be described in detail below. To avoid excessive and unnecessary detail, well-known structures or functions will not be described in detail in the following embodiments. Unless otherwise defined, the technical and scientific terms used in the following embodiments have the same meaning as commonly understood by those skilled in the art to which this invention pertains.

[0061] according to X The penetration properties of X-rays vary for different objects, resulting in different light signal intensities generated on the photosensitive film 15. Based on this property, the local grayscale value of the obtained image is detected and compared with the template image to detect foreign objects such as metal and stones.

[0062] Example:

[0063] This detection device is installed on a disc header silage corn harvester to perform real-time online detection of foreign objects such as metal and stones.

[0064] To achieve the above objectives, a method and device for real-time foreign object detection in a disc header corn silage harvester are proposed. The specific implementation steps are as follows:

[0065] Step 1: Device Installation

[0066] Rectangular support frame 2 is welded to the middle of the rear support beam of the disc header frame of the forage harvester. A cavity is opened on the header base directly below it to hold various components. X X-ray emission device 1 X The X-ray receiving device 4 is covered by a clamping plate 3 and fixed with bolts. It is powered by an on-board power supply and the device is 30-50mm away from the disc.

[0067] Step 2: Capturing and Transmitting Optical Signals

[0068] The driver operates the disc header silage corn harvester, and the inverter converts the on-board DC power to 220V AC power to power the entire detection device. X The X-ray emitter emits energy levels of 70-90kV.X The X-ray passes through the clamping plate 3 to reach the detection area, penetrates the corn plant and foreign objects such as metal and stones, and finally shines onto the photosensitive film 15, which captures the X-ray. X The light signal energy of X-rays δ i , ,in α i Energy loss coefficient, r for X Radiation radius of rays β The attenuation coefficient of the radiation intensity. d To maintain the thickness of the clamping plate; the photoelectric converter 17 converts the light signal sensed by the photosensitive adhesive plate 15 into a photosensitive adhesive plate. δ i Converted into electrical signals E i , ,in ζ To improve photoelectric conversion efficiency, photoelectric converter 17 converts electrical signals... E i The signal is transmitted via the circuit board to the 5G smart signal transmitter 5, which then transmits the electrical signal. E i The signal was transmitted to a 315MHz channel, and the in-cab image processing and foreign object recognition system 6 received the electrical signal from that channel. E i And through AD conversion and encoding, the electrical signal is converted into an analog-to-digital converter. E i Converted to digital image signal X i The image to be detected is obtained.

[0069] Step 3: Filtering and noise reduction processing

[0070] The received data image is filtered and denoised using a recursive filtering and denoising algorithm, and the output signal is obtained according to Formula 2. Y , , X For input, Y Set the number of iterations for the output. n It is 10. γ Weighting coefficients are set during detection. γ The calculation formula is 20% of the denoised result is used as part of the input signal, and then compared with the input signal. X Combined with the new input signal for noise reduction, it affects subsequent noise reduction, reducing and eliminating random noise in dynamic images.

[0071] Step 4: Image Enhancement Processing

[0072] The image after filtering and denoising is enhanced using the multi-scale Retinex algorithm with adaptive weights. This algorithm employs four Gaussian functions with different scale parameter values, each applied to the image to be detected, and the result is calculated using the convolution average formula. , n The number of images to be detected. σ 1 2 This is a scale parameter, with values ​​of 0.2, 0.4, 0.6, and 0.8. S Given the coordinates of all points in the data image, calculate and output the weighted average of the pixels within the Gaussian neighborhood of a given pixel, continuing until the traversal is complete; use the Softmax function. Weights are assigned to the convolutional average of the outputs from the four channels with different scale parameter values. ω i The weighting coefficients are assigned, with channels having more incident components receiving a larger weight, ultimately resulting in image data that preserves the metallic appearance. The grayscale values ​​of the original characteristics of foreign objects such as stones.

[0073] Step 5: Foreign Object Identification and Processing

[0074] Foreign object identification processing is performed on the images after steps 2 and 3 using a template matching algorithm based on grayscale values, with a local matching area of ​​7-8 mm. 2 The similarity threshold is set to 90%, and the similarity condition is that the gray value of the image to be detected is less than the gray value of the template image. The template image and the image to be detected are input. When the similarity is greater than the set threshold, the foreign object is detected and the foreign object information is transmitted to the detection alarm 11 to issue an alarm. The response time of the detection method is less than 0.02s.

[0075] Step 6: Brake and stop

[0076] Power is cut off, the disc header silage corn harvester is automatically stopped, and the foreign object is marked on the display screen. The entire process from detecting the foreign object to braking to stop takes less than 0.1 seconds.

[0077] The operator presses the start button on the device. X The X-ray emitter 1 emits highly stable X-rays through the X-ray tube 13. X The transmitter is equipped with a radiation detector 14 to detect the radiation. X Whether the radiation is emitted stably.

[0078] X After passing through X-ray tube 13, the X-ray penetrates the target and reaches... X The photosensitive film 15 on the X-ray receiving device 4 senses... XThe light signal brought by the ray is transmitted to the photoelectric converter 17, which converts the light signal into an electrical signal. The electrical signal is connected to the 5G smart signal transmitter 5 mounted on the circuit board 16 through the circuit board 16. The 5G smart signal transmitter 5 is fixed on the circuit board 16 by a rectangular slot 18.

[0079] The 5G intelligent signal transmitter 5 transmits 315MHz radio waves, and the image processing and foreign object identification system 6 in the driver's cab receives the radio waves of the designated channel, obtains the electrical signal, and converts it into a digital image signal through AD conversion and encoding for image processing and foreign object identification.

[0080] The first step is to use a filtering and denoising module, which converts the digital signal into an image and then filters and denoises it. By comparing the mean square error and peak signal-to-noise ratio (PSNR), which are objective evaluation metrics for image denoising, this invention employs a recursive filtering and denoising algorithm. The denoised output is then re-entered into the input in a proportional manner, affecting the output of subsequent denoising, reducing noise in dynamic images, and eliminating irrelevant interference information in the image.

[0081] The image enhancement module utilizes a multi-scale Retinex algorithm with adaptive weights. It uses three Gaussian functions with different scale parameter values ​​to process the acquired image, calculating and outputting the weighted average of pixels within the Gaussian neighborhood of each pixel. This process continues until the end of the iteration. Based on the magnitude of the incident component, the algorithm then... The formula assigns weights to retain useful information that highlights the original features of the image, thereby improving image quality.

[0082] Finally, there's the foreign object detection module, which uses a template matching method based on... X The X-rays penetrate foreign objects such as metal and stones to obtain differences in local grayscale values ​​of the image, and identify the image to be detected. When the similarity is greater than a set threshold, foreign objects such as metal and stones entering the disc header of the silage corn harvester are detected. The detection method has a response time of less than 0.02s.

[0083] The detected images are transmitted to the real-time display screen 8 in real time for the operator to observe. When the foreign object recognition module detects foreign objects such as metal or stones, it will issue an alarm command to the vehicle display and alarm system, the detection alarm 11 will sound, the foreign object will be marked on the display screen, and the vehicle will be stopped at the same time.

[0084] The manual braking button 9 on the real-time display screen 8 allows for manual braking to stop the vehicle.

[0085] After the foreign object detection is completed, the operator clicks the reset button 10 on the real-time display screen 8, the alarm sound is turned off, and the system returns to its initial state.

[0086] The embodiments of the present invention have been described in detail above, but the content described is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the scope of the present invention should be included within the protection scope of the present invention.

Claims

1. A real-time online foreign object detection device for a disc header silage corn harvester, comprising: X X-ray emission device (1), X The X-ray receiving device (4), the 5G intelligent signal transmitter (5), and the real-time display and alarm system (7) are characterized by: X The X-ray emitter (1) includes high stability X X-ray source (12), X-ray tube (13), X-ray detector (14), clamping plate (3) are fastened. X The X-ray emitter (1) is fixed in the cavity between the rectangular support frame (2) and the clamping plate (3). A silicone gasket (19) with a thickness of 10-12 mm is laid at the bottom of the cavity. The rectangular support frame (2) is welded to the middle of the rear support beam of the disc cutter frame of the silage harvester. The length is 1000-1200 mm, the width is 100-120 mm, and the height is 50-60 mm. The clamping plate (3) is a transparent plate and is fastened to the rectangular support frame (2) by bolts. The length of the clamping plate (3) is 1000-1200 mm, the width is 100-120 mm, and the thickness is 8-10 mm. X The X-ray emitter (1) is powered by the DC power supply on the corn silage machine vehicle; X The X-ray receiving device (4) includes a photosensitive adhesive plate (15), a circuit board (16), and a photoelectric converter (17). The photosensitive adhesive plate (15) has a groove in the center of its lower surface. The upper part of the photoelectric converter (17) is embedded in the groove and connected to the photosensitive adhesive plate (15). The circuit board (16) has a groove in the center of its upper surface. The lower part of the photoelectric converter (17) is embedded in the groove and connected to the circuit board (16). The photoelectric converter (17) is 40-50mm long, 40-50mm wide, and 40-45mm high. The groove is 40-50mm long and 40-50mm wide. The circuit board has a rectangular slot (18) with a length of 20-30mm, a width of 10-15mm, and a depth of 4-6mm. The 5G smart signal transmitter (5) is fastened inside the rectangular slot (18) by the elastic pressure between it and the elastic pressure plate. The clamping pressure plate (3) is fastened. X X-ray receiving device (4), fixed on X Inside the cavity of the disc cutter base directly below the X-ray emitter (1), X The position of the X-ray receiving device (4) and X The X-ray emitter (1) is positioned vertically. X The distance between the X-ray receiving device (4) and the disc cutter is 90~100mm, the length is 1000~1200mm, the width is 100~120mm, and the height is 50~60mm. X X-ray emitter (1) and X The area between the X-ray receiving devices is the detection area, which has a length of 1000~1200mm, a width of 100~120mm, and a penetration height of 550~600mm.

2. The real-time online foreign object detection device for a disc header silage corn harvester according to claim 1, wherein the cutter of the silage corn harvester header is a disc cutter, the number of discs is 2, the harvesting width is 1.5~2.5m, the operating speed is 6~10km / h, and the matching power range is 150~200 horsepower.

Images