Intelligent pineapple harvesting machine based on visual recognition

Abstract

The present application relates to the technical field of pineapple harvesting, and discloses an intelligent pineapple harvesting machine based on visual recognition, which comprises a vehicle shell, a storage mechanism, a harvesting mechanism, a pineapple recognition system and a control system. The bottom of the vehicle shell is provided with self-adaptive row spacing wheels. The storage mechanism is installed on the vehicle shell. The harvesting mechanism comprises a lifting assembly and an elastic roller harvesting assembly. The elastic roller harvesting assembly is installed on the vehicle shell through the lifting assembly. The elastic roller harvesting assembly is used for cutting pineapples and conveying the pineapples to the storage mechanism. The pineapple recognition system is installed on the lifting assembly and is used for measuring the planting row spacing of the pineapples and identifying the positions of the pineapples. The control system is installed on the vehicle shell. The self-adaptive row spacing wheels, the lifting assembly, the storage mechanism, the elastic roller harvesting assembly and the pineapple recognition system are electrically connected with the control system. The present application can realize automatic and rapid pineapple picking, reduce the damage of the pineapples and improve the picking efficiency.

Description

Technical Field

[0001] This invention relates to the field of pineapple harvesting technology, and in particular to an intelligent pineapple harvesting machine based on visual recognition. Background Technology

[0002] Currently, the global pineapple planting area and yield are increasing year by year, and pineapple harvesting technology has become a major obstacle restricting the development of the pineapple planting industry. Existing pineapple harvesting machines include mechanical claw harvesting, roller or lever-feed harvesting, and double-row parallel pineapple harvesting. Among these, devices using mechanical claws as the harvesting mechanism are prone to damaging pineapple trees, and the concentrated force of the claw tips can easily damage the pineapple skin, resulting in slow harvesting efficiency. Roller or lever-feed harvesting can easily damage the plant or fruit. Double-row parallel pineapple harvesting requires manual operation, resulting in low efficiency and high labor costs. Furthermore, the wheels of traditional harvesters often travel on the ridge slopes due to varying planting ridge spacing, easily causing the harvester to tip over.

[0003] To address this, a visual recognition-based intelligent pineapple harvester is proposed. Summary of the Invention

[0004] The purpose of this invention is to provide a visual recognition-based intelligent pineapple harvester, which aims to solve or improve at least one of the above-mentioned technical problems.

[0005] To achieve the above objectives, the present invention provides the following solution: The present invention provides an intelligent pineapple harvester based on visual recognition, comprising: The vehicle body shell has adaptive spacing wheels mounted on its bottom. A storage mechanism, which is mounted on the vehicle body shell; A harvesting mechanism, comprising a lifting assembly and an elastic roller harvesting assembly, wherein the elastic roller harvesting assembly is mounted on the vehicle body shell via the lifting assembly; the elastic roller harvesting assembly is used to cut the pineapple and transport the pineapple to the storage mechanism. A pineapple recognition system, installed on the lifting assembly, is used to measure the row spacing of pineapple plantings and identify the position of the pineapples; The control system is mounted on the vehicle body shell, and the adaptive spacing wheels, the lifting assembly, the storage mechanism, the elastic roller harvesting assembly, and the pineapple recognition system are all electrically connected to the control system.

[0006] According to the present invention, a vision-based intelligent pineapple harvester includes an elastic roller harvesting assembly comprising: An upper fixing plate is installed on the lifting end of the lifting assembly; A lower fixing plate is installed at the bottom of the upper fixing plate via several connecting rods; The cutting part is mounted on the lower fixing plate; Two motors are provided, and the two motors are installed on both sides of the upper fixed plate. The output shaft of the motor is connected to a pineapple guide plate via a connecting rod. The motor is electrically connected to the control system. The rollers are provided in two, and the two rollers are slidably connected to the top surfaces of the lower fixed plate on both sides. A feeding interval is provided between the two rollers. A roller expansion device is installed between the rollers and the upper fixed plate. A slider is installed on the top of the rollers, and the two sliders are slidably connected to the two connecting rods respectively. A spring is mounted between the two sliders.

[0007] According to the present invention, an intelligent pineapple harvester based on visual recognition is provided, wherein the adaptive row spacing wheel includes two wheel shells, the two wheel shells are installed on the bottom sides of the vehicle body shell; two anti-sinking wheels are installed on the wheel shells side by side, and a wheel spacing adjustment component is installed between the two wheel shells; both the anti-sinking wheels and the wheel spacing adjustment component are electrically connected to the control system.

[0008] According to the present invention, a vision-based intelligent pineapple harvester is provided, wherein the wheel spacing adjustment component includes two strip-shaped brackets and four symmetrically arranged wheel connecting frames. Two wheel connecting frames are respectively installed on the opposite side walls of the two wheel shells, and two horizontally arranged wheel connecting rods are provided between the two wheel shells. The two ends of the wheel connecting rods are respectively slidably connected to the two wheel connecting frames. Two strip-shaped brackets are installed side by side at the bottom of the vehicle body shell. An intermediate connecting frame is fixedly installed between the two strip-shaped brackets. Several Y-shaped brackets are rotatably connected between the strip-shaped brackets and the wheel connecting frame. A hydraulic cylinder is installed between the strip-shaped brackets and the wheel connecting rod. The hydraulic cylinder is electrically connected to the control system.

[0009] According to the present invention, a vision-based intelligent pineapple harvester includes a receiving mechanism comprising: A pineapple storage box, which is installed on the vehicle body shell; A conveyor belt is mounted on the vehicle body shell and is located between the pineapple storage box and the feeding interval; the conveyor belt is electrically connected to the control system. A conveyor belt housing, which covers both sides of the conveyor belt.

[0010] According to the present invention, a vision-based intelligent pineapple harvester includes a first conveyor belt, a second conveyor belt, a third conveyor belt, and a horizontal conveyor belt sequentially mounted on the vehicle body shell. The feeding end of the first conveyor belt faces the feeding interval, the discharging end of the third conveyor belt faces the feeding end of the horizontal conveyor belt, and the discharging end of the horizontal conveyor belt faces the pineapple storage box. The first conveyor belt, the second conveyor belt, the third conveyor belt, and the horizontal conveyor belt are all electrically connected to the control system.

[0011] According to the present invention, a visual recognition-based intelligent pineapple harvester is provided, wherein the angle between the first conveyor belt and the horizontal plane is 13° to 17°. The angle between the second conveyor belt and the horizontal plane, and the angle between the third conveyor belt and the horizontal plane, are both 10° to 14°.

[0012] According to the present invention, a visual recognition-based intelligent pineapple harvester is provided, wherein the cutting part includes two symmetrically arranged blades, both blades are detachably connected to the bottom of the lower fixing plate, and the blades are located directly below the roller, with the vertices of the two blades in contact. The blade edge is provided with a first inclined side and a second inclined side, the included angle between the extension lines corresponding to the two first inclined sides is 33°, and the included angle between the extension lines corresponding to the two second inclined sides is 120°.

[0013] According to the present invention, a visual recognition-based intelligent pineapple harvester is provided, wherein the pineapple recognition system uses a camera.

[0014] According to the present invention, a vision-based intelligent pineapple harvester is provided, wherein the lifting assembly adopts a lead screw device.

[0015] The present invention discloses the following technical effects: This invention utilizes a pineapple identification system to measure the row spacing of pineapple plantings and identify the pineapple's position. The data is transmitted to a control system, which adjusts the spacing of the adaptive row-spacing wheels based on the actual row spacing. When the harvester reaches the approximate head position of the pineapple, the control system adjusts the height of the lifting assembly and the elastic roller harvesting assembly according to the pineapple's position. The elastic roller harvesting assembly cuts the pineapple and transports it to the receiving mechanism. After harvesting, the lifting assembly raises the height of the elastic roller harvesting assembly, allowing the pineapple to slide naturally into the receiving mechanism, reducing skin abrasion and improving quality. This invention enables fully automated and rapid pineapple harvesting, reduces damage, and improves harvesting efficiency. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the structure of the present invention; Figure 2 This is a schematic diagram of the harvesting mechanism in this invention; Figure 3 This is a schematic diagram of the wheel spacing adjustment component in this invention; Figure 4 This is a schematic diagram of the conveyor belt structure in this invention.

[0018] The components include: 1. Conveyor belt shell; 2. Conveyor belt; 21. First conveyor belt; 22. Second conveyor belt; 23. Third conveyor belt; 24. Horizontal conveyor belt; 3. Wheel shell; 4. Wheel spacing adjustment assembly; 41. Y-shaped bracket; 42. Strip bracket; 43. Intermediate connecting frame; 44. Hydraulic cylinder; 45. Wheel connecting frame; 46. Wheel connecting rod; 5. Anti-sinking wheel; 6. Harvesting mechanism; 7. Vehicle body shell; 8. Pineapple storage box; 9. Screw device; 10. Camera; 11. Upper fixing plate; 12. Roller expansion device; 13. Roller; 14. Pineapple guide plate; 15. Lower fixing plate; 16. Motor; 17. Spring; 18. Blade. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0021] Reference Figures 1-4 This invention provides a vision-based intelligent pineapple harvester, comprising: The vehicle body shell 7 has adaptive spacing wheels installed at its bottom; Storage mechanism, which is installed on the vehicle body shell 7; Harvesting mechanism 6 includes a lifting assembly and an elastic roller harvesting assembly. The elastic roller harvesting assembly is mounted on the vehicle body shell 7 via the lifting assembly. The elastic roller harvesting assembly is used to cut the pineapple and transport the pineapple to the storage mechanism. The pineapple recognition system is installed on the lifting assembly and is used to measure the row spacing of pineapple planting and identify the position of the pineapples. The control system is installed on the vehicle body shell 7. The adaptive spacing wheels, lifting components, storage mechanism, elastic roller harvesting components and pineapple recognition system are all electrically connected to the control system. With this setup, the present invention uses a pineapple recognition system to measure the row spacing of pineapple plantings and identify the position of the pineapples, transmitting the data to the control system. The control system adjusts the row spacing of the adaptive row spacing wheels according to the actual row spacing of the pineapple plantings. When the harvester reaches the approximate position of the pineapple head, the control system controls the lifting component to rise and fall according to the pineapple's position, adjusting the height of the elastic roller harvesting component. This height adjustment is based on the pineapple's growth position. The elastic roller harvesting component cuts the pineapple and transports it to the storage mechanism. After harvesting, the lifting component raises the height of the elastic roller harvesting component, allowing the pineapple to slide naturally into the rear storage mechanism, reducing pineapple skin wear and improving pineapple quality. The present invention enables fully automatic and rapid pineapple harvesting, reduces pineapple damage, and improves harvesting efficiency.

[0022] Further optimization of the scheme includes the following flexible roller harvesting components: Upper fixing plate 11 is installed on the lifting end of the lifting assembly; The lower fixing plate 15 is installed at the bottom of the upper fixing plate 11 via several connecting rods; The cutting section is mounted on the lower fixed plate 15. Two motors 16 are provided and are installed on both sides of the upper fixed plate 11. The output shaft of the motor 16 is connected to the pineapple guide plate 14 via a connecting rod. The motor 16 is electrically connected to the control system. There are two rollers 13, which are slidably connected to the top surfaces of the lower fixed plate 15 on both sides. There is a feeding interval between the two rollers 13. A roller expansion device 12 is installed between the rollers 13 and the upper fixed plate 11. A slider is installed on the top of the rollers 13, and the two sliders are slidably connected to the two connecting rods respectively. Spring 17 is installed between the two sliders; Motor 16 drives pineapple guide plate 14 to expand outward to a 120° angle, forming a funnel-shaped guide structure. When the pineapple enters the feeding interval between rollers 13, spring 17 is compressed, causing the spacing between rollers 13 to adaptively decrease (adjustment range 30-150mm), which can adapt to different pineapple sizes. Soft fixation is achieved through the friction of the silicone rollers 13. The cutting section completes the stem cutting. Motor 16 drives pineapple guide plate 14 to rotate inward, pushing the cut pineapple to move inward and entering the storage mechanism through the feeding interval of rollers 13 for storage. The rollers 13 can effectively prevent the citrus from swinging violently during cutting, improving cutting efficiency.

[0023] Further optimization of the scheme: the adaptive spacing wheel includes two wheel shells 3, which are installed on the bottom sides of the body shell 7; two anti-sinking wheels 5 are installed on the wheel shells 3 side by side, and a wheel spacing adjustment component 4 is installed between the two wheel shells 3. The anti-sinking wheels 5 and the wheel spacing adjustment component 4 are electrically connected to the control system; the anti-sinking wheels 5 are driven by an internal servo motor.

[0024] Further optimization of the scheme: the wheel spacing adjustment component 4 includes two strip brackets 42 and four symmetrically arranged wheel connecting brackets 45. Two wheel connecting brackets 45 are respectively installed on the opposite side walls of the two wheel shells 3. Two horizontally arranged wheel connecting rods 46 are provided between the two wheel shells 3. The two ends of the wheel connecting rods 46 are slidably connected to the two wheel connecting brackets 45 respectively. Two type brackets 42 are installed side by side at the bottom of the vehicle body shell 7. An intermediate connecting frame 43 is fixedly installed between the two type brackets 42. Several Y-type brackets 41 are rotatably connected between the strip bracket 42 and the wheel connecting frame 45. A hydraulic cylinder 44 is installed between the strip bracket 42 and the wheel connecting rod 46. The hydraulic cylinder 44 is electrically connected to the control system. In this embodiment, there are eight Y-shaped brackets 41. The height between the strip bracket 42 and the wheel connecting rod 46 is adjusted by the hydraulic cylinder 44. When the piston end of the hydraulic cylinder 44 extends, the angle between the Y-shaped bracket 41 and the ground increases, the distance between the far points of the two Y-shaped brackets 41 in the horizontal direction decreases, and the spacing between the two anti-sinking wheels 5 decreases. Conversely, the spacing between the two anti-sinking wheels 5 increases, thereby realizing the adjustment of the spacing between the two anti-sinking wheels 5.

[0025] Further optimization of the solution includes storage mechanisms such as: Pineapple storage box 8, which is installed on the vehicle body shell 7; Conveyor belt 2 is mounted on the vehicle body shell 7 and is located between the pineapple storage box 8 and the feeding interval; conveyor belt 2 is electrically connected to the control system. Conveyor belt housing 1 covers both sides of conveyor belt 2.

[0026] Further optimization of the scheme: the conveyor belt 2 includes a first conveyor belt 21, a second conveyor belt 22, a third conveyor belt 23, and a horizontal conveyor belt 24, which are sequentially installed on the vehicle body shell 7. The feeding end of the first conveyor belt 21 faces the feeding interval, the discharging end of the third conveyor belt 23 faces the feeding end of the horizontal conveyor belt 24, and the discharging end of the horizontal conveyor belt 24 faces the pineapple storage box 8. The first conveyor belt 21, the second conveyor belt 22, the third conveyor belt 23, and the horizontal conveyor belt 24 are all electrically connected to the control system. Baffles are provided on both sides of the first conveyor belt 21, the second conveyor belt 22, the third conveyor belt 23, and the horizontal conveyor belt 24 to prevent pineapples from falling out and to ensure pineapple harvesting efficiency. After the pineapples enter the conveyor belt 2, in order to prevent the pineapples from piling up and causing blockage at the transmission port, the conveyor belt 2 is designed to transport the pineapples to the right along the conveyor belt 2. The pineapples are then delivered to the high point at the rear of the pineapple collection box 8 by the multi-stage lifting tracks of the first conveyor belt 21, the second conveyor belt 22, the third conveyor belt 23 and the horizontal conveyor belt 24. This allows the pineapples to spread outwards under gravity, effectively preventing the pineapples from piling up, balancing the center of gravity of the overall harvester and maintaining driving stability.

[0027] In a further optimized design, the angle between the first conveyor belt 21 and the horizontal plane is 13° to 17°, and in this embodiment, it is preferably 15°. The angle between the second conveyor belt 22 and the horizontal plane, and the angle between the third conveyor belt 23 and the horizontal plane are both 10° to 14°, and are preferably 12° in this embodiment.

[0028] The design is further optimized so that the cutting part includes two symmetrically arranged blades 18. Both blades 18 are detachably connected to the bottom of the lower fixing plate 15 and the blades 18 are located directly below the roller 13. The apexes of the two blades 18 are in contact. The cutting edge of the blade 18 is provided with a first bevel and a second bevel. The angle between the extension lines corresponding to the two first bevels is 33° and the angle between the extension lines corresponding to the two second bevels is 120°. The pineapple stem is cut using two blades 18. Since the stem is prone to swaying and uneven force when the straight blade is used to cut it, two symmetrical blades 18 with included angles of 120° and 33° are used.

[0029] The design was further optimized by adopting a lead screw device 9 for the lifting assembly.

[0030] The pineapple recognition system has been further optimized by using camera 10.

[0031] Based on the above-described apparatus, the present invention provides an image processing method for a specific application, comprising: In image recognition tasks, image preprocessing is a crucial step to ensure the accuracy of subsequent segmentation and recognition. Since the original images often suffer from uneven lighting and complex backgrounds, this invention employs an adaptive histogram equalization (CLAHE) method based on the LAB color space to enhance image illumination. Specifically, the images captured by camera 10 are first scaled to a standard size and converted from the BGR color space to the LAB space. The luminance channel L is then processed using CLAHE to improve contrast in local areas. This method effectively enhances the detail representation of the target area, especially in low-light conditions or with severe background interference, improving the overall recognizability of the image.

[0032] After brightness enhancement, the image is converted back to BGR space and further converted to grayscale to reduce the computational complexity of subsequent processing. Then, a Gaussian filter is applied to smooth the image, reducing high-frequency noise while preserving edge structure information. To facilitate the extraction of boundary information of the target region, the Canny operator is used for edge detection in the processed image. This method is sensitive to subtle changes and can effectively depict the features of the pineapple's outer contour. This preprocessing step provides a clear and clean input image for image segmentation and deep learning recognition, improving the robustness of the overall system.

[0033] After initial image preprocessing, this paper employs a combination of threshold-based image segmentation and morphological operations to further extract the pineapple target region. First, the smoothed grayscale image is binarized using the Otsu adaptive thresholding method to automatically select the optimal threshold for initial separation of the foreground (pineapple) from the background. This method boasts strong adaptability and eliminates the need for manual parameter tuning, making it particularly suitable for image scenes with significant background variations. This initial segmentation effectively distinguishes the pineapple's main outline, providing candidate target regions for subsequent accurate identification.

[0034] Building upon binary images, this paper introduces morphological operations for post-processing to further optimize segmentation, suppress noise, and fill holes in the pineapple target region. Specifically, this includes two processes: opening (removing small background noise) and closing (connecting broken edges and filling holes). The integrity of the target region is enhanced by setting appropriate structuring element sizes. Furthermore, in some experiments, contour extraction algorithms (such as cv2.findContours) are used to locate the segmented regions to obtain target bounding boxes for subsequent object detection model annotation and training. This segmentation strategy demonstrates significant advantages in improving model training efficiency and localization accuracy.

[0035] Based on effective target region segmentation, this invention employs the YOLOv8 (You Only Look Once version 8) model, based on deep learning, for accurate identification and classification of pineapple targets. YOLOv8 is a high-performance one-stage detector in the current target detection field, possessing advantages such as lightweight design, high speed, and high detection accuracy. The model structure consists of three parts: Backbone, Neck, and Head, achieving effective detection of multi-scale targets by fusing feature information from different scales. This invention utilizes pre-processed images and corresponding labels to construct a training set and employs transfer learning to fine-tune the YOLOv8 model, thereby quickly adapting it to the pineapple recognition task.

[0036] During training, this invention uses the CIoU loss function to optimize the bounding box regression task and employs data augmentation strategies (such as random cropping, horizontal flipping, and color perturbation) to expand the training samples and improve the model's generalization ability. Test results show that the YOLOv8 model can accurately locate and identify pineapple targets in multiple test image scenarios, demonstrating good robustness and real-time performance. Compared to traditional image processing methods, the YOLOv8-based recognition process not only has higher accuracy but also adapts to complex backgrounds and targets of different shapes, making it suitable for applications such as agricultural product identification and automated harvesting.

[0037] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this invention, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0038] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. For those skilled in the art, other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively describe all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A visual recognition-based intelligent pineapple harvester, characterized in that, include: The vehicle body shell (7) has adaptive spacing wheels installed at its bottom; A storage mechanism is installed on the vehicle body shell (7); Harvesting mechanism (6), the harvesting mechanism (6) includes a lifting assembly and an elastic roller harvesting assembly, the elastic roller harvesting assembly is installed on the vehicle body shell (7) through the lifting assembly; the elastic roller harvesting assembly is used to cut the pineapple and transport the pineapple to the storage mechanism; A pineapple recognition system, installed on the lifting assembly, is used to measure the row spacing of pineapple plantings and identify the position of the pineapples; The control system is installed on the vehicle body shell (7), and the adaptive spacing wheels, the lifting assembly, the storage mechanism, the elastic roller harvesting assembly and the pineapple recognition system are all electrically connected to the control system; The elastic roller harvesting assembly includes: Upper fixing plate (11), the upper fixing plate (11) is installed on the lifting end of the lifting assembly; The lower fixing plate (15) is installed at the bottom of the upper fixing plate (11) by a number of connecting rods; A cutting section is mounted on the lower fixing plate (15); Two motors (16) are provided, and the two motors (16) are installed on both sides of the upper fixing plate (11). The output shaft of the motor (16) is equipped with a pineapple guide plate (14) through a connecting rod 2. The motor (16) is electrically connected to the control system. Two rollers (13) are provided, and the two rollers (13) are slidably connected to the top surfaces of the lower fixed plate (15) respectively. A feeding interval is provided between the two rollers (13). A roller expansion device (12) is installed between the roller (13) and the upper fixed plate (11). A slider is installed on the top of the roller (13), and the two sliders are slidably connected to the two connecting rods respectively. A spring (17) is mounted between the two sliders.

2. The intelligent pineapple harvester based on visual recognition according to claim 1, characterized in that: The adaptive spacing wheel includes two wheel shells (3), which are installed on the bottom sides of the vehicle body shell (7); two anti-sinking wheels (5) are installed on the wheel shells (3) side by side, and a wheel spacing adjustment component (4) is installed between the two wheel shells (3). The anti-sinking wheels (5) and the wheel spacing adjustment component (4) are electrically connected to the control system.

3. The intelligent pineapple harvester based on visual recognition according to claim 2, characterized in that: The wheel spacing adjustment assembly (4) includes two strip brackets (42) and four symmetrically arranged wheel connecting brackets (45). Two wheel connecting brackets (45) are respectively installed on the opposite side walls of the two wheel shells (3). Two horizontally arranged wheel connecting rods (46) are provided between the two wheel shells (3). The two ends of the wheel connecting rods (46) are slidably connected to the two wheel connecting brackets (45). Two strip-shaped brackets (42) are installed side by side at the bottom of the vehicle body shell (7). An intermediate connecting frame (43) is fixedly installed between the two strip-shaped brackets (42). Several Y-shaped brackets (41) are rotatably connected between the strip-shaped brackets (42) and the wheel connecting frame (45). A hydraulic cylinder (44) is installed between the strip-shaped brackets (42) and the wheel connecting rod (46). The hydraulic cylinder (44) is electrically connected to the control system.

4. The intelligent pineapple harvester based on visual recognition according to claim 1, characterized in that: The storage mechanism includes: Pineapple storage box (8), the pineapple storage box (8) is installed on the vehicle body shell (7); A conveyor belt (2) is installed on the vehicle body shell (7) and is located between the pineapple storage box (8) and the feeding interval; the conveyor belt (2) is electrically connected to the control system. A conveyor belt housing (1) is provided on both sides of the conveyor belt (2).

5. The intelligent pineapple harvester based on visual recognition according to claim 4, characterized in that: The conveyor belt (2) includes a first conveyor belt (21), a second conveyor belt (22), a third conveyor belt (23) and a horizontal conveyor belt (24) installed sequentially on the vehicle body shell (7). The feeding end of the first conveyor belt (21) faces the feeding interval, the discharging end of the third conveyor belt (23) faces the feeding end of the horizontal conveyor belt (24), and the discharging end of the horizontal conveyor belt (24) faces the pineapple storage box (8). The first conveyor belt (21), the second conveyor belt (22), the third conveyor belt (23) and the horizontal conveyor belt (24) are all electrically connected to the control system.

6. The intelligent pineapple harvester based on visual recognition according to claim 5, characterized in that: The angle between the first conveyor belt (21) and the horizontal plane is 13° to 17°; The angle between the second conveyor belt (22) and the horizontal plane, and the angle between the third conveyor belt (23) and the horizontal plane are both 10° to 14°.

7. The intelligent pineapple harvester based on visual recognition according to claim 1, characterized in that: The cutting section includes two symmetrically arranged blades (18). Both blades (18) are detachably connected to the bottom of the lower fixing plate (15). The blades (18) are located directly below the roller (13). The apexes of the two blades (18) are in contact. The cutting edge of the blade (18) is provided with a first inclined edge and a second inclined edge. The angle between the extension lines corresponding to the two first inclined edges is 33°, and the angle between the extension lines corresponding to the two second inclined edges is 120°.

8. The intelligent pineapple harvester based on visual recognition according to claim 1, characterized in that: The pineapple recognition system uses a camera (10).

9. The intelligent pineapple harvester based on visual recognition according to claim 1, characterized in that: The lifting assembly adopts a lead screw device (9).

Images