Automatic inspection robot for tiny pests and operation method

Abstract

The invention discloses an automatic inspection robot for tiny pests and an operation method, and relates to the technical field of agricultural robots. According to the invention, through cooperative work of the automatic focusing microscope lens and the binocular camera, microscopic image acquisition of tiny pests is realized, and imaging definition and detail capturing precision are ensured; an algorithm framework integrated by the control center supports multi-modal data fusion, pest types can be accurately identified, the quantity can be quantified, the growth period can be evaluated, and decision support can be provided; due to the integrated design of the movable chassis and the mechanical arm, the robot can operate autonomously in a complex farmland environment, and dynamic path adjustment and target tracking are supported; the method is based on a prevention and control strategy of accurate analysis, the pesticide application amount can be remarkably reduced, the chemical residue risk is reduced, and the method conforms to the green agriculture and sustainable development principle.

Description

Technical Field

[0001] This invention relates to the field of agricultural robot technology, specifically to an automatic inspection robot for tiny pests and its operating method. Background Technology

[0002] With the rapid development of precision agriculture technology, accurate detection and control of crop pests and diseases have become crucial for improving agricultural production efficiency and ensuring crop yield and quality. Traditional pest detection methods mainly rely on manual inspections or low-resolution image acquisition systems mounted on drones. These methods have significant limitations in identifying tiny pests (such as aphids, spider mites, and thrips) and their distribution on target parts such as leaves and flowers, and cannot achieve high-precision imaging and real-time analysis. In addition, existing pest identification technologies are mostly limited to single image processing algorithms, lacking comprehensive assessment of information such as pest quantity, life cycle (e.g., egg stage, larval stage, adult stage), and dynamic distribution. This results in insufficient targeting of control measures and easily leads to problems such as pesticide overuse, environmental pollution, or delays in control.

[0003] In recent years, the application of robotics technology in agriculture has become increasingly widespread. For example, agricultural robots used for large-scale pesticide spraying, mechanical weeding, or crop harvesting have gradually matured. However, existing agricultural robots are mainly designed for macro-scale tasks, and there are still significant technological gaps in the precise inspection and dynamic analysis of tiny pests.

[0004] Specifically, existing technical solutions have the following main problems:

[0005] 1. There is a lack of high-precision image acquisition and recognition systems specifically designed for tiny pests. Existing cameras or sensors have insufficient resolution to capture the microscopic features of pests, resulting in low detection accuracy.

[0006] 2. Existing agricultural robots typically employ fixed or rigid structures, lacking flexible target positioning and dynamic attitude adjustment capabilities, and are unable to adapt to the changing crop morphologies in complex farmland environments;

[0007] 3. Insufficient ability to analyze dynamic information such as the growth period, quantity, and distribution of pests makes it difficult to support real-time decision-making and the formulation of precise control strategies.

[0008] In view of the above problems, there is an urgent need to develop a robotic system and its operation method that can realize automatic inspection of tiny pests, precise target positioning, and multi-dimensional dynamic analysis, so as to meet the needs of precision agriculture for efficient and environmentally friendly pest control. Summary of the Invention

[0009] The purpose of this invention is to provide an automatic inspection robot for tiny pests and its operation method, so as to solve the technical problems mentioned in the background art.

[0010] To achieve the above objectives, the present invention provides the following technical solution: an automatic inspection robot for tiny pests, comprising a mobile chassis, a robotic arm, an autofocusing microscope lens, and a control center;

[0011] The mobile chassis is used to carry the robot body and supports automatic navigation or movement along a preset path to the vicinity of the crop-damaged area;

[0012] The robotic arm is equipped with a binocular camera for sensing and locating target parts such as leaves or flowers;

[0013] The autofocus microscope lens is mounted at the end of the robotic arm and is used to acquire high-resolution images of the target area.

[0014] The control center integrates leaf / flower target recognition algorithms and pest recognition algorithms. The control center is used to process the acquired images, count the number of pests, and analyze the growth period.

[0015] Furthermore, the mobile chassis is equipped with a GPS module and a lidar, thereby supporting real-time environment mapping and path planning, and adapting to various farmland terrains.

[0016] The robotic arm is a multi-degree-of-freedom robotic arm that supports precise attitude control and end-effector positioning.

[0017] The autofocus microscope lens supports dynamic focus adjustment and image stabilization, making it suitable for high-resolution imaging at the microscopic level.

[0018] A method for operating an automated inspection robot for minute pests includes at least the following steps:

[0019] S1: The automatic inspection robot for tiny pests moves to the vicinity of the crop-damaged area via a mobile chassis or a preset path.

[0020] S2: The robot arm uses a binocular camera at the end of its arm to sense and locate the target part of the leaf or flower.

[0021] S3: The robotic arm adjusts its posture and moves the autofocus microscope lens to the vicinity of the target;

[0022] S4: The autofocus microscope lens acquires high-resolution images of the target area and transmits them to the control center;

[0023] S5: The control center analyzes the image through the leaf / flower target recognition algorithm and the pest recognition algorithm, and outputs information such as the number of pests and their growth period.

[0024] Furthermore, the leaf / flower target recognition algorithm is based on a convolutional neural network and supports the output of target type and three-dimensional coordinates.

[0025] Furthermore, the pest identification algorithm combines a target detection network and a classification network to support multi-scale feature extraction and pest morphological analysis.

[0026] Compared with the prior art, the beneficial effects of the present invention are:

[0027] 1. High-precision detection capability: This invention achieves microscopic image acquisition of tiny pests through the coordinated work of an autofocus microscope lens and a binocular camera, ensuring image clarity and detail capture accuracy.

[0028] 2. Intelligent analysis function: The algorithm framework integrated in the control center of this invention supports multimodal data fusion, which can accurately identify pest species, quantify the quantity and assess the growth period, and provide decision support;

[0029] 3. Strong adaptability and flexibility: The integrated design of the mobile chassis and robotic arm in this invention enables the robot to operate autonomously in complex farmland environments, supporting dynamic path adjustment and target tracking.

[0030] 4. Environmental friendliness: Based on precise analysis, this invention provides a control strategy that can significantly reduce pesticide application and lower the risk of chemical residues, which is in line with the principles of green agriculture and sustainable development. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of 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.

[0032] Figure 1 This is a schematic diagram of the overall structure of the present invention;

[0033] Figure 2 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0034] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0035] The robot and method of this invention are applicable to the fields of precision agriculture, plant protection, and intelligent farmland management, enabling high-precision detection, positioning, and dynamic analysis of minute pests.

[0036] Example 1:

[0037] Please see Figure 1An automated inspection robot for tiny pests includes:

[0038] Mobile chassis: Used to support the overall structure of the robot, supporting automatic navigation or preset path planning based on sensor data, enabling the robot to move precisely to areas where crops may be at risk; the mobile chassis adopts a tracked or wheeled drive structure, equipped with a global positioning system (GPS) module and lidar sensor, supporting real-time environment mapping, path planning and autonomous navigation, and has a built-in high-capacity power battery, suitable for various farmland terrains such as mud and slopes;

[0039] The robotic arm is equipped with a binocular camera for real-time perception and positioning of target parts such as crop leaves and flowers, providing three-dimensional spatial coordinate information. The robotic arm is a multi-degree-of-freedom serial robotic arm with a binocular camera integrated into the end effector, supporting precise attitude control and end-effector positioning. The field of view is suitable for crop scanning, and the image resolution meets high-precision requirements.

[0040] Autofocus microscope lens: Fixedly mounted on the end effector of the robotic arm, used to acquire high-resolution microscopic images of the target area, supporting dynamic focus adjustment to adapt to imaging needs of different distances and sizes; the autofocus microscope lens is equipped with a high-speed autofocus module and image stabilization system, which can capture the morphological details of tiny pests, such as body size, antennae and other features.

[0041] Control Center: Integrates advanced leaf / flower target recognition algorithms and pest recognition algorithms. It is responsible for receiving image data, performing real-time processing, pest feature extraction, quantity statistics, and growth period assessment, and outputting analysis results. The control center is based on a high-performance embedded processor, integrates a deep learning framework, supports the real-time operation of algorithms, and is equipped with a wireless communication module for data transmission.

[0042] Example 2:

[0043] This embodiment proposes the following technical solution based on the above embodiments:

[0044] See Figure 2 An automated inspection method for minute pests based on the aforementioned robot, comprising the following steps:

[0045] The mobile chassis moves to the vicinity of potentially hazardous areas of crops via an automatic navigation system or pre-planned path, ensuring that the robot's positioning accuracy meets operational requirements; the mobile chassis uses LiDAR and GPS modules for environmental perception and path planning, or moves autonomously along a pre-planned path;

[0046] The binocular camera at the end of the robotic arm scans and senses the crops, identifies and locates target parts such as leaves or flowers, and generates spatial coordinate data of the target. The binocular camera performs a stereo scan, uses a deep learning-based target detection algorithm to identify the target parts, and calculates their three-dimensional spatial coordinates.

[0047] The robotic arm dynamically adjusts its posture based on the target coordinates to precisely move the autofocus microscope lens to the optimal imaging position near the target; the robotic arm also dynamically adjusts joint angles using inverse kinematics algorithms to achieve precise positioning.

[0048] The autofocus microscope lens acquires high-resolution microscopic images of the target area and transmits them to the control center in real time via a data transmission interface; the autofocus microscope lens automatically adjusts the focal length according to the target distance, and the acquired image data is transmitted via a high-speed data interface;

[0049] The control center uses leaf / flower target recognition algorithms and pest recognition algorithms to perform in-depth analysis on the acquired images, extracting key information such as pest species, quantity distribution, and growth period, and generating output reports. The control center runs an integrated algorithm to preprocess, extract features, and classify the images, including confirming pest species, counting their numbers, and determining their growth period. The analysis results are transmitted in real time to user terminal devices via a wireless communication module or stored in a local database.

[0050] The leaf / flower target recognition algorithm is based on a convolutional neural network (CNN) architecture. The training dataset covers leaf and flower image samples of various crops (such as rice, cotton, vegetables, etc.) and supports the output of target type, 3D coordinates and confidence scores.

[0051] Pest identification algorithms integrate target detection networks (such as the YOLO series) and residual classification networks (such as the ResNet series), support multi-scale feature extraction and attention mechanisms, adapt to pest images of different sizes and postures, identify a variety of tiny pest species, count their numbers, and assess their reproductive period through morphological analysis.

[0052] Based on the above embodiments one and two, a specific application description is proposed:

[0053] Taking the identification of mites on strawberry leaves in a facility horticulture setting as an example, the robot of this invention performs the operation according to the following process, with the specific parameters as follows: the positioning error of the mobile chassis is within 0.5 meters, and the battery life is not less than 8 hours; the working robotic arm has a six-degree-of-freedom structure, a maximum extension distance of 1.2 meters, an end-effector positioning accuracy of ±5 centimeters, a binocular camera with a field of view of 120° and a resolution of 1920×1080 pixels; the autofocus microscope lens has a magnification of 10-100 times and a resolution of not less than 2560×1920 pixels; the image processing speed of the control center is not less than 30 frames / second, the leaf / flower target recognition algorithm accuracy is over 95%, and the pest recognition algorithm accuracy is over 90%.

[0054] 1. The mobile chassis moves along a preset path or autonomously to the strawberry growing area within the greenhouse horticulture facility;

[0055] 2. A binocular camera scans the strawberry plant, and a target recognition algorithm is used to detect leaf targets suspected of being infested by mites, and their coordinates are calculated;

[0056] 3. The robotic arm adjusts its posture, moving the microscope lens to a position 5-10 cm above the leaf;

[0057] 4. The microscope lens acquires high-resolution images and transmits them to the control center;

[0058] 5. Analysis of images by the central nervous system confirmed that the pest was a red spider mite, with a number of approximately 30 individuals per square centimeter, and the main reproductive stage was the adult stage;

[0059] 6. The analysis results are wirelessly transmitted to the user's mobile terminal, and precise application suggestions are provided, such as targeted spraying of low-dose pesticides or biological control measures.

[0060] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. An automated inspection robot for tiny pests, characterized in that: The mobile chassis, the working mechanical arm, the auto-focusing microscope lens and the control center are included. The mobile chassis is used to carry the robot body, support automatic navigation or preset path movement to the vicinity of crop damage area. The working mechanical arm is equipped with binocular camera for sensing and positioning the target site such as leaf or flower. The auto-focusing microscope lens is installed at the end of the working mechanical arm, and is used to collect high-resolution images of the target site. The control center integrates leaf / flower target recognition algorithm and pest recognition algorithm, and is used to process the collected images, count the number of pests and analyze the growth period.

2. The micro-pest automatic inspection robot according to claim 1, characterized in that: The mobile chassis is equipped with GPS module and laser radar, thereby supporting real-time environment mapping and path planning, and adapting to various farmland terrains.

3. A method for operating a micro-pest automatic inspection robot, wherein the micro-pest automatic inspection robot is as claimed in any one of claims 1-2, and wherein the method comprises the following steps: The method comprises at least the following steps: ​ S1: The micro-pest automatic inspection robot moves to the vicinity of crop damage area through automatic navigation or preset path movement by the mobile chassis; S2: The binocular camera at the end of the mechanical arm senses and positions the leaf or flower target site; S3: The working mechanical arm adjusts the posture and moves the auto-focusing microscope lens to the vicinity of the target; S4: The auto-focusing microscope lens collects high-resolution images of the target site and transmits them to the control center; S5: The control center analyzes the images through the leaf / flower target recognition algorithm and pest recognition algorithm, and outputs the number of pests, growth period and other information.

4. The method of claim 3, wherein the method further comprises: detecting the micro-pests on the plant; and determining the type of the micro-pests. The leaf / flower target recognition algorithm is based on convolutional neural network, and supports the output of target type and three-dimensional coordinates.

5. The method of claim 3, wherein the method further comprises: detecting the micro-pests on the plant; and determining the type of the micro-pests. The pest recognition algorithm combines target detection network and classification network, and supports multi-scale feature extraction and pest morphology analysis.

Images