CN-121990122-A - Automatic water surface garbage identification and cleaning device and method
Abstract
The invention discloses an automatic water surface garbage identification and cleaning device and method, which relate to the technical field of environmental protection equipment and comprise a ship body platform, a sensing system, a calculation and control system, a garbage cleaning execution system and an energy and return system, wherein an optical sensing module in the sensing system is used for acquiring water surface target data, a positioning module is used for positioning the ship body, a water surface garbage identification model is deployed in the calculation and control system and is used for processing data acquired by the optical sensing module in real time and outputting garbage identification information, the garbage cleaning execution system is controlled by the calculation and control system and is used for executing garbage salvaging and collecting operation, and the energy and return system is used for monitoring the electric quantity of a device and the garbage bin capacity and carrying out return judgment and execution according to the electric quantity of the device and the garbage bin capacity, so that the problem that full-process automation and intelligent operation from garbage identification, salvaging to endurance management can not be realized in the prior art can be solved.
Inventors
- DONG BIN
- AN HAOYU
- Peng Congge
Assignees
- 安徽农业大学
Dates
- Publication Date
- 20260508
- Application Date
- 20260211
Claims (10)
- 1. An automatic water surface garbage identifying and cleaning device, which is characterized by comprising: The ship body platform is provided with a power steering system and a garbage storage bin; The sensing system is arranged on the ship body platform and comprises an optical sensing module, a positioning module and an obstacle avoidance module, wherein the optical sensing module is used for acquiring water surface target data, and the positioning module is used for positioning the ship body; The computing and control system is provided with a water surface garbage identification model and is used for processing data acquired by the optical sensing module in real time and outputting garbage identification information; The garbage cleaning execution system is arranged on the ship platform and is controlled by the calculation and control system and used for executing garbage salvaging and collecting operation; the energy and return system comprises a battery pack and a battery management system and is used for monitoring the electric quantity of the device and the capacity of the garbage bin, and carrying out return judgment and execution according to the electric quantity of the device and the capacity of the garbage bin.
- 2. The automatic water surface garbage identification and cleaning device according to claim 1, wherein the optical sensing module comprises a spectral imaging sensor for acquiring reflectivity data of a water surface target on a plurality of continuous or discrete wavebands to acquire raw hyperspectral image data.
- 3. The automatic water surface garbage recognition and cleaning device according to claim 2, wherein the execution of the calculation and control system comprises the steps of: Preprocessing the original hyperspectral image data; Extracting characteristic wave bands from the preprocessed image; inputting the image with the extracted features into the water garbage identification model to obtain a pixel level segmentation map or a target level detection frame.
- 4. The apparatus according to claim 3, wherein the characteristic band extraction is performed by at least one of principal component analysis, independent component analysis, and linear discriminant analysis.
- 5. The automatic water surface garbage recognition and cleaning device according to claim 1, wherein the water surface garbage recognition model is obtained through training by the following steps: Acquiring hyperspectral image samples of a marked data set and an unlabeled data set, and generating corresponding labels for the hyperspectral image samples of the marked data set; performing primary data augmentation on the labeled data set containing the label, and performing K times of data augmentation on the labeled data set without the label; predicting the label-free data after K times of augmentation through a model, and calculating average probability distribution to generate a pseudo label, and sharpening the pseudo label; Mixing the amplified labeled data with the sharpened unlabeled data, and generating mixed data; Calculating cross entropy loss of the labeled data and mean square error loss of the unlabeled data, and weighting and summing the cross entropy loss and the mean square error loss to obtain an overall loss function; And optimizing model parameters based on the integral loss function until the model converges.
- 6. The automatic water surface garbage identification and cleaning device of claim 1, wherein the computing and control system is further configured to: Generating and dynamically updating a water surface garbage distribution thermodynamic diagram based on the identification information of the continuous multiframes and positioning data provided by the positioning module; and carrying out real-time planning and adjustment on a cleaning route based on the water surface garbage distribution thermodynamic diagram and the environmental information provided by the obstacle avoidance module.
- 7. The automatic water surface garbage identification and cleaning device according to claim 6, wherein the water surface garbage distribution thermodynamic diagram is mapped in a geographic coordinate system, and the distribution density, class or cleaning priority of garbage is characterized by using different colors or numerical gradients.
- 8. The automatic water surface garbage recognition and cleaning device according to claim 1, wherein the obstacle avoidance module comprises a millimeter wave radar.
- 9. The automatic water surface garbage identification and cleaning device according to claim 1, wherein the energy source and return system is configured to: Monitoring the residual electric quantity of the battery pack and the loaded capacity of the garbage storage bin in real time; when the residual electric quantity is lower than a first preset threshold value or the loaded capacity reaches a second preset threshold value, determining that the return voyage needs to be executed; triggering the calculation and control system to plan an optimal path for returning to a preset base, and controlling the power steering system to drive the hull platform to return to the ship along the path.
- 10. A method for automatically identifying and cleaning water surface garbage, which is applied to the device as claimed in any one of claims 1 to 9, and comprises the following steps: Collecting water surface target data through an optical sensing module; the collected data is processed in real time through a water surface garbage identification model in the calculation and control system, and garbage identification information is output; generating and updating a water surface garbage distribution thermodynamic diagram according to the identification information of the continuous multiframes and the positioning data of the positioning module; Fusing thermodynamic diagrams with environment perception information of the obstacle avoidance module, and planning a cleaning route; Controlling a ship body platform to navigate along a planned route, and controlling a garbage cleaning execution system to execute salvage operation in a garbage dense area; the operation state is monitored through the energy and the return system, and the control device automatically returns when the return condition is met.
Description
Automatic water surface garbage identification and cleaning device and method Technical Field The invention relates to the technical field of environmental protection equipment, in particular to an automatic water surface garbage recognition and cleaning device and method. Background Along with the promotion of the urban process and the enhancement of human activities, the problem of water surface floating garbage pollution in water areas such as rivers, lakes, reservoirs and the like is more serious. Such garbage can destroy water landscapes and ecological environments, and can also threaten shipping safety, water ecological balance and coastal facilities. Currently, the cleaning of water surface garbage is mainly carried out by means of manual salvaging and semi-automatic ship cleaning, wherein manual salvaging is carried out by workers driving ships to perform visual searching and manual salvaging operations. The mode has the obvious defects of high labor intensity, low operation efficiency, limited coverage range, high cost, higher safety risk of personnel in complex water surface environment operation and the like, and is difficult to meet the requirement of large-scale and continuous garbage cleaning. Semi-automated cleaning vessels are typically equipped with simple collection devices such as conveyor belts, robotic arms, etc., but their identification and decision making capabilities are highly dependent on manual remote control or preset fixed airlines. They lack intelligent perceptibility to the water surface environment, cannot autonomously identify garbage in real time and dynamically plan the optimal cleaning path. Therefore, the operation efficiency, the intelligent level and the complex environment adaptability are still insufficient. In recent years, with the development of machine vision and artificial intelligence technology, some schemes for attempting to use a camera to identify garbage on the water surface are presented. However, when the technology based on the common RGB visible light imaging is used for dealing with complex and changeable water surface environments, the technology has fundamental limitations that objects with similar colors and textures but different materials, such as green plastic bottles, duckweed, dark plastic bags, shadows or sludge, are difficult to distinguish effectively, the environment robustness is poor, the recognition effect is extremely easy to be interfered by water surface reflection, ripple, weather change and illumination conditions, the recognition rate is unstable, and the false detection and omission rate is high. The problems cause that the prior art is difficult to realize high-precision, high-robustness and real-time identification of garbage under a complex water surface environment, and further cannot support full-automatic and intelligent cleaning operation in a true sense. In particular, there is an urgent need for a systematic solution that can reliably identify garbage, intelligently dynamically plan paths, effectively avoid obstacles, and implement autonomous endurance management in a complex environment. Disclosure of Invention The invention provides an automatic water surface garbage identification and cleaning device and method, which can solve the problem that full-flow automatic and intelligent operation from garbage identification, salvage to cruising management cannot be realized in water surface garbage cleaning in the prior art. In order to solve the above problems, the present invention provides an automatic water surface garbage recognition and cleaning device, comprising: The ship body platform is provided with a power steering system and a garbage storage bin; The sensing system is arranged on the ship body platform and comprises an optical sensing module, a positioning module and an obstacle avoidance module, wherein the optical sensing module is used for acquiring water surface target data, and the positioning module is used for positioning the ship body; The computing and control system is provided with a water surface garbage identification model and is used for processing data acquired by the optical sensing module in real time and outputting garbage identification information; The garbage cleaning execution system is arranged on the ship platform and is controlled by the calculation and control system and used for executing garbage salvaging and collecting operation; the energy and return system comprises a battery pack and a battery management system and is used for monitoring the electric quantity of the device and the capacity of the garbage bin, and carrying out return judgment and execution according to the electric quantity of the device and the capacity of the garbage bin. Compared with the prior art, the automatic water surface garbage identifying and cleaning device provided by the invention has the following beneficial effects: The intelligent operation monitoring device comprises a sensing system integratin