Detection system for in-cabin environment of port ship

Abstract

The invention is suitable for the technical field of environment scanning detection, and provides a detection system for the in-cabin environment of a port ship, which comprises a space information acquisition module, an in-cabin three-dimensional mapping module, a material identification and positioning module, a cabin cleaning equipment path planning module and a management and control software communication module. According to the detection system for the in-cabin environment of the port ship, through the arrangement of the space information acquisition module, the in-cabin three-dimensional mapping module, the material identification and positioning module, the cabin cleaning equipment path planning module and the management and control software communication module, the system encapsulates multiple functions, so that equipment carrying the scheme can be modularly controlled, and the detection efficiency is improved. Meanwhile, the system has good applicability, can directly utilize data, including three-dimensional mapping, object recognition and path planning, on the basis of collecting spatial information, integrates a communication function, performs information interaction with a master control system, and realizes modularization and intelligentization.

Description

Technical Field

[0001] The invention belongs to the technical field of environmental scanning detection, and in particular relates to a detection system for the cabin environment of a port ship. Background Art

[0002] Chinese patent CN116151714B discloses a three-dimensional visualization safety production control system (invention) for port yards, including the following parts: thermal image acquisition and splicing module, point cloud three-dimensional data processing module, thermal image material pile three-dimensional data processing module and three-dimensional visualization display module; according to the thermal imaging temperature monitoring module of the coal pile detection method and system as the temperature monitoring method, the obtained full-field multi-angle thermal image temperature map is transformed and spliced ​​through the thermal image temperature processing module. The material pile point cloud data is obtained by scanning the material pile by laser radar, and the data is sent to the three-dimensional material pile generation module to generate a three-dimensional model of the material pile. The data of the temperature processing module and the three-dimensional model data of the material pile are processed by the three-dimensional material pile thermal imaging processing module to perform layered thermal imaging data processing on the three-dimensional material pile, and finally the data is sent to the visualization module of the safety control system for three-dimensional display.

[0003] Its detection methods all have applicable scenarios, but there is no relevant technology for identifying the environment inside the cargo ship cabin at the port. In terms of the realization of the detection function, the schemes and technologies adopted are similar, such as target detection, SLAM, etc., but limited by the needs of the scene, it is impossible to directly utilize the data based on the collection of spatial information, nor can it perform three-dimensional mapping, object recognition and path planning, integrate communication functions, and it is unable to interact with the main control system to achieve modularization and intelligence. Summary of the invention

[0004] The present invention provides a detection system for the cabin environment of a port ship, aiming to solve the problems that the existing detection method is limited to the requirements of the scene and cannot directly utilize the data based on the collected spatial information.

[0005] The present invention is implemented as follows: a detection system for the cabin environment of a port ship includes a spatial information acquisition module, a cabin three-dimensional mapping module, a material identification and positioning module, a cabin cleaning equipment path planning module and a control software communication module;

[0006] The spatial information planning module realizes multi-modal spatial information collection, wherein a receiving laser radar and an RGB-D camera are provided for real-time data collection, from which key features such as landmarks and corner points are extracted to provide accurate basic spatial information for subsequent processes;

[0007] The in-cabin three-dimensional mapping module uses SLAM technology to establish a spatial coordinate system and determine the equipment's operating boundaries. By processing the collected multimodal data, the equipment's positioning in three-dimensional space and the preliminary construction of its motion trajectory are achieved.

[0008] Preferably, the material identification and positioning module processes image data collected by a camera and utilizes computer vision technology to achieve identification and positioning of the slag to be cleaned in the environment.

[0009] Preferably, the path planning module of the cleaning equipment combines the generated environmental map, the location of the slag to be cleaned and the obstacle avoidance algorithm, so that the equipment can plan a safe and efficient path, thereby realizing the path planning function of the cleaning equipment in the environment.

[0010] Preferably, the control software communication module realizes effective communication between various modules to ensure real-time data transmission and information sharing, so that the entire system can work in coordination, adopts a distributed communication architecture, and supports efficient and reliable data transmission and instruction scheduling.

[0011] Preferably, the material identification and positioning module is trained based on the YOLOv8 deep learning model.

[0012] Preferably, the global path planning of the cleaning equipment path planning module is based on map information and uses the HybridA* algorithm to search for the optimal path.

[0013] Preferably, the local path planning of the cleaning equipment path planning module is based on using an RRT algorithm to achieve rapid local path adjustment near the current position.

[0014] Preferably, the control software communication module uses WebSocket real-time communication technology to ensure that data with high real-time requirements can be transmitted quickly.

[0015] Beneficial Effects

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: a detection system for the cabin environment of a port ship of the present invention encapsulates multiple functions by setting a spatial information acquisition module, a cabin three-dimensional mapping module, a material identification and positioning module, a cabin cleaning equipment path planning module and a management and control software communication module. The system includes spatial information acquisition, target detection, three-dimensional mapping, path planning and interactive communication, so that the equipment equipped with this solution can be controlled modularly and has good applicability. It can directly utilize data on the basis of collecting spatial information, including three-dimensional mapping, object recognition and path planning, and integrates communication functions to interact with the main control system, thereby realizing modularization and intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a functional architecture diagram of the solution of the present invention;

[0018] Figure 2 It is a system workflow diagram of the present invention;

[0019] Figure 3 It is a schematic diagram of system data transmission of the present invention.

[0020] In the figure: 1-spatial information acquisition module, 2-cabin 3D mapping module, 3-material identification and positioning module, 4-cabin cleaning equipment path planning module, 5-control software communication module. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0022] See also Figure 1-3 , the present invention provides a technical solution: a detection system for the cabin environment of a port ship, including a spatial information acquisition module 1, a cabin three-dimensional mapping module 2, a material identification and positioning module 3, a cabin cleaning equipment path planning module 4 and a control software communication module 5;

[0023] The spatial information planning module 1 realizes multi-modal spatial information collection, which is equipped with a receiving laser radar and an RGB-D camera for real-time data collection, from which key features such as landmarks and corner points are extracted to provide accurate basic spatial information for subsequent processes;

[0024] The in-cabin three-dimensional mapping module 2 uses SLAM technology to establish a spatial coordinate system and determine the equipment's operating boundaries. By processing the collected multimodal data, the equipment's positioning in three-dimensional space and the preliminary construction of its motion trajectory are achieved.

[0025] In this embodiment, the functions of the spatial information acquisition module 1 include data standardization: formulating a general solution based on the currently common data interface, and realizing standardized data acquisition by developing data acquisition and processing modules for laser radar and RGB-D cameras;

[0026] Data preprocessing: After the collected data is preprocessed by denoising and filtering to improve the quality, the real-time collected spatial information is displayed on the software interface as original images and point cloud images;

[0027] Multimodal data fusion: realize the coordinate system conversion between LiDAR point cloud data and RGB-D camera images to ensure the accuracy of data fusion. Use advanced data fusion algorithms point cloud registration and image-point cloud fusion to ensure accurate alignment of multimodal data;

[0028] Feature extraction: Select the LiDAR landmark extraction algorithm and the image corner extraction algorithm to achieve the extraction and recognition of key features.

[0029] The functions of the in-cabin 3D mapping module 2 include SLAM algorithm integration: using the SLAM algorithm LIO-SAM suitable for multimodal data to achieve high-precision positioning and map construction, and debugging and optimizing the algorithm;

[0030] Build a three-dimensional map: Use the trajectory information generated by SLAM to build a three-dimensional environment map while ensuring the accuracy and real-time performance of the map.

[0031] Furthermore, the material identification and positioning module 3 processes the image data collected by the camera and utilizes computer vision technology to realize the identification and positioning of the slag to be cleaned in the environment.

[0032] Material identification and positioning module 3 is trained based on the YOLOv8 deep learning model.

[0033] In this embodiment, the main workflow of this module includes data preprocessing: preprocessing the collected image data, including denoising, adjusting brightness and contrast, and image size standardization, so as to improve the stability and effect of subsequent processing;

[0034] Model training: Build a relatively complete data set by collecting data on site or using public data. At the same time, select a deep learning model suitable for material recognition. Here, select YOLOv8, train it, and adjust the parameters to adapt to the recognition task.

[0035] Model integration and optimization: Optimize the trained models, such as pruning and quantization, to reduce the model size and improve the reasoning speed. At the same time, iteratively update the model based on feedback from actual application scenarios to ensure that the model can adapt to new material types and scenario changes, and conduct system-level testing to ensure that the material identification and classification module works together with other modules to achieve the expected results.

[0036] Furthermore, the path planning module 4 of the cleaning equipment combines the generated environmental map, the location of the slag to be cleaned and the obstacle avoidance algorithm, so that the equipment can plan a safe and efficient path, thereby realizing the path planning function of the cleaning equipment in the environment.

[0037] The global path planning of the cleaning equipment path planning module 4 is based on map information and uses the Hybrid A* algorithm to search for the optimal path.

[0038] The local path planning of the cleaning equipment path planning module 4 is based on the use of the RRT algorithm to achieve rapid local path adjustment near the current position.

[0039] In this embodiment, the cleaning equipment path planning module 4 also needs to integrate the path planning module into the entire system and perform system-level testing to ensure the accuracy and stability of the path planning.

[0040] Furthermore, the management and control software communication module 5 realizes effective communication between various modules to ensure real-time data transmission and information sharing, so that the entire system can work in coordination, adopts a distributed communication architecture, and supports efficient and reliable data transmission and instruction scheduling.

[0041] The management and control software communication module 5 uses WebSocket real-time communication technology to ensure that data with high real-time requirements can be transmitted quickly.

[0042] In this embodiment, the tasks of the control software communication module 5 include communication protocol integration: using the mass communication protocol MQTT to achieve low latency and efficient data transmission;

[0043] Build a message queue: used for asynchronous communication, reduce the coupling between modules, and improve system maintainability;

[0044] 5G communication: Connect network routers or switches, industrial-grade 5G CPE and other devices to achieve remote communication with other systems through a secure, highly reliable, and low-latency 5G network.

[0045] The working principle and use process of the present invention: After the present invention is installed, as an overall architecture, the present method is mainly used in the detection system of port clearance equipment. The existing equipment can integrate the various functional modules of the system according to the present method or design the system according to the present method. First, use spatial information acquisition equipment such as cameras and lidars to obtain environmental images and point cloud data, and transmit the environmental images to the material identification and positioning module 3 through corresponding information transmission channels, and transmit the point cloud data to the three-dimensional mapping module in the cabin to complete the transmission of multimodal data. Then, the three-dimensional mapping module 2 in the cabin and the material identification and positioning module 3 process the data respectively, and the two modules work simultaneously, and output the three-dimensional coordinate system of the surrounding environment and the object to be located respectively. The location of the cleaning target is transmitted to the cleaning equipment path planning module 4. The cabin three-dimensional mapping module 2 pre-processes the incoming data and uses LIO-SAM to generate a three-dimensional map of the cabin environment. The material identification and positioning module 3 is based on YOLOv8. After the image is pre-processed, the target object in the original image is marked and displayed with a colored frame, and the spatial position of the object relative to the camera is output at the terminal of the module. The cleaning equipment path planning module 4 will plan a suitable mobile cleaning strategy based on the incoming data, and output the result to the management and control software communication module 5. In addition to the planning strategy results, the intermediate information generated by other modules during the working process will also be transmitted to the management and control software communication module 5 in the form of logs.

[0046] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A detection system for the cabin environment of a port ship, characterized in that: It includes a spatial information acquisition module (1), a cabin three-dimensional mapping module (2), a material identification and positioning module (3), a cabin cleaning equipment path planning module (4) and a management and control software communication module (5); The spatial information planning module (1) realizes multi-modal spatial information collection, wherein a receiving laser radar and an RGB-D camera are provided for real-time data collection, from which key features such as landmarks and corner points are extracted to provide accurate basic spatial information for subsequent processes; The in-cabin three-dimensional mapping module (2) uses SLAM technology to establish a spatial coordinate system and determine the equipment operation boundary, and realizes the positioning of the equipment in the three-dimensional space and the preliminary construction of the motion trajectory by processing the collected multi-modal data.

2. A detection system for the cabin environment of a port ship as claimed in claim 1, characterized in that: The material identification and positioning module (3) processes the image data collected by the camera and uses computer vision technology to realize the identification and positioning of the slag to be cleaned in the environment.

3. A detection system for the cabin environment of a port ship as claimed in claim 1, characterized in that: The cleaning equipment path planning module (4) enables the equipment to plan a safe and efficient path by combining the generated environmental map, the location of the slag to be cleaned and the obstacle avoidance algorithm, thereby realizing the path planning function of the cleaning equipment in the environment.

4. A detection system for the cabin environment of a port ship as claimed in claim 1, characterized in that: The control software communication module (5) realizes effective communication between various modules to ensure real-time data transmission and information sharing, so that the entire system can work in coordination, adopts a distributed communication architecture, and supports efficient and reliable data transmission and instruction scheduling.

5. A detection system for the cabin environment of a port ship as claimed in claim 2, characterized in that: The material identification and positioning module (3) is trained based on the YOLOv8 deep learning model.

6. A detection system for the cabin environment of a port ship as claimed in claim 3, characterized in that: The global path planning of the cleaning equipment path planning module (4) is based on map information and uses the Hybrid A* algorithm to search for the optimal path.

7. A detection system for the cabin environment of a port ship as claimed in claim 3, characterized in that: The local path planning of the cleaning equipment path planning module (4) is based on the use of the RRT algorithm to achieve rapid local path adjustment near the current position.

8. A detection system for the cabin environment of a port ship as claimed in claim 4, characterized in that: The control software communication module (5) uses WebSocket real-time communication technology to ensure that data with high real-time requirements can be quickly transmitted.

Images