CN-121170722-B - Ship flow detection and analysis method based on laser radar

CN121170722BCN 121170722 BCN121170722 BCN 121170722BCN-121170722-B

Abstract

The invention discloses a ship flow detection and analysis method based on a laser radar, and relates to the technical field of laser radar positioning. The method comprises the steps of utilizing a historical point cloud detection data sequence of a laser radar to predict and obtain a predicted ship state distribution sequence of a target water area in a preset time window, optimizing a contemporaneous video monitoring angle sequence based on the predicted sequence, collecting video stream data according to an adaptive monitoring angle sequence, formulating an adaptive image dimension reduction mechanism according to current image interference intensity and the predicted ship state distribution sequence, carrying out data dimension reduction on the video stream data to obtain a key image frame sequence, monitoring and obtaining the current point cloud detection data sequence of the target water area in the preset time window through the laser radar, carrying out ship flow statistics and ship type identification by combining the key image frame sequence, and outputting a ship flow detection result sequence. The invention effectively improves the perception performance of ship flow detection and saves calculation force.

Inventors

WANG DENGCAI
JU CHUN
QIN QIN
CHEN YAO
DU XIANGJIN
DING SHUAI

Assignees

江苏长天智远交通科技有限公司

Dates

Publication Date: 20260505
Application Date: 20251119

Claims (7)

1. The method for detecting and analyzing the ship flow based on the laser radar is characterized by comprising the following steps of: Predicting and acquiring a predicted ship state distribution sequence of a target water area in a preset time window by utilizing a history point cloud detection data sequence of the laser radar; Optimizing a video monitoring angle in the preset time window based on the predicted ship state distribution sequence to obtain an adaptive monitoring angle sequence, and acquiring data of the target water area in the preset time window according to the adaptive monitoring angle sequence to obtain video stream data; an adaptive image dimension reduction mechanism is formulated according to the current image interference intensity and the predicted ship state distribution sequence, and data dimension reduction is carried out on the video stream data to obtain a key image frame sequence; Acquiring a current point cloud detection data sequence of a target water area in a preset time window through laser radar monitoring, carrying out ship flow statistics and ship type identification by combining the key image frame sequence, and outputting a ship flow detection result sequence; the method for predicting and acquiring the predicted ship state distribution sequence of the target water area in the preset time window by utilizing the history point cloud detection data sequence of the laser radar comprises the following steps: Based on a historical point cloud detection record of a target water area, acquiring a sample point cloud detection data sequence set by taking a preset data acquisition interval and a time span of a preset time window as characteristic constraints, and taking historical point cloud detection data sequences corresponding to different sample point cloud detection data sequences in a subsequent historical time window as sample prediction point cloud detection data sequences to acquire a sample prediction point cloud detection data sequence set; The sample point cloud detection data sequence set is used as input, the sample prediction point cloud detection data sequence set is used as supervision, and the long-time and short-time memory network is trained to converge to generate a point cloud data prediction engine; The point cloud data prediction engine is utilized to obtain a predicted point cloud detection data sequence in the preset time window according to the historical point cloud detection data sequence prediction; And extracting the ship state according to the predicted point cloud detection data sequence to obtain a predicted ship state distribution sequence, wherein the predicted ship state distribution comprises predicted ship quantity, predicted ship distribution dispersion and predicted ship navigation speed average.
2. The method of lidar-based ship flow detection analysis of claim 1, wherein optimizing the video surveillance angle acquisition adaptation surveillance angle sequence within the preset time window based on the predicted ship state distribution sequence comprises: Carrying out flow detection complexity analysis according to the predicted ship state distribution sequence, and setting a monitoring angle optimization depth sequence based on the predicted flow detection complexity coefficient sequence; setting an adaptive optimizing convergence number sequence according to the monitoring angle optimizing depth sequence, wherein the ratio of the monitoring angle optimizing depth to the preset standard optimizing depth is multiplied by the preset optimizing convergence number to obtain the adaptive optimizing convergence number; And in a ship monitoring simulation space of the target water area, based on the predicted ship state distribution sequence and the adaptive optimizing convergence frequency sequence, maximizing the information value as an optimizing target, performing iterative optimizing search on the video monitoring angle in the preset time window until the adaptive optimizing convergence frequency is reached, stopping optimizing, and outputting the adaptive monitoring angle sequence.
3. The method of lidar-based ship flow detection analysis of claim 2, wherein performing flow detection complexity analysis from the predicted ship state distribution sequence and setting a monitoring angle optimization depth sequence based on the predicted flow detection complexity coefficient sequence comprises: Randomly selecting first predicted ship state distribution from the predicted ship state distribution sequence, and acquiring first predicted ship quantity, first predicted ship distribution dispersion and first predicted ship navigation speed average value of the first predicted ship state distribution; And carrying out flow detection complexity assessment according to the first predicted ship quantity, the first predicted ship distribution dispersion and the first predicted ship navigation speed average value, outputting a first predicted flow detection complexity coefficient, and adding the first predicted flow detection complexity coefficient to a predicted flow detection complexity coefficient sequence, wherein the first predicted flow detection complexity coefficient is positively correlated with the first predicted ship quantity, the first predicted ship distribution dispersion and the first predicted ship navigation speed average value.
4. The method for detecting and analyzing the ship flow based on the laser radar according to claim 2, wherein in the ship monitoring simulation space of the target water area, based on the predicted ship state distribution sequence and the adaptive optimizing convergence number sequence, with the information value maximized as an optimizing target, performing iterative optimization search on the video monitoring angle in the preset time window, comprising: randomly selecting first predicted ship state distribution and acquiring corresponding first adaptive optimizing convergence times; Randomly selecting a first initial monitoring angle within a monitoring angle adjustment threshold of the video equipment; In a ship monitoring simulation space of the target water area, monitoring simulation is carried out according to the first predicted ship state distribution and the first initial monitoring angle, and a first monitoring simulation result is output; configuring video monitoring credible weight distribution based on the monitoring distance of the video equipment, wherein the video monitoring credible weight is inversely related to the monitoring distance; according to the video monitoring credible weight distribution, carrying out weighted fitting on the number of the ships covered in the first monitoring simulation result, and outputting a first information value; and continuing iterative optimization by taking the monitoring angle adjustment threshold value as an optimization space until the first adaptive optimization convergence times are reached, setting an initial monitoring angle corresponding to the maximum information value in the output optimization process as a first adaptive monitoring angle, and adding the first adaptive monitoring angle to the adaptive monitoring angle sequence.
5. A method of lidar-based marine flow detection analysis according to claim 3, wherein formulating an adaptive image dimension reduction mechanism based on the current image disturbance intensity and the predicted marine state distribution sequence comprises: Determining a current image interference intensity sequence of video acquisition based on an environmental monitoring data sequence evaluation of the target water area in the preset time window, wherein the environmental monitoring data at least comprises illumination conditions and meteorological conditions; Randomly selecting first predicted ship state distribution, and acquiring a corresponding first predicted flow detection complex coefficient and first current image interference intensity; Determining a first image dimension-reduction compensation coefficient based on the first predicted flow detection complexity coefficient and a first current image interference intensity weighted evaluation, wherein the first image dimension-reduction compensation coefficient and the first predicted flow detection complexity coefficient are in negative correlation and in positive correlation with the first current image interference intensity; And setting the product of the first image dimension reduction compensation coefficient and the preset standard image dimension reduction ratio as a first adaptive image dimension reduction ratio, and sequentially analyzing to obtain an adaptive image dimension reduction ratio sequence as an adaptive image dimension reduction mechanism.
6. The method for detecting and analyzing ship flow based on laser radar according to claim 5, wherein obtaining a current point cloud detection data sequence of a target water area within a preset time window through laser radar monitoring, performing ship flow statistics and ship type identification in combination with the key image frame sequence, and outputting a ship flow detection result sequence, comprises: Formulating an adaptive point cloud dimension reduction ratio sequence based on the adaptive image dimension reduction ratio sequence, wherein the sum of the adaptive image dimension reduction ratio and the adaptive point cloud dimension reduction ratio at the same time node is 1; performing data dimension reduction on the current point cloud detection data sequence according to the adaptive point cloud dimension reduction proportion to obtain a key point cloud data sequence; based on the deep convolutional neural network, carrying out ship flow statistics and ship type identification according to the key point cloud data sequence and the key image frame sequence, carrying out weighted fusion on the output result according to the adaptive fusion weight sequence, and outputting a ship flow detection result sequence.
7. The method for lidar-based ship flow detection analysis of claim 6, wherein the process of adapting the fusion weight sequence comprises: Randomly selecting a first adaptive image dimension reduction ratio and a first adaptive point cloud dimension reduction ratio under the same time node; multiplying the ratio of the preset standard image dimension reduction ratio to the first adaptive image dimension reduction ratio by a preset initial image weight to obtain a first adaptive image weight, wherein the preset initial image weight is 0.6, and the first adaptive image weight is more than or equal to 0.3 and less than or equal to 0.8; Subtracting the first adaptive image weight from 1 to obtain a first adaptive point cloud weight, and adding the first adaptive image weight and the first adaptive point cloud weight to the adaptive fusion weight sequence as a first adaptive fusion weight.

Description

Ship flow detection and analysis method based on laser radar Technical Field The invention relates to the technical field of laser radar positioning, in particular to a method for detecting and analyzing ship flow based on a laser radar. Background With the intelligent development of water traffic transportation, the demands of ship flow detection on efficiency and accuracy are continuously improved, and the comprehensive acquisition of ship position, type and other information can be realized by adopting a detection mode combining laser radar and machine vision in the prior art. However, the fluctuation of meteorological conditions such as illumination intensity, rain and fog can directly influence the image quality of machine vision and the detection stability of a laser radar, so that the detection precision is reduced, the perception performance is greatly fluctuated, and meanwhile, the prior art mostly adopts a fixed acquisition and processing strategy, so that a great amount of calculation force is wasted. Disclosure of Invention The application provides a ship flow detection and analysis method based on a laser radar, and aims to solve the technical problems of unstable ship flow detection and perception performance and calculation waste in the prior art. In view of the above problems, the present application provides a method for detecting and analyzing ship flow based on laser radar, comprising: Predicting and acquiring a predicted ship state distribution sequence of a target water area in a preset time window by utilizing a history point cloud detection data sequence of the laser radar; Optimizing a video monitoring angle in the preset time window based on the predicted ship state distribution sequence to obtain an adaptive monitoring angle sequence, and acquiring data of the target water area in the preset time window according to the adaptive monitoring angle sequence to obtain video stream data; an adaptive image dimension reduction mechanism is formulated according to the current image interference intensity and the predicted ship state distribution sequence, and data dimension reduction is carried out on the video stream data to obtain a key image frame sequence; and monitoring and acquiring a current point cloud detection data sequence of the target water area in a preset time window through a laser radar, carrying out ship flow statistics and ship type identification by combining the key image frame sequence, and outputting a ship flow detection result sequence. One or more technical schemes provided by the application have at least the following technical effects or advantages: The application provides a ship flow detection analysis method based on a laser radar, which realizes prospective guidance of a detection process by predicting ship state distribution, dynamically optimizes a video monitoring angle to improve data acquisition pertinence, intelligently formulates an image dimension reduction mechanism by combining environment interference, screens key image frames from a source, effectively reduces redundant data processing, and finally ensures recognition accuracy, simultaneously effectively improves the perception performance of ship detection in a complex environment and saves calculation resources by means of deep fusion of key point clouds and the key image frames. Drawings In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. Fig. 1 is a flow chart of a method for detecting and analyzing ship flow based on laser radar according to an embodiment of the present application. Detailed Description The application provides a ship flow detection and analysis method based on a laser radar, which is used for solving the technical problems of unstable ship flow detection sensing performance and calculation waste in the prior art. The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. It will be apparent that the described embodiments are only some, but not all, embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application. It should be noted that the terms "comprises" and "comprising" are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or server that comprises a list of steps or elements is not necessarily limited to those steps or elements expres