CN114693729A - Depth tracking identification method based on laser scanning - Google Patents
Depth tracking identification method based on laser scanning Download PDFInfo
- Publication number
- CN114693729A CN114693729A CN202210136111.XA CN202210136111A CN114693729A CN 114693729 A CN114693729 A CN 114693729A CN 202210136111 A CN202210136111 A CN 202210136111A CN 114693729 A CN114693729 A CN 114693729A
- Authority
- CN
- China
- Prior art keywords
- data
- point cloud
- scanning
- cloud data
- abnormal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
- G06T7/251—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments involving models
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/24—Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
Abstract
The application provides a depth tracking identification method based on laser scanning, which comprises the following steps: defining a key scanning area based on historical data of a canal danger, specifically comprising: acquiring and processing historical data of a water channel dangerous case, and dividing a scanning area of a water channel section according to the historical data; acquiring point cloud data of a water channel section limited area by using a three-dimensional laser scanner; scanning to obtain abnormal data in the point cloud data; identifying the dangerous case type according to the abnormal data in the scanning; dynamically tracking based on abnormal data and dangerous case types; and analyzing the abnormal point cloud data to generate dangerous case alarm information. The beneficial effects of the system include: the monitoring is carried out on the spot without detection personnel, the monitoring safety is improved, the sewer or the water channel can be timely found and identified when the dangerous case occurs through dynamic monitoring, the type of the dangerous case and the specific position of the dangerous case can be identified, a large amount of memory of the system is saved through selective scanning and data storage, and the high efficiency of the system operation is ensured.
Description
[ technical field ] A method for producing a semiconductor device
The invention relates to the technical field of intelligent laser scanning, in particular to a depth tracking identification method based on laser scanning.
[ background of the invention ]
Aiming at the difficult-to-enter environments such as a water channel and a sewer pipeline, portable scanning and small scene scanning play an important role.
The existing expressways and urban highways are mainly divided into three types, namely cement concrete roads, asphalt concrete roads, regenerated asphalt concrete roads and the like, a sewer or a water channel is arranged on the highways every hundreds of meters, so that water cannot be accumulated on the highways in severe weather such as heavy rainy days and the like, and smooth running of the highways is ensured, but the sewers can accumulate sundries and sludge in the sewers along with long-time use, the sewers can be blocked along with the increase of the sundries and the sludge, once the sewers are blocked, rainwater and sewage cannot be discharged out and then can overflow the highways, so that running of vehicles can be seriously influenced, at present, urban sewer pipe networks are complicated, and in the process of detecting the smooth running condition of the sewer pipe networks, liquid level conditions in the sewer pipe networks are generally detected by infrared detection equipment and laser detection equipment, and then analyzing and processing liquid level signals detected by the infrared detection equipment and the laser detection equipment to finally obtain the unblocked condition of the sewer pipe network. The liquid level condition in the current sewer pipe network is surveyed through the manual work, can not carry out real-time detection, and when sewer pipe network blockked up and lead to the liquid level to rise, maintainer can not discover earlier and in time handle, leads to sewer pipe network's maintenance efficiency lower. Therefore, the sewer needs to be monitored for dangerous cases and early warned in time, and the dangerous cases specifically comprise: some dangerous situations such as blockage, deformation, water flow, floaters, rat damage and the like need to be dynamically monitored for a long time to ensure the normal function of a sewer or a ditch.
[ summary of the invention ]
The depth tracking identification method based on laser scanning provided by the embodiment of the invention comprises the following steps:
defining a key scanning area based on historical data of the canal danger: acquiring and processing historical data of a water channel dangerous case, and dividing a scanning area of a water channel section according to the historical data;
acquiring point cloud data of a water channel section limited area by using a three-dimensional laser scanner;
scanning to obtain abnormal data in the point cloud data: correcting the overwater and underwater scanning data and acquiring an abnormal point cloud data set;
identifying the type of the dangerous case according to abnormal data in scanning: detecting foreign matters in the water channel, identifying types of the foreign matters, monitoring the form change of the water channel, and monitoring the abnormal water flow of the water channel;
and dynamically tracking based on abnormal data and dangerous case types: dynamic monitoring of foreign matters and abnormal water flow;
and analyzing the abnormal point cloud data to generate dangerous case alarm information.
Preferably, the canal danger-based historical data defines a key scanning area, including:
acquiring and processing historical data of a canal dangerous case;
and dividing the scanning area of the cross section of the water channel according to historical data.
Preferably, the acquiring and processing historical data of the canal danger comprises:
acquiring historical data of a dangerous case of a cross section of a water channel; the historical data comprises dangerous case types and point cloud data thereof, the frequency of dangerous case occurrence, position partitions of dangerous case occurrence and specific position coordinates in the partitions; the position subareas comprise the top, the left side surface, the right side surface, the bottom and the water surface of the cross section of the water channel;
extracting the characteristics of different dangerous cases in the historical data, establishing a standard dangerous case matrix, wherein row vectors are partitions at all positions, column vectors are dangerous case characteristics, and counting the types and times of the dangerous cases of the partitions at all the positions in the historical data table.
Preferably, the dividing the scanning area of the cross section of the water channel according to the historical data comprises:
dividing the cross section of the water channel into a plurality of preset areas:
dividing the position subareas into a preset number of subareas with equal size;
clustering the historical data in the sub-regions;
dividing the historical data into corresponding sub-areas according to specific position coordinates in the partitions;
acquiring a risk model, wherein the historical data is divided into high-risk, medium-risk and risk-free areas as training data of the risk model;
preprocessing point cloud data to be processed into data with the same characteristics as the training data of the risk model;
determining whether the preprocessed data accord with the characteristics of the high-risk area or not according to the risk model, and if so, determining that the sub-area belongs to the high-risk area;
performing the processing on all sub-areas of all position partitions, and combining adjacent areas with the same risk in the same position partition; wherein, the high risk and medium risk areas are used as the areas to be scanned, and the risk-free areas are non-scanning areas.
Preferably, the acquiring point cloud data of the limited area of the cross section of the water channel by using the three-dimensional laser scanner comprises the following steps:
setting a scanning range of a three-dimensional laser scanner according to scanning division, and scanning the cross section of the water channel by using the three-dimensional laser scanner; wherein the three-dimensional laser scanner comprises:
the device comprises a support assembly, a rotary driving assembly arranged above the support assembly, a laser sensor rotating assembly, a GPS positioning assembly arranged above the support assembly, a laser sensor arranged on the laser sensor rotating assembly and a control box assembly arranged below the support assembly;
the rotation driving component is used for driving the laser sensor rotating component to rotate;
the control box assembly is connected with the rotary driving assembly and the laser sensor through cables;
acquiring required point cloud data: scanning the high risk area and the medium risk area, not scanning the risk-free area, and setting different scanning periods for the high risk area and the medium risk area;
and dynamically adjusting the scanning period according to the update of the historical data: setting a scanning period according to the actual situation, setting a shorter scanning period in a high risk area, and setting a longer scanning period in a medium risk area;
the point cloud data obtained by each scanning is added into historical data to update the historical data;
and if the judgment result of the risk area is changed after the historical data is updated, adjusting the scanning period of the area according to the changed judgment result.
Preferably, the scanning obtains abnormal data in the point cloud data, including:
scanning a standard ditch without danger to obtain a standard point cloud data set; the standard point cloud data set comprises data scanned above water and data scanned under water;
the underwater scanning data correction on water comprises the following steps:
preprocessing the acquired point cloud data to obtain preprocessed point cloud data;
performing edge identification on the preprocessed point cloud data to obtain a plurality of edge points;
connecting all the edge points at one time to obtain an overwater and underwater boundary line and a water channel boundary line;
dividing the point cloud data into an above-water part and an under-water part according to an above-water and under-water boundary;
performing the above processing on the standard point cloud data set to obtain a water channel boundary line obtained by water and underwater scanning;
comparing the boundary line of the water channel obtained by the standard point cloud data set with point cloud data to be corrected, and if the deviation exists, acquiring the offset of each corresponding position coordinate;
carrying out summation and average processing on the offset to obtain a final correction quantity, and correcting the position coordinate;
acquiring an abnormal point cloud data set, comprising:
taking a standard point cloud data set as a sample set, and extracting reflection intensity information and coordinate data of the data as characteristic values;
training a data point prediction model through a decision tree algorithm, inputting point cloud data to be processed into the prediction model, and outputting a normal point cloud data set;
and removing the normal point cloud data set on the basis of the original point cloud data to obtain abnormal data in the point cloud data.
Preferably, the identifying the type of danger according to the abnormal data in the scanning comprises:
detecting foreign matters in a water channel and identifying types:
the foreign matter detection mainly comprises the steps of monitoring floaters and mouse damage, and mapping the abnormal point cloud data to generate a distance image and a reflection intensity image;
carrying out point cloud segmentation and clustering on the abnormal point cloud data according to the distance image and the reflection intensity image to obtain a plurality of point cloud areas;
extracting the features of each cloud area, classifying the extracted feature vectors to identify the target, comprising: training a target detection model of the mouse; inputting the extracted characteristic vector into a target detection model, and outputting a recognition result as a mouse damage; extracting characteristic values of point cloud data in a preset range near an overwater and underwater boundary, comparing the characteristic values with a standard point cloud data set, and judging whether abnormal data exist or not; if yes, outputting the identification result as a floating object;
monitoring the form change of the water channel:
after the foreign matter detection, floating objects and rat damage data in the abnormal data are removed, and residual point cloud data are obtained;
extracting edge data from the point cloud data through an edge extraction algorithm;
similarly, the edge data is extracted from the standard point cloud data by the edge extraction algorithm, and the edge extraction algorithm can adopt a Robert operator to extract the edge in the image according to a formula:
comparing the two groups of data to obtain similarity;
setting a threshold value, and considering that the ditch is not deformed when the similarity is higher than the threshold value;
monitoring abnormal water flow of the ditch:
and according to the pre-obtained water and water boundary, the horizontal height of the boundary is obtained by taking the lowest point of the ditch as a reference, and different height thresholds are set to correspond to different water flow situations.
Preferably, the dynamic tracking based on abnormal data and dangerous case type includes:
dynamic monitoring of foreign bodies:
erecting a three-dimensional laser scanner at a preset position of a water channel, acquiring abnormal data and identifying the type of the abnormal data at a certain position, and uploading the abnormal data to a server;
the server dynamically tracks floaters and the type of the rat hazard in the abnormal data, and the method comprises the following steps: extracting characteristic values of the floaters or the dangerous cases of the rats; the server determines a position point needing to be called for dynamic tracking according to the water flow direction and the position information of the abnormal data, sends a scanning instruction to the three-dimensional laser scanner at the position point, and transmits point cloud data obtained by scanning the position point back to the server; the server identifies the dangerous case of the uploaded data according to the characteristic value, and whether the uploaded data is the dangerous case to be tracked or not is judged; if so, sending the instruction to the next site, and if not, recovering the normal working state of the site; acquiring original canal scanning point cloud data, preprocessing the point cloud data, and then reconstructing a three-dimensional curved surface to construct a three-dimensional model of the canal; acquiring the abnormal data and the type of the abnormal data, marking the dangerous case on the three-dimensional model according to the position information of the abnormal data and connecting the same dangerous case monitored by different sites to obtain a movement track of the dangerous case;
dynamic monitoring of abnormal water flow:
acquiring the water flow conditions of all the sites, and comparing water flow difference values of adjacent sites one by one;
setting a threshold value of the difference value, and when the threshold value is exceeded, indicating that the two compared points are blocked;
the two points adjust the scanning period to a preset value, and the maximum scanning times are set;
if the water flow difference between the two sites is reduced below the threshold value before the maximum scanning times are not reached, no processing is carried out, otherwise, the position information and the water flow difference between the two sites are uploaded to a server.
Preferably, the analyzing abnormal point cloud data to generate dangerous case alarm information includes:
acquiring abnormal point cloud data which is dynamically tracked and monitored, and counting the number of floaters and the positions of the floaters within a certain time;
establishing a prediction model of the number of the floating objects corresponding to the blocking probability through a linear regression algorithm, and predicting the size of the blocking probability;
setting a probability threshold value, and when the prediction result is higher than the threshold value, generating blockage alarm information by a server, wherein the blockage alarm information comprises the number of floaters, monitored sites and the probability of predicting blockage;
acquiring monitoring data of the form change of the water channel;
setting a minimum value of edge similarity according to the construction requirement of the canal, and generating canal deformation alarm information by the server when the monitoring data is lower than the set minimum value, wherein the canal deformation alarm information comprises edge data and similarity of the canal;
counting the frequency of the occurrence of the rat hazard dangerous case in a certain time in the abnormal point cloud data;
establishing a prediction model of the mouse damage flooding risk value corresponding to the number of the mouse damage dangerous situations through a linear regression algorithm, and predicting the height of the risk of the mouse damage flooding;
and setting a threshold value of the risk, and when the prediction result is higher than the threshold value, generating the mouse damage alarm information by the server, wherein the mouse damage alarm information comprises the number of times of mouse damage in a certain time and the predicted risk value of the flood of the mouse damage.
The technical scheme provided by the embodiment of the invention has the following beneficial effects:
firstly, detection personnel is not needed to monitor on the spot, the monitoring safety is improved, and moreover, dynamic monitoring ensures that when a sewer or a water channel is in a dangerous situation, the dangerous situation of which type and the specific position of the dangerous situation can be found and identified in time, so that maintenance personnel can remove the dangerous situation as soon as possible, and finally, data are selectively scanned and stored, a large amount of memory of a system is saved, redundant data are reduced, and the high efficiency of the operation of the system is ensured.
[ description of the drawings ]
Fig. 1 is a flowchart of a depth tracking identification method based on laser scanning according to the present invention.
Fig. 2 is a system configuration diagram of the three-dimensional laser scanner of the present invention.
[ detailed description ] embodiments
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in detail with reference to the accompanying drawings and specific embodiments.
FIG. 1 is a flow chart of a compact integrated laser depth tracking identification method according to the present invention. As shown in fig. 1, the depth tracking and identifying system based on laser scanning in this embodiment may specifically include:
step 101, defining a key scanning area based on historical data of the canal dangerous case.
Historical data of the canal danger is obtained and processed.
Acquiring historical data of a channel section dangerous case, wherein the historical data comprises: the dangerous case type and point cloud data thereof, the frequency of occurrence of various dangerous cases, the position partition of the dangerous case occurrence, and the specific position coordinates in the partition, wherein the position partition specifically comprises: and (3) extracting characteristics of different dangerous cases in the historical data, establishing a standard dangerous case matrix, wherein a row vector is each position partition, a column vector is a dangerous case characteristic, and counting the type and the frequency of the dangerous cases in each position partition in the historical data table.
And dividing the scanning area of the cross section of the water channel according to historical data.
Divide the ditch section into a plurality of predetermined regions, specifically include: dividing the position subareas into a preset number of subareas with equal size, clustering historical data in the subareas, dividing the historical data into corresponding subareas according to specific position coordinates in the subareas to obtain a risk model, dividing the historical data serving as training data of the risk model into high-risk, medium-risk and no-risk areas, preprocessing point cloud data to be processed into data with the same characteristics as the training data of the risk model, determining whether the preprocessed data accord with the characteristics of the high-risk area or not according to the risk model, if so, determining that the subarea belongs to the high-risk area, performing the processing on all subareas of all the position subareas, merging adjacent same-risk areas in the same position subarea, and taking the high-risk and medium-risk areas as areas to be scanned, the risk-free area is a non-scanning area, for example, the bottom of the cross section of the canal is divided into 10 preset sub-areas, the risk types of the sub-areas are obtained by processing the historical data of the 10 sub-areas through a risk model, and the adjacent areas with the same risk are combined to obtain an area to be scanned and the non-scanning area.
And 102, acquiring point cloud data of a limited area of the cross section of the water channel by using a three-dimensional laser scanner.
The scanning range of setting for three-dimensional laser scanner according to scanning division utilizes three-dimensional laser scanner to scan the ditch section, three-dimensional laser scanner specifically includes: the laser positioning system comprises a support assembly, a rotary driving assembly arranged above the support assembly, a laser sensor rotary assembly, a GPS positioning assembly arranged above the support assembly, a laser sensor arranged on the laser sensor rotary assembly and a control box assembly arranged below the support assembly, wherein the rotary driving assembly is used for driving the laser sensor rotary assembly to rotate, and the control box assembly is connected with the rotary driving assembly and the laser sensor through cables; acquiring required point cloud data, specifically scanning a high risk area and a medium risk area, not scanning a risk-free area, setting different scanning periods for the high risk area and the medium risk area, and dynamically adjusting the scanning periods according to the update of historical data, wherein the method specifically comprises the following steps: setting a scanning period according to an actual situation, setting a shorter scanning period in a high risk area, setting a longer scanning period in a medium risk area, taking the point cloud data acquired by each scanning into the historical data to update the historical data, and if the judgment result of the risk area is changed after the historical data is updated, adjusting the scanning period of the area according to the changed judgment result, for example: the scanning period of a certain high-risk area is set to be 2 days, no dangerous case occurs after one month of scanning, after the scanning result of the month is taken into the historical data, the area is judged to be a medium-risk area, and then the scanning period of the area is changed to be 5 days.
And 103, scanning to obtain abnormal data in the point cloud data.
Scanning a standard and non-dangerous ditch to obtain a standard point cloud data set, wherein the point cloud data set comprises data scanned above water and data scanned under water.
And (5) correcting the underwater and above water scanning data.
Preprocessing the acquired point cloud data to obtain preprocessed point cloud data, performing edge recognition on the preprocessed point cloud data to obtain a plurality of edge points, connecting the edge points at one time to obtain an overwater and underwater boundary line and a canal boundary line, dividing the point cloud data into an overwater part and an underwater part according to the overwater and underwater boundary line, performing the above processing on a standard point cloud data set to obtain an overwater and underwater scanning acquired canal boundary line, comparing the canal boundary line obtained by the standard point cloud data set with the point cloud data to be corrected, acquiring the offset of each corresponding position coordinate if the offset exists, and performing summation and average processing on the offset to obtain a final correction amount to correct the position coordinate; for example, if most of the data points are found to have an offset to the right by comparison, the offsets are added and averaged to obtain a correction amount, and the data points are adjusted to the left according to the correction amount.
And acquiring an abnormal point cloud data set.
Taking a standard point cloud data set as a sample set, extracting reflection intensity information and coordinate data of the data as characteristic values, training a data point prediction model through a decision tree algorithm, inputting point cloud data to be processed into the prediction model, outputting a normal point cloud data set, and removing the normal point cloud data set on the basis of original point cloud data to obtain abnormal data in the point cloud data, such as: and (3) taking the original point cloud data set as A, obtaining a normal point cloud data set B after processing by a data point prediction model, and removing the same data in the data set A and the data set B to obtain a data set C, wherein the data set C is a required abnormal point cloud data set.
And 104, identifying the dangerous case type according to the abnormal data in the scanning.
Foreign object detection and type identification in a canal.
The foreign matter detection mainly comprises floaters and rat damage monitoring, the abnormal point cloud data are mapped to generate a distance image and a reflection intensity image, point cloud segmentation and clustering are carried out on the field point cloud data according to the distance image and the reflection intensity image to obtain a plurality of point cloud areas, feature extraction is carried out on each point cloud area, and extracted feature vectors are classified to identify targets, and the method specifically comprises the following steps: training a target detection model of the mouse, inputting the extracted characteristic vector into the target detection model, and outputting a recognition result as a mouse damage; detecting whether abnormal data exist in a preset range near the overwater and underwater boundary, and if the abnormal data exist, outputting a recognition result as a floater, for example: and detecting whether abnormal data exist in a strip-shaped area which is 5 cm above the boundary and 5 cm below the boundary, wherein if the abnormal data exist, the abnormal data are from the water surface floating objects.
The morphology of the raceway itself is monitored.
After the foreign matter detection, floating objects and rat damage data in the abnormal data are removed, residual point cloud data are obtained, edge data are extracted from the point cloud data through an edge extraction algorithm, the edge data are extracted from standard point cloud data through the edge extraction algorithm, the edge extraction algorithm can adopt a Robert operator, and edges in the image are extracted according to a formula, wherein the formula is as follows:
and comparing the two groups of data to obtain similarity, setting a threshold value, and considering that the ditch is not deformed when the similarity is higher than the threshold value.
Abnormal water flow of the canal is monitored.
According to the pre-obtained water and water boundary, the horizontal height of the boundary is obtained by taking the lowest point of the ditch as a reference, and different height thresholds are set to correspond to different water flow situations, for example: two height thresholds are set, namely 50 cm and 80 cm, the water level is low when the horizontal height of the above-water and underwater boundary is lower than 50 cm, the water level is normal when the horizontal height is higher than 50 cm and lower than 80 cm, and the water level is abnormal when the horizontal height is higher than 80 cm.
And 105, dynamically tracking based on the abnormal data and the dangerous case type.
And (5) dynamically monitoring the foreign matters.
The method comprises the following steps that a three-dimensional laser scanner is erected at a preset position of a water channel, after a certain position obtains abnormal data and identifies the type of the abnormal data, the abnormal data are uploaded to a server, and the server dynamically tracks floaters and the type of the dangerous case of the rats in the abnormal data, and specifically comprises the following steps: extracting a characteristic value of the flotage or mouse hazard, determining a site needing to be called for dynamic tracking by the server according to a water flow direction and position information of abnormal data, sending a scanning instruction to the three-dimensional laser scanner at the site, transmitting point cloud data obtained by scanning the site back to the server, identifying the uploaded data by the server according to the characteristic value, judging whether the uploaded data is a hazard needing to be tracked, if so, sending the instruction to the next site, and if not, recovering the site to be in a normal working state; acquiring original canal scanning point cloud data, performing three-dimensional curved surface reconstruction after preprocessing to construct a three-dimensional model of the canal, acquiring the abnormal data and the type thereof, marking a dangerous case on the three-dimensional model according to the position information of the abnormal data, and connecting the same dangerous case monitored by different sites to obtain a dangerous case movement track, for example: when the floating object is monitored at the site A, B, C sites are sequentially arranged along the water flow direction, the server sends a scanning instruction to the site B, and when the floating object is displayed in the scanning result returned by the site B, the server sends the scanning instruction to the site C.
And dynamically monitoring abnormal water flow.
And acquiring the water flow conditions of all the sites, comparing the water flow difference values of adjacent sites one by one, setting a threshold value of the difference value, when the water flow conditions exceed the threshold value, indicating that the two sites for comparison are blocked, adjusting the scanning period to a preset value by the two sites, setting the maximum scanning times, if the water flow difference value between the two sites is reduced below the threshold value before the maximum scanning times is not reached, not processing, and otherwise, uploading the position information and the water flow difference value of the two sites to a server.
And 106, analyzing the abnormal point cloud data to generate dangerous case alarm information.
Acquiring abnormal point cloud data of dynamic tracking monitoring, counting the number of floaters and sites where the floaters are monitored within a certain time, establishing a prediction model of the floaters number corresponding to the blocking probability through a linear regression algorithm, predicting the blocking probability, setting a threshold of the probability, and generating blocking alarm information by a server when the prediction result is higher than the threshold, wherein the blocking alarm information specifically comprises: the number of floaters, the monitored sites and the probability of blockage occurrence are predicted; acquiring monitoring data of the form change of the canal, setting a minimum value of edge similarity according to the construction requirement of the canal, and generating canal deformation alarm information by a server when the monitoring data is lower than the set minimum value, wherein the canal deformation alarm information specifically comprises: edge data and similarity of the water channel; counting the frequency of the occurrence of the rat damage dangerous case in the abnormal point cloud data within a certain time, establishing a prediction model of a rat damage flooding risk value corresponding to the frequency of the rat damage dangerous case through a linear regression algorithm, predicting the level of the risk of the rat damage flooding, setting a threshold value of the level of the risk, and generating rat damage alarm information by a server when the prediction result is higher than the threshold value, wherein the rat damage alarm information specifically comprises: the frequency of the damage caused by the rats in a certain time and the predicted risk value of the flooding caused by the damage caused by the rats.
The device for realizing the method can comprise a laser radar module, a power management system, an integrated device shell, a Linux embedded board card for data processing, embedded software and a display screen. The laser radar module adopts a multilayer Printed Circuit Board (PCB) process, is small in size and low in power consumption, the display screen is connected with the embedded board card through an high-definition multimedia interface (HDMI), has a touch function, can display a real-time scanning effect through the display screen, simultaneously stores scanned data into a hard disk in a point cloud format universal in the industry, is connected with the embedded board card through a Universal Serial Bus (USB) to realize communication, and can be introduced into a computer to perform post-processing. In the mechanical structure, the scanning device is fixed in the environments which are difficult to enter such as a water channel, a sewer pipeline and the like through a telescopic bracket and a thin connecting rod for scanning. When the equipment goes deep far away, the touch screen can be separated from the equipment, and a user can see the scanning effect at a short distance. The target tracking and risk prediction are carried out by carrying out a target model on the region of interest where the target is located in the scanning scene and applying the target model to the step method.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Programs for implementing the information governance of the present invention may be written in computer program code for carrying out operations of the present invention in one or more programming languages, including an object oriented programming language such as Java, python, C + +, or a combination thereof, as well as conventional procedural programming languages, such as the C language or similar programming languages.
The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
In the embodiments provided in the present invention, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. For example, the above-described device embodiments are merely illustrative, and for example, the division of the units is only one logical functional division, and other divisions may be realized in practice.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.
The integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute some steps of the methods according to the embodiments of the present invention.
And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
Claims (9)
1. A depth tracking identification method based on laser scanning is characterized by comprising the following steps:
defining a key scanning area based on historical data of the canal danger: acquiring and processing historical data of a water channel dangerous case, and dividing a scanning area of a water channel section according to the historical data;
acquiring point cloud data of a water channel section limited area by using a three-dimensional laser scanner;
scanning to obtain abnormal data in the point cloud data: correcting the overwater and underwater scanning data and acquiring an abnormal point cloud data set;
identifying the type of the dangerous case according to abnormal data in scanning: detecting foreign matters in the water channel, identifying types of the foreign matters, monitoring the form change of the water channel, and monitoring the abnormal water flow of the water channel;
and dynamically tracking based on abnormal data and dangerous case types: dynamic monitoring of foreign matters and abnormal water flow;
and analyzing the abnormal point cloud data to generate dangerous case alarm information.
2. The method of claim 1, wherein the defining a region of key scanning based on historical data of the raceway danger comprises:
acquiring and processing historical data of a canal dangerous case;
and dividing the scanning area of the cross section of the water channel according to historical data.
3. The method of claim 2, wherein the obtaining and processing historical data of raceway hazards comprises:
acquiring historical data of a dangerous case of a cross section of a water channel; the historical data comprises dangerous case types and point cloud data thereof, the frequency of dangerous case occurrence, position partitions of dangerous case occurrence and specific position coordinates in the partitions; the position subareas comprise the top, the left side surface, the right side surface, the bottom and the water surface of the cross section of the water channel;
extracting the characteristics of different dangerous cases in the historical data, establishing a standard dangerous case matrix, wherein row vectors are partitions at all positions, column vectors are dangerous case characteristics, and counting the types and times of the dangerous cases of the partitions at all the positions in the historical data table.
4. The method of claim 2, wherein the partitioning the scan area of the channel section from historical data comprises:
dividing the cross section of the water channel into a plurality of preset areas:
dividing the position subareas into a preset number of subareas with equal size;
clustering the historical data in the sub-regions;
dividing the historical data into corresponding sub-areas according to specific position coordinates in the partitions;
acquiring a risk model, wherein the historical data is divided into high-risk, medium-risk and risk-free areas as training data of the risk model;
preprocessing point cloud data to be processed into data with the same characteristics as the training data of the risk model;
determining whether the preprocessed data accord with the characteristics of the high-risk area or not according to the risk model, and if so, determining that the sub-area belongs to the high-risk area;
performing the processing on all sub-areas of all position partitions, and combining adjacent same-risk areas in the same position partition; wherein, the high risk and medium risk areas are used as the areas to be scanned, and the risk-free areas are non-scanning areas.
5. The method of claim 1, wherein the acquiring point cloud data of a defined area of a cross-section of a canal using a three-dimensional laser scanner comprises:
setting a scanning range of a three-dimensional laser scanner according to scanning division, and scanning the cross section of the water channel by using the three-dimensional laser scanner; wherein the three-dimensional laser scanner comprises:
the device comprises a support assembly, a rotary driving assembly arranged above the support assembly, a laser sensor rotating assembly, a GPS positioning assembly arranged above the support assembly, a laser sensor arranged on the laser sensor rotating assembly and a control box assembly arranged below the support assembly;
the rotation driving component is used for driving the laser sensor rotating component to rotate;
the control box assembly is connected with the rotary driving assembly and the laser sensor through cables;
acquiring required point cloud data: scanning the high risk area and the medium risk area, not scanning the risk-free area, and setting different scanning periods for the high risk area and the medium risk area;
and dynamically adjusting the scanning period according to the update of the historical data: setting a scanning period according to the actual situation, setting a shorter scanning period in a high risk area, and setting a longer scanning period in a medium risk area;
the point cloud data obtained by each scanning is added into historical data to update the historical data;
and if the judgment result of the risk area is changed after the historical data is updated, adjusting the scanning period of the area according to the changed judgment result.
6. The method of claim 1, wherein the scanning results in anomalous data in the point cloud data, comprising:
scanning a standard ditch without danger to obtain a standard point cloud data set; the standard point cloud data set comprises data scanned above water and data scanned under water;
the underwater scanning data correction on water comprises the following steps:
preprocessing the acquired point cloud data to obtain preprocessed point cloud data;
performing edge identification on the preprocessed point cloud data to obtain a plurality of edge points;
connecting all the edge points at one time to obtain an overwater and underwater boundary line and a water channel boundary line;
dividing the point cloud data into an above-water part and an under-water part according to an above-water and under-water boundary;
performing the above processing on the standard point cloud data set to obtain a water channel boundary line obtained by water and underwater scanning;
comparing the boundary line of the water channel obtained by the standard point cloud data set with point cloud data to be corrected, and if the deviation exists, acquiring the offset of each corresponding position coordinate;
carrying out summation and average processing on the offset to obtain a final correction quantity, and correcting the position coordinate;
acquiring an abnormal point cloud data set, comprising:
taking a standard point cloud data set as a sample set, and extracting reflection intensity information and coordinate data of the data as characteristic values;
training a data point prediction model through a decision tree algorithm, inputting point cloud data to be processed into the prediction model, and outputting a normal point cloud data set;
and removing the normal point cloud data set on the basis of the original point cloud data to obtain abnormal data in the point cloud data.
7. The method of claim 1, wherein identifying the type of risk based on the anomalous data in the scan comprises:
foreign matter detection and type identification in the ditch:
the foreign matter detection mainly comprises the steps of monitoring floaters and mouse damage, and mapping the abnormal point cloud data to generate a distance image and a reflection intensity image;
carrying out point cloud segmentation and clustering on the abnormal point cloud data according to the distance image and the reflection intensity image to obtain a plurality of point cloud areas;
extracting the features of each cloud area, classifying the extracted feature vectors to identify the target, comprising: training a target detection model of the mouse; inputting the extracted feature vector into a target detection model, and outputting a recognition result as a mouse damage; extracting characteristic values of point cloud data in a preset range near the overwater and underwater boundary, comparing the characteristic values with a standard point cloud data set, and judging whether abnormal data exist or not; if yes, outputting the identification result as a floating object;
monitoring the form change of the water channel:
after the foreign matter detection, floating objects and rat damage data in the abnormal data are removed, and residual point cloud data are obtained;
extracting edge data from the point cloud data through an edge extraction algorithm;
similarly, the edge data is extracted from the standard point cloud data by the edge extraction algorithm, the edge extraction algorithm can adopt a Robert operator to extract the edge in the image according to a formula, and the formula is as follows:
comparing the two groups of data to obtain similarity;
setting a threshold value, and considering that the ditch is not deformed when the similarity is higher than the threshold value;
monitoring abnormal water flow of the ditch:
and according to the pre-obtained water and water boundary, the horizontal height of the boundary is obtained by taking the lowest point of the water channel as a reference, and different height thresholds are set to correspond to different water flow conditions.
8. The method of claim 1, wherein the dynamically tracking based on anomaly data and a type of risk comprises:
dynamic monitoring of foreign bodies:
erecting a three-dimensional laser scanner at a preset position of a water channel, acquiring abnormal data and identifying the type of the abnormal data at a certain position, and uploading the abnormal data to a server;
the server carries out dynamic tracking to floaters and dangerous case types of rats in the abnormal data, and the method comprises the following steps: extracting characteristic values of the floaters or the dangerous cases of the rats; the server determines a position point needing to be called for dynamic tracking according to the water flow direction and the position information of the abnormal data, sends a scanning instruction to the three-dimensional laser scanner at the position point, and transmits point cloud data obtained by scanning the position point back to the server; the server identifies the dangerous case of the uploaded data according to the characteristic value, and whether the uploaded data is a dangerous case to be tracked or not is judged; if so, sending the instruction to the next site, and if not, recovering the normal working state of the site; acquiring original canal scanning point cloud data, and performing three-dimensional curved surface reconstruction after preprocessing to construct a three-dimensional model of the canal; acquiring the abnormal data and the type of the abnormal data, marking the dangerous case on the three-dimensional model according to the position information of the abnormal data and connecting the same dangerous case monitored by different sites to obtain a movement track of the dangerous case;
dynamic monitoring of abnormal water flow:
acquiring the water flow conditions of all the sites, and comparing water flow difference values of adjacent sites one by one;
setting a threshold value of the difference value, and when the threshold value is exceeded, indicating that the two compared points are blocked;
the two sites adjust the scanning period to a preset value, and the maximum scanning times are set;
if the water flow difference between the two sites is reduced below the threshold value before the maximum scanning times are not reached, no processing is carried out, otherwise, the position information and the water flow difference between the two sites are uploaded to a server.
9. The method of claim 1, wherein the analyzing the outlier point cloud data to generate hazardous alarm information comprises:
acquiring abnormal point cloud data which is dynamically tracked and monitored, and counting the number of floaters and the positions of the floaters within a certain time;
establishing a prediction model of the number of the floating objects corresponding to the blocking probability through a linear regression algorithm, and predicting the size of the blocking probability;
setting a probability threshold value, and when the prediction result is higher than the threshold value, generating blockage alarm information by a server, wherein the blockage alarm information comprises the number of floating objects, monitored sites and the probability of blockage prediction;
acquiring monitoring data of the form change of the water channel;
setting a minimum value of edge similarity according to the construction requirement of the canal, and generating canal deformation alarm information by the server when the monitoring data is lower than the set minimum value, wherein the canal deformation alarm information comprises edge data and similarity of the canal;
counting the frequency of the occurrence of the rat hazard dangerous case in a certain time in the abnormal point cloud data;
establishing a prediction model of the mouse damage flooding risk value corresponding to the mouse damage dangerous case frequency through a linear regression algorithm, and predicting the level of the risk of the mouse damage flooding;
and setting a threshold value of the risk, and when the prediction result is higher than the threshold value, generating the mouse damage alarm information by the server, wherein the mouse damage alarm information comprises the number of times of mouse damage in a certain time and the predicted risk value of the flood of the mouse damage.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210136111.XA CN114693729A (en) | 2022-02-15 | 2022-02-15 | Depth tracking identification method based on laser scanning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210136111.XA CN114693729A (en) | 2022-02-15 | 2022-02-15 | Depth tracking identification method based on laser scanning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114693729A true CN114693729A (en) | 2022-07-01 |
Family
ID=82137467
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210136111.XA Pending CN114693729A (en) | 2022-02-15 | 2022-02-15 | Depth tracking identification method based on laser scanning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114693729A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115311354A (en) * | 2022-09-20 | 2022-11-08 | 中国铁建电气化局集团有限公司 | Foreign matter risk area identification method, device, equipment and storage medium |
-
2022
- 2022-02-15 CN CN202210136111.XA patent/CN114693729A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115311354A (en) * | 2022-09-20 | 2022-11-08 | 中国铁建电气化局集团有限公司 | Foreign matter risk area identification method, device, equipment and storage medium |
| CN115311354B (en) * | 2022-09-20 | 2024-01-23 | 中国铁建电气化局集团有限公司 | Foreign matter risk area identification method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108230344B (en) | Automatic identification method for tunnel water leakage diseases | |
| CN110208803B (en) | Method for monitoring and identifying ship anti-collision radar at bridge | |
| US20230005176A1 (en) | Throwing position acquisition method and apparatus, computer device and storage medium | |
| CN110705759B (en) | Water level early warning monitoring method and device, storage medium and electronic equipment | |
| CN108765453B (en) | Expressway agglomerate fog identification method based on video stream data | |
| CN111310736B (en) | Rapid identification method for unloading and stacking of vehicles in protection area | |
| WO2020111934A1 (en) | A method and system for detection of natural disaster occurrence | |
| CN109186706A (en) | A method of for the early warning of Urban Storm Flood flooding area | |
| CN114842450A (en) | Driving region detection method, device and equipment | |
| CN114693729A (en) | Depth tracking identification method based on laser scanning | |
| CN104063691B (en) | Lane line quick determination method based on improved Hough transform | |
| CN105141887A (en) | Submarine cable area video alarming method based on thermal imaging | |
| US20210104059A1 (en) | Method for size estimation by image recognition of specific target using given scale | |
| CN112633722A (en) | Vehicle-mounted road safety risk assessment system and method | |
| CN113963314A (en) | Rainfall monitoring method and device, computer equipment and storage medium | |
| CN102054174A (en) | Method and system for positioning licence plate | |
| CN116739183B (en) | Mine safety risk early warning prediction system | |
| CN113496182A (en) | Road extraction method and device based on remote sensing image, storage medium and equipment | |
| CN115471980B (en) | Mud-rock flow disaster early warning method | |
| CN115439792A (en) | Monitoring method and system based on artificial intelligence | |
| CN116524686A (en) | Mine slope disaster monitoring system and method | |
| CN115937539A (en) | Cable well monitoring and early warning method and system based on multi-signal decision | |
| Li-Yong et al. | A lane detection technique based on adaptive threshold segmentation of lane gradient image | |
| Paar et al. | Vision-based terrestrial surface monitoring | |
| Guindon | A framework for the development and assessment of object recognition modules from high-resolution satellite images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20221103 Address after: 511500 No. 06-07, floor t0218, industrial building, Tian'an Zhigu science and Technology Industrial Park, No. 18, Chuangxing Avenue, Qingyuan high tech Industrial Development Zone, Qingyuan City, Guangdong Province Applicant after: Guangdong Normal University Weizhi Information Technology Co.,Ltd. Applicant after: South China Normal University (Qingyuan) science and Technology Innovation Research Institute Co.,Ltd. Address before: 510000 No. 503, 504 and 510, floor 5, No. 886, Tianhe North Road, Tianhe District, Guangzhou City, Guangdong Province Applicant before: Guangzhou Weizhi digital twin Intelligent Information Technology Co.,Ltd. |
|
| TA01 | Transfer of patent application right |