CN115661666B - Bridge identification method and device in remote sensing image, electronic equipment and medium - Google Patents
Bridge identification method and device in remote sensing image, electronic equipment and medium Download PDFInfo
- Publication number
- CN115661666B CN115661666B CN202211587457.8A CN202211587457A CN115661666B CN 115661666 B CN115661666 B CN 115661666B CN 202211587457 A CN202211587457 A CN 202211587457A CN 115661666 B CN115661666 B CN 115661666B
- Authority
- CN
- China
- Prior art keywords
- bridge
- area
- identification
- overpass
- identification result
- 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.)
- Active
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The application provides a method and a device for identifying a bridge in a remote sensing image, electronic equipment and a medium. Respectively acquiring first identification results of an underwater bridge, a pedestrian bridge and an overpass in a current remote sensing image based on a bridge identification network; determining a second identification result based on the width of the region of any water bridge and/or the coincidence rate of the region of any water bridge and the region of any pedestrian bridge; identifying the road of the current remote sensing image based on a road identification network to obtain road identification results of different types of roads; updating the identification result of the overpass in the first identification result based on the incidence relation between the road identification result and the position of the area of the overpass in the first identification result to obtain a third identification result; and determining a bridge identification result corresponding to the current remote sensing image based on the second identification result and the third identification result. The method can accurately and quickly realize the identification of the full bridge in the remote sensing image.
Description
Technical Field
The application relates to the technical field of image processing, in particular to a method and a device for identifying a bridge in a remote sensing image, electronic equipment and a medium.
Background
The bridge belongs to important transportation facilities, the identification of the bridge in the remote sensing image is not only beneficial to the updating and maintenance of geographic information, but also beneficial to urban building planning, natural disaster assessment, path selection and the like, and the application value in the fields of military affairs and civil affairs is very high.
At present, bridge identification based on deep learning mainly identifies a single type of bridge through a segmentation or detection mode. The method for identifying the single type of bridge can only output the result of one type of bridge and cannot meet the requirement for comprehensively knowing the condition of the bridge.
Disclosure of Invention
An object of the embodiments of the present application is to provide a method, an apparatus, an electronic device, and a medium for identifying a bridge in a remote sensing image, so as to solve the above problems in the prior art, and accurately identify different types of bridges in the remote sensing image.
In a first aspect, a method for identifying a bridge in a remote sensing image is provided, and the method may include:
respectively acquiring first identification results of an underwater bridge, a pedestrian bridge and an overpass in the current remote sensing image based on a bridge identification network; the first identification result comprises the position of the area of the corresponding type of bridge and the width of the corresponding area;
determining a second identification result based on the width of the area of any water bridge in the first identification result and/or the coincidence rate of the area of any water bridge and the area of any pedestrian bridge;
identifying the road of the current remote sensing image based on a road identification network to obtain road identification results of different types of roads, wherein the road identification results comprise the positions of areas of the corresponding types of roads;
updating the identification result of the overpass in the first identification result based on the incidence relation between the road identification result and the position of the area of the overpass in the first identification result to obtain a third identification result;
and determining a bridge identification result corresponding to the current remote sensing image based on the second identification result and the third identification result.
In a second aspect, there is provided an apparatus for identifying a bridge in a remote sensing image, the apparatus may include:
the acquisition unit is used for respectively acquiring first identification results of an overwater bridge, a pedestrian bridge and an overpass in the current remote sensing image based on a bridge identification network; the first identification result comprises the position of the area of the corresponding type of bridge and the width of the corresponding area;
the determining unit is used for determining a second recognition result based on the width of the region of any one overwater bridge and/or the coincidence rate of the region of any one overwater bridge and the region of any one pedestrian bridge in the first recognition result;
the identification unit is used for identifying the road of the current remote sensing image based on a road identification network to obtain road identification results of different types of roads, and the road identification results comprise the positions of areas of the corresponding types of roads;
the updating unit is used for updating the identification result of the overpass in the first identification result based on the incidence relation between the road identification result and the position of the area of the overpass in the first identification result to obtain a third identification result;
the determining unit is further configured to determine a bridge identification result corresponding to the current remote sensing image based on the second identification result and the third identification result.
In a third aspect, an electronic device is provided, which includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete communication with each other through the communication bus;
a memory for storing a computer program;
a processor adapted to perform the method steps of any of the above first aspects when executing a program stored in the memory.
In a fourth aspect, a computer-readable storage medium is provided, having stored therein a computer program which, when executed by a processor, performs the method steps of any of the above first aspects.
The method for identifying the bridge in the remote sensing image, provided by the embodiment of the application, is used for respectively obtaining first identification results of the bridge on water, the pedestrian bridge and the overpass in the current remote sensing image based on a bridge identification network; the method comprises the positions of areas of bridges of corresponding types and the widths of the corresponding areas; determining a second identification result based on the width of any one region of the water bridge and/or the coincidence rate of any one region of the water bridge and any one region of the pedestrian bridge in the first identification result; identifying the road of the current remote sensing image based on a road identification network to obtain road identification results of different types of roads, wherein the road identification results comprise the positions of the areas of the corresponding types of roads; updating the identification result of the overpass in the first identification result based on the incidence relation between the road identification result and the position of the area of the overpass in the first identification result to obtain a third identification result; and determining a bridge identification result corresponding to the current remote sensing image based on the second identification result and the third identification result. According to the method, the identified results of the overwater bridge, the pedestrian bridge, the road and the overpass are corrected, the bridge which is possibly identified by errors is deleted, the problem of false detection among different types of bridges is reduced, and the identification of the whole bridge in the remote sensing image can be accurately and quickly realized.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and that those skilled in the art can also obtain other related drawings based on the drawings without inventive efforts.
Fig. 1 is a schematic flow chart of a method for identifying a bridge in a remote sensing image according to an embodiment of the present disclosure;
fig. 2 is a schematic view of a determination of an overwater bridge according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of an apparatus for identifying a bridge in a remote sensing image according to an embodiment of the present disclosure;
fig. 4 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present application without any creative effort belong to the protection scope of the present application.
The method for identifying the bridge in the remote sensing image can be applied to a server and can also be applied to a terminal. The server may be a physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as cloud service, a cloud database, cloud computing, a cloud function, cloud storage, network service, cloud communication, middleware service, domain name service, security service, content Delivery Network (CDN), big data and an artificial intelligence platform. The Terminal may be a User Equipment (UE) such as a Mobile phone, a smart phone, a notebook computer, a digital broadcast receiver, a Personal Digital Assistant (PDA), a tablet computer (PAD), etc. with high computing power, a handheld device, a vehicle-mounted device, a wearable device, a computing device or other processing device connected to a wireless modem, a Mobile Station (MS), a Mobile Terminal (Mobile Terminal), etc.
According to the method for identifying the bridge in the remote sensing image, the results of the above-water bridge, the pedestrian bridge, the road and the overpass are filtered and fused, the bridge which is possibly identified wrongly is deleted or corrected, the problem of false detection among different types of bridges is reduced, and the results of various bridges are output simultaneously. The method can accurately, completely and quickly realize the identification of the full bridge in the remote sensing image.
The preferred embodiments of the present application will be described in conjunction with the drawings of the specification, it should be understood that the preferred embodiments described herein are only for illustrating and explaining the present application, and are not intended to limit the present application, and the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
Fig. 1 is a schematic flow chart of a method for identifying a bridge in a remote sensing image according to an embodiment of the present disclosure. As shown in fig. 1, the method may include:
step S110, based on the bridge identification network, first identification results of the overwater bridge, the pedestrian bridge and the overpass in the current remote sensing image are respectively obtained.
The first identification result comprises the position of the region of the corresponding type of bridge and the width of the corresponding region.
(1) For an above-water bridge:
identifying the overwater bridge of the current remote sensing image based on a preset rotating frame identification model, such as an Oriented R-CNN network model, and obtaining the positions of a plurality of rotating frames output by the rotating frame identification model; each rotating frame represents one area of any overwater bridge; and determining the minimum circumscribed rectangle of the multiple overlapped rotating frames as the region of the overwater bridge based on the positions of the multiple rotating frames, and determining the position and the width of the minimum circumscribed rectangle as the position of the region of the overwater bridge and the width of the corresponding region.
Specifically, the orientited R-CNN network is used for identifying the overwater bridge in the current remote sensing image, and coordinates of four angular points of a rotating frame corresponding to all overwater bridges in the current remote sensing image output by the Orentited R-CNN network and the corresponding rotating frame, namely the position of the rotating frame, are obtained. Each rotating frame represents one area of any overwater bridge;
then, randomly selecting one rotating frame as the current rotating frame, and traversing the rest rotating frames to execute the following steps: calculate IoU between the two rotating frames. The calculation mode of IoU is dividing the intersection area of two rotating frames by the parallel area of two rotating frames. If IoU >0, it is added to the overlapped frame group of the current rotated frame. And iteratively searching the overlapped frame group of the rest rotating frames until the number of the rest rotating frames is 0. And for each group of overlapping frames, taking the minimum circumscribed rectangle of all the overlapping frames as the region of the corresponding overwater bridge. And determining the minimum circumscribed rectangle corresponding to each overlapped frame group as the area of an overwater bridge.
And finally, determining the position and the width of the minimum circumscribed rectangle as the position of the region of the overwater bridge and the width of the corresponding region.
As shown in fig. 2, a group of overlapping frames including a plurality of mutually overlapping rotating frames exists in the bridge portion of the overwater bridge in the current remote sensing image, and a minimum circumscribed rectangle a of the group of overlapping frames is obtained.
The method comprises the steps that an overwater bridge in a remote sensing image is marked by using a rotating frame, and the label format is (x 1, y1, x2, y2, x3, y3, x4, y 4), namely label data; the remote sensing image is cut into an image set with the size of 1024 multiplied by 1024, the image in the image concentration point keeps the characteristics of the overwater bridge, the retention proportion of the part of the overwater bridge in the remote sensing image to the original size of the overwater bridge is calculated, and the image in the image set with the deletion retention proportion smaller than 0.7 is used as a training sample.
The preset rotating frame recognition model is obtained based on the label data and training samples.
(2) For a footbridge:
and identifying the pedestrian bridge of the current remote sensing image based on a preset semantic segmentation model, such as a Segformer network model, so as to obtain all connected domains in a segmentation result output by the semantic segmentation model. And determining each connected domain as a pedestrian bridge area, and determining the position and the width of the corresponding connected domain as the position of the pedestrian bridge area and the width of the corresponding area.
Specifically, a Segformer network is used for segmenting a pedestrian bridge in a current remote sensing image to obtain a segmentation result output by the Segformer network, all connected domains in the segmentation result are found at first, and each connected domain represents a pedestrian bridge area. The connected domain is the region composed of the number 1 in the segmentation result, so that the position and the width of the connected domain can be obtained, and the position and the width of the corresponding connected domain are determined as the position of the region of the pedestrian bridge and the width of the corresponding region.
The preset semantic segmentation model is obtained by training based on the remote sensing image and a pedestrian bridge area marked in the remote sensing image.
(3) For overpasses:
detecting the overpass of the current remote sensing image based on a preset target identification network, such as a Yolov4 network, and obtaining a rectangular frame output by the target identification network; and determining each rectangular frame as the area of one overpass, and determining the position and the width of the corresponding rectangular frame as the position of the area of the overpass and the width of the corresponding area.
Specifically, the Yolov4 network is used for identifying the overpass in the current remote sensing image to obtain coordinates of an overpass rectangular frame output by the Yolov4 network in the current remote sensing image, so that the position and the width of the corresponding rectangular frame are determined as the position of the area of the overpass and the width of the corresponding area.
The preset target recognition network is obtained by training based on the remote sensing image and the overpass area marked with the rectangular frame in the remote sensing image.
The training method of each model is the same as that of the conventional model, and the embodiment of the present application is not limited thereto.
And S120, determining a second identification result based on the width of the area of any one overwater bridge and/or the overlapping rate of the area of any one overwater bridge and the area of any one pedestrian bridge in the first identification result.
In specific implementation, the width of the region of any overwater bridge is compared with a preset width threshold (such as 10), and/or the coincidence rate of the region of any overwater bridge and the region of any pedestrian bridge is compared with a preset coincidence threshold; wherein the preset width threshold value is the preset minimum width of the bridge deck of the overwater bridge;
(1) The specific way of comparing the area width of any overwater bridge with a preset width threshold (e.g. 10) comprises:
if the area width of any overwater bridge is not larger than the preset width threshold value, the bridge deck of the overwater bridge is indicated to be narrow, the bridge may not be an overwater bridge and may be a pedestrian bridge, namely, the identification is wrong, at the moment, the identification result of the overwater bridge in the first identification result needs to be corrected, namely, the bridge type of the overwater bridge corresponding to the first identification result is updated to be the pedestrian bridge, the identification result corresponding to the preset width threshold value is obtained, and the identification result is changed compared with the first identification result;
if the area width of any water bridge is larger than the preset width threshold value, it is indicated that the bridge deck of the water bridge is normal, namely, the identification is correct, at the moment, the identification result of the water bridge in the first identification result does not need to be corrected, namely, the identification result of the water bridge in the first identification result is kept unchanged, and at the moment, the identification result corresponding to the preset width threshold value is unchanged from the first identification result.
(2) The specific mode of comparing the coincidence rate of the area of any overwater bridge and the area of any pedestrian bridge with a preset coincidence threshold value comprises the following steps:
acquiring a first image corresponding to the area of any overwater bridge and a second image corresponding to the area of any pedestrian bridge, and acquiring a superposed part image of the first image and the second image; and calculating the ratio of the number of pixels of the superposed part image to the number of pixels of the second image, and determining the ratio as the superposition rate of the area of the corresponding overwater bridge and the area of the corresponding pedestrian bridge.
If the coincidence rate of the area of any overwater bridge and the area of any pedestrian bridge is not smaller than a preset coincidence threshold (such as 0.2), the fact that the two areas are large in overlapping area and errors exist in recognition is indicated, at the moment, the recognition result corresponding to the pedestrian bridge in the first recognition result is deleted, the recognition result corresponding to the preset coincidence threshold is obtained, and the recognition result is changed compared with the first recognition result.
If the coincidence rate of the area of any overwater bridge and the area of any pedestrian bridge is smaller than a preset coincidence threshold (such as 0.2), the overlapping area of the two areas is small, the identification result corresponding to the pedestrian bridge in the first identification result does not need to be corrected, namely the identification result corresponding to the pedestrian bridge in the first identification result is kept unchanged, and the identification result corresponding to the preset coincidence threshold obtained at the moment is unchanged from the first identification result.
And obtaining a second recognition result based on the recognition result corresponding to the preset width threshold and/or the recognition result corresponding to the preset coincidence threshold. And when the recognition result corresponding to the preset width threshold and the recognition result corresponding to the preset coincidence threshold are not changed, the second recognition result is the same as the first recognition result.
And S130, identifying the road of the current remote sensing image based on the road identification network to obtain the road identification results of different types of roads.
And the road identification result comprises the position of the area of the corresponding type of road. The road identification network may be a D-LinkNet network, and the different types of roads may include railways, expressways, ordinary highways, and minor roads.
The first type of roads, which can be intersected with the overpass, among the different types of roads, is expressways, ordinary roads, and the second type of roads, which cannot be intersected with the overpass, is railways and paths.
The road recognition network is trained in the same way as the existing model, and the embodiment of the present application is not limited herein.
And step S140, updating the recognition result of the overpass in the first recognition result based on the incidence relation between the road recognition result and the position of the area of the overpass in the first recognition result to obtain a third recognition result.
In specific implementation, a D-LinkNet network is used for segmenting roads in a current remote sensing image to obtain segmentation results output by the D-LinkNet network, a first type of roads in different types of roads, such as expressways and common roads, are determined according to pixel values in the segmentation results, and the values of the expressways and the common roads in the segmentation results are respectively 2 and 3;
detecting whether an overlapping area exists between the position of the area of each road in the first type of road and the position of the area of any overpass, namely determining whether any overpass is overlapped with any expressway or common highway;
if no overlapping area exists, deleting the identification result corresponding to the overpass in the first identification result to obtain a third identification result, namely the identification result corresponding to the overpass in the first identification result is different from the identification result corresponding to the overpass in the third identification result;
if the overlapped area exists, the overlapped area corresponding to the corresponding overpass is reserved in the first recognition result as a recognition result, and a third recognition result is obtained, namely the recognition result corresponding to the overpass in the first recognition result is the same as the recognition result corresponding to the overpass in the third recognition result.
Further, acquiring all connected regions and corresponding connected region areas in the regions of all overpasses in the third recognition result; and reserving the target connected domain with the largest area in each overpass region, and deleting the other connected domains except the target connected domain in the corresponding region.
And S150, determining a bridge identification result corresponding to the current remote sensing image based on the second identification result and the third identification result.
And combining the obtained second recognition result and the third recognition result to obtain a bridge recognition result corresponding to the current remote sensing image, wherein the bridge recognition result comprises recognition results of an overwater bridge, a pedestrian bridge and an overpass.
According to the method for identifying the bridge in the remote sensing image, the results of the overwater bridge, the pedestrian bridge, the road and the overpass are filtered and fused, the bridge which is possibly identified wrongly is deleted or corrected, the false detection problem among different types of bridges is reduced, and the results of various bridges are output at the same time. The method can accurately, completely and quickly realize the identification of the full bridge in the remote sensing image.
Corresponding to the above method, an embodiment of the present application further provides an apparatus for identifying a bridge in a remote sensing image, as shown in fig. 3, the apparatus includes:
an obtaining unit 310, configured to obtain first recognition results of an overwater bridge, a pedestrian bridge, and an overpass in the current remote sensing image, respectively, based on a bridge recognition network; the first identification result comprises the position of the area of the corresponding type of bridge and the width of the corresponding area;
a determining unit 320, configured to determine a second recognition result based on a width of any one region of the marine bridge and/or a coincidence rate of any one region of the marine bridge and any one region of the pedestrian bridge in the first recognition result;
the identification unit 330 is configured to identify a road of a current remote sensing image based on a road identification network to obtain road identification results of different types of roads, where the road identification results include positions of areas of the corresponding types of roads;
an updating unit 340, configured to update the identification result of the overpass in the first identification result based on the association relationship between the road identification result and the location of the area of the overpass in the first identification result, so as to obtain a third identification result;
the determining unit 320 is further configured to determine, based on the second recognition result and the third recognition result, a bridge recognition result corresponding to the current remote sensing image.
The functions of the functional units of the identification device for bridges in remote sensing images provided by the embodiments of the present application can be implemented through the above method steps, and therefore, detailed working processes and beneficial effects of the units in the device provided by the embodiments of the present application are not repeated herein.
An electronic device is further provided in the embodiment of the present application, as shown in fig. 4, and includes a processor 410, a communication interface 420, a memory 430, and a communication bus 440, where the processor 410, the communication interface 420, and the memory 430 complete communication with each other through the communication bus 440.
A memory 430 for storing computer programs;
the processor 410, when executing the program stored in the memory 430, implements the following steps:
respectively acquiring first identification results of an underwater bridge, a pedestrian bridge and an overpass in the current remote sensing image based on a bridge identification network; the first identification result comprises the position of the area of the corresponding type of bridge and the width of the corresponding area;
determining a second identification result based on the width of the area of any water bridge in the first identification result and/or the coincidence rate of the area of any water bridge and the area of any pedestrian bridge;
identifying the road of the current remote sensing image based on a road identification network to obtain road identification results of different types of roads, wherein the road identification results comprise the positions of areas of the corresponding types of roads;
updating the recognition result of the overpass in the first recognition result based on the incidence relation between the road recognition result and the position of the area of the overpass in the first recognition result to obtain a third recognition result;
and determining a bridge identification result corresponding to the current remote sensing image based on the second identification result and the third identification result.
The aforementioned communication bus may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus.
The communication interface is used for communication between the electronic equipment and other equipment.
The Memory may include a Random Access Memory (RAM) or a Non-Volatile Memory (NVM), such as at least one disk Memory. Alternatively, the memory may be at least one memory device located remotely from the processor.
The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components.
As the implementation manner and the beneficial effects of the problem solving of each device of the electronic device in the foregoing embodiment can be implemented by referring to each step in the embodiment shown in fig. 1, detailed working processes and beneficial effects of the electronic device provided in the embodiment of the present application are not repeated herein.
In another embodiment provided by the present application, a computer-readable storage medium is further provided, where instructions are stored in the computer-readable storage medium, and when the instructions are executed on a computer, the computer is caused to execute the method for identifying a bridge in a remote sensing image according to any one of the above embodiments.
In another embodiment provided by the present application, there is further provided a computer program product containing instructions, which when run on a computer, causes the computer to execute the method for identifying a bridge in a remote sensing image according to any one of the above embodiments.
As will be appreciated by one of skill in the art, the embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, embodiments of the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and so forth) having computer-usable program code embodied therein.
Embodiments of the present application are described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including the preferred embodiment and all changes and modifications that fall within the true scope of the embodiments of the present application.
It is apparent that those skilled in the art can make various changes and modifications to the embodiments of the present application without departing from the spirit and scope of the embodiments of the present application. Thus, if such modifications and variations of the embodiments of the present application fall within the scope of the claims of the embodiments of the present application and their equivalents, the embodiments of the present application are also intended to include such modifications and variations.
Claims (8)
1. A method for identifying a bridge in a remote sensing image is characterized by comprising the following steps:
respectively acquiring first identification results of an underwater bridge, a pedestrian bridge and an overpass in the current remote sensing image based on a bridge identification network; the first identification result comprises the position of the area of the corresponding type of bridge and the width of the corresponding area;
determining a second identification result based on the width of the area of any water bridge in the first identification result and/or the coincidence rate of the area of any water bridge and the area of any pedestrian bridge;
identifying the road of the current remote sensing image based on a road identification network to obtain road identification results of different types of roads, wherein the road identification results comprise the positions of areas of the corresponding types of roads;
updating the recognition result of the overpass in the first recognition result based on the incidence relation between the road recognition result and the position of the area of the overpass in the first recognition result to obtain a third recognition result;
determining a bridge identification result corresponding to the current remote sensing image based on the second identification result and the third identification result;
wherein, based on bridge identification network, obtain the first recognition result of bridge, pedestrian bridge and overpass on the water in the current remote sensing image respectively, include:
identifying the overwater bridge of the current remote sensing image based on a preset rotating frame identification model to obtain the positions of a plurality of rotating frames output by the rotating frame identification model; each rotating frame represents one area of any overwater bridge; determining the minimum circumscribed rectangle of a plurality of overlapped rotating frames as the region of an overwater bridge based on the positions of the rotating frames, and determining the position and the width of the minimum circumscribed rectangle as the position of the region of the overwater bridge and the width of the corresponding region;
identifying a pedestrian bridge of the current remote sensing image based on a preset semantic segmentation model to obtain all connected domains in a segmentation result output by the semantic segmentation model; determining each connected domain as a pedestrian bridge area, and determining the position and the width of the corresponding connected domain as the position of the pedestrian bridge area and the width of the corresponding area;
detecting the overpass of the current remote sensing image based on a preset target recognition network to obtain a rectangular frame output by the target recognition network; determining each rectangular frame as an area of the overpass, and determining the position and the width of the corresponding rectangular frame as the position of the area of the overpass and the width of the corresponding area;
the different types of roads comprise a first type of road intersected with the overpass and a second type of road not intersected with the overpass;
updating the recognition result of the overpass in the first recognition result based on the incidence relation between the road recognition result and the position of the area of the overpass in the first recognition result to obtain a third recognition result, wherein the third recognition result comprises the following steps:
determining a first type of road in the different types of roads;
detecting whether an overlapping area exists between the position of the area of each road in the first type of road and the position of the area of any overpass;
if the overlapped area does not exist, deleting the identification result corresponding to the corresponding overpass in the first identification result to obtain a third identification result;
and if the overlapped area exists, keeping the identification result corresponding to the corresponding overpass in the first identification result to obtain a third identification result.
2. The method of claim 1, wherein determining the second recognition result based on the width of the area of any one of the water bridges and/or the overlapping rate of the area of any one of the water bridges and the area of any one of the footbridge in the first recognition result comprises:
if the width of the region of any overwater bridge is not larger than a preset width threshold value, the bridge type of the corresponding overwater bridge in the first identification result is updated to be the pedestrian bridge, and/or if the coincidence rate of the region of any overwater bridge and the region of any pedestrian bridge is not smaller than a preset coincidence threshold value, the identification result corresponding to the corresponding pedestrian bridge in the first identification result is deleted, and a second identification result is obtained.
3. The method of claim 2, wherein the method further comprises:
and if the width of the area of any water bridge is larger than a preset width threshold value and/or the coincidence rate of the area of any water bridge and the area of any pedestrian bridge is smaller than a preset coincidence threshold value, determining that the obtained second identification result is the first identification result.
4. The method of claim 1, wherein obtaining the coincidence of the area of any marine bridge with the area of any pedestrian bridge comprises:
acquiring a first image corresponding to a region of any overwater bridge and a second image corresponding to a region of any pedestrian bridge;
and determining the ratio of the number of pixels of the superposed image of the first image and the second image to the number of pixels of the second image as the superposition rate of the area of any overwater bridge and the area of any pedestrian bridge.
5. The method of claim 1, wherein after obtaining the third recognition result, the method further comprises:
acquiring all connected regions and corresponding connected region areas in the region of each overpass in the third recognition result;
and reserving the target connected domain with the largest area in each overpass region, and deleting the other connected domains except the target connected domain in the corresponding region.
6. An identification device for a bridge in a remote sensing image, the device comprising:
the acquisition unit is used for respectively acquiring first identification results of an underwater bridge, a pedestrian bridge and an overpass in the current remote sensing image based on a bridge identification network; the first identification result comprises the position of the area of the corresponding type of bridge and the width of the corresponding area;
the determining unit is used for determining a second recognition result based on the width of the region of any one overwater bridge and/or the coincidence rate of the region of any one overwater bridge and the region of any one pedestrian bridge in the first recognition result;
the identification unit is used for identifying the road of the current remote sensing image based on a road identification network to obtain road identification results of different types of roads, and the road identification results comprise the positions of areas of the corresponding types of roads;
the updating unit is used for updating the identification result of the overpass in the first identification result based on the incidence relation between the road identification result and the position of the area of the overpass in the first identification result to obtain a third identification result;
the determining unit is further configured to determine a bridge identification result corresponding to the current remote sensing image based on a second identification result and a third identification result;
the obtaining unit is specifically configured to:
identifying the overwater bridge of the current remote sensing image based on a preset rotating frame identification model to obtain the positions of a plurality of rotating frames output by the rotating frame identification model; each rotating frame represents one area of any overwater bridge; determining the minimum circumscribed rectangle of a plurality of overlapped rotating frames as the region of an overwater bridge based on the positions of the rotating frames, and determining the position and the width of the minimum circumscribed rectangle as the position of the region of the overwater bridge and the width of the corresponding region;
identifying a pedestrian bridge of the current remote sensing image based on a preset semantic segmentation model to obtain all connected domains in a segmentation result output by the semantic segmentation model; determining each connected domain as a pedestrian bridge area, and determining the position and the width of the corresponding connected domain as the position of the pedestrian bridge area and the width of the corresponding area;
detecting the overpass of the current remote sensing image based on a preset target recognition network to obtain a rectangular frame output by the target recognition network; determining each rectangular frame as an area of an overpass, and determining the position and the width of the corresponding rectangular frame as the position of the area of the overpass and the width of the corresponding area;
the different types of roads comprise a first type of road intersected with the overpass and a second type of road not intersected with the overpass; the update unit is specifically configured to:
determining a first type of road in the different types of roads;
detecting whether an overlapping area exists between the position of the area of each road in the first type of road and the position of the area of any overpass;
if the overlapped area does not exist, deleting the identification result corresponding to the corresponding overpass in the first identification result to obtain a third identification result;
and if the overlapped area exists, keeping the identification result corresponding to the corresponding overpass in the first identification result to obtain a third identification result.
7. An electronic device, characterized in that the electronic device comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory are communicated with each other through the communication bus;
a memory for storing a computer program;
a processor for implementing the method steps of any of claims 1-5 when executing a program stored on a memory.
8. A computer-readable storage medium, characterized in that a computer program is stored in the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of the claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211587457.8A CN115661666B (en) | 2022-12-12 | 2022-12-12 | Bridge identification method and device in remote sensing image, electronic equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211587457.8A CN115661666B (en) | 2022-12-12 | 2022-12-12 | Bridge identification method and device in remote sensing image, electronic equipment and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115661666A CN115661666A (en) | 2023-01-31 |
| CN115661666B true CN115661666B (en) | 2023-04-07 |
Family
ID=85019260
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211587457.8A Active CN115661666B (en) | 2022-12-12 | 2022-12-12 | Bridge identification method and device in remote sensing image, electronic equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115661666B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112052777A (en) * | 2020-09-01 | 2020-12-08 | 国交空间信息技术（北京）有限公司 | Cross-water bridge extraction method and device based on high-resolution remote sensing image |
| CN113989632A (en) * | 2021-09-13 | 2022-01-28 | 西安电子科技大学 | Bridge detection method and device for remote sensing image, electronic equipment and storage medium |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101814144B (en) * | 2010-03-25 | 2012-03-07 | 西安电子科技大学 | Water-free bridge target identification method in remote sensing image |
| CN101976347A (en) * | 2010-10-21 | 2011-02-16 | 西北工业大学 | Method for recognizing overwater bridge in remote sensing image on basis of Mean Shift segmentation |
| US8798372B1 (en) * | 2012-03-07 | 2014-08-05 | Hrl Laboratories, Llc | Method for detecting bridges using lidar point cloud data |
| CN108921173B (en) * | 2018-06-01 | 2019-10-25 | 中南大学 | A kind of deep learning method of combination OSM and remote sensing image extraction overpass |
-
2022
- 2022-12-12 CN CN202211587457.8A patent/CN115661666B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112052777A (en) * | 2020-09-01 | 2020-12-08 | 国交空间信息技术（北京）有限公司 | Cross-water bridge extraction method and device based on high-resolution remote sensing image |
| CN113989632A (en) * | 2021-09-13 | 2022-01-28 | 西安电子科技大学 | Bridge detection method and device for remote sensing image, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115661666A (en) | 2023-01-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109785637B (en) | Analysis and evaluation method and device for vehicle violation | |
| CN110991215B (en) | Lane line detection method and device, storage medium and electronic equipment | |
| CN113033497B (en) | Lane line identification method, device, equipment and computer readable storage medium | |
| CN115984273B (en) | Road disease detection method, device, computer equipment and readable storage medium | |
| US20220262111A1 (en) | Road surface inspection apparatus, road surface inspection method, and program | |
| EP4202365A1 (en) | Method, apparatus, and computer program product for identifying and correcting lane geometry in map data | |
| CN110889388A (en) | Violation identification method, device, equipment and storage medium | |
| CN111414878B (en) | Social attribute analysis and image processing method and device for land parcels | |
| CN115661666B (en) | Bridge identification method and device in remote sensing image, electronic equipment and medium | |
| CN113496182B (en) | Road extraction method and device based on remote sensing image, storage medium and equipment | |
| CN110057377B (en) | Path navigation method and related product | |
| CN116206326A (en) | Training method of missing detection model, missing detection method and device of diversion area | |
| CN115249407B (en) | Indicator light state identification method and device, electronic equipment, storage medium and product | |
| CN115114312A (en) | Map data updating method and device and electronic equipment | |
| CN114998863A (en) | Target road identification method, target road identification device, electronic equipment and storage medium | |
| CN115712749A (en) | Image processing method and device, computer equipment and storage medium | |
| CN114155508B (en) | Road change detection method, device, equipment and storage medium | |
| CN114863285B (en) | Method, device, equipment and storage medium for identifying target road | |
| CN115658839A (en) | POI data mining method and device, computer equipment and storage medium | |
| CN114926454B (en) | Parking space detection method and device and electronic equipment | |
| CN113837155B (en) | Image processing method, map data updating device and storage medium | |
| CN115982306B (en) | Method and device for identifying retrograde behavior of target object | |
| CN113642533B (en) | Lane level positioning method and electronic equipment | |
| EP4202364A1 (en) | Method, apparatus, and computer program product for identifying and correcting intersection lane geometry in map data | |
| JP2006178694A (en) | Method for detecting candidate of constitutional segment of road mark image, and program capable of detecting candidate of constitutional segment of road mark image |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |