CN113108792A - Wi-Fi fingerprint map reconstruction method and device, terminal equipment and medium - Google Patents
Wi-Fi fingerprint map reconstruction method and device, terminal equipment and medium Download PDFInfo
- Publication number
- CN113108792A CN113108792A CN202110289719.1A CN202110289719A CN113108792A CN 113108792 A CN113108792 A CN 113108792A CN 202110289719 A CN202110289719 A CN 202110289719A CN 113108792 A CN113108792 A CN 113108792A
- Authority
- CN
- China
- Prior art keywords
- resolution
- image
- super
- map
- fingerprint
- 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
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000012549 training Methods 0.000 claims abstract description 41
- 238000003062 neural network model Methods 0.000 claims abstract description 28
- 238000012360 testing method Methods 0.000 claims description 68
- 238000004590 computer program Methods 0.000 claims description 21
- 230000006870 function Effects 0.000 claims description 14
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 claims description 11
- 238000012795 verification Methods 0.000 claims description 10
- 238000005070 sampling Methods 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 8
- 239000011159 matrix material Substances 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 239000000126 substance Substances 0.000 claims description 4
- 238000013480 data collection Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 9
- 238000010586 diagram Methods 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000013508 migration Methods 0.000 description 2
- 230000005012 migration Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000005389 magnetism Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
- G01C21/206—Instruments for performing navigational calculations specially adapted for indoor navigation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/02—Systems using the reflection of electromagnetic waves other than radio waves
- G01S17/06—Systems determining position data of a target
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/86—Combinations of lidar systems with systems other than lidar, radar or sonar, e.g. with direction finders
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
- G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
- G01S5/0257—Hybrid positioning
Abstract
The invention discloses a Wi-Fi fingerprint map reconstruction method, a device, terminal equipment and a medium, wherein the method comprises the following steps: constructing a scene map, collecting Wi-Fi fingerprint data of collection points on the scene map, and generating a Wi-Fi fingerprint map; acquiring a signal intensity value corresponding to any AP access point according to the Wi-Fi fingerprint map, and generating a corresponding signal intensity distribution image; carrying out nearest neighbor downsampling operation on the signal intensity distribution image to obtain a low-resolution image; training a preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model; and (4) taking the low-resolution images corresponding to all the AP access points as the input of the super-resolution reconstruction model, and integrating the super-resolution images output by the super-resolution reconstruction model into a reconstructed Wi-Fi fingerprint map. The method can effectively solve the problem that the reconstructed Wi-Fi fingerprint map is smooth and loses details in the prior art, improves the reconstruction effect, and greatly reduces the training time.
Description
Technical Field
The invention relates to the technical field of indoor positioning, in particular to a Wi-Fi fingerprint map reconstruction method, a Wi-Fi fingerprint map reconstruction device, terminal equipment and a medium.
Background
The satellite-based navigation system is widely popularized outdoors and obtains higher precision, but the positioning precision in an indoor scene is obviously reduced due to the shielding of the wall of a building, so that the precision requirement of indoor positioning cannot be met.
Indoor positioning refers to the realization of position location in an indoor environment, and along with the continuous development of information technology, the practicability and the necessity of indoor positioning become more and more remarkable. Today's major indoor location technologies include Wi-Fi, terrestrial magnetism, bluetooth, Radio Frequency Identification (RFID), acoustic, visible, and computer vision technologies, among others. With the wide application of wireless Ethernet and the popularization of Wi-Fi mobile equipment, the cost of implementing the indoor positioning scheme based on Wi-Fi is reduced, the positioning precision can generally reach about 3-5m, and a better positioning effect is achieved indoors.
The core operation of Wi-Fi fingerprint positioning is to match the acquired information value with a Wi-Fi fingerprint map (Wi-Fi signal database) established previously to determine the positioning result of a point to be measured. The algorithm is mainly divided into an off-line acquisition stage and an on-line positioning stage. And the Wi-Fi fingerprint map reconstruction is to generate a dense Wi-Fi fingerprint map by combining a sparse Wi-Fi fingerprint map with information such as scenes, so that the cost of acquiring and maintaining a database is reduced.
In order to improve the positioning accuracy and reduce the acquisition cost, the current research on the reconstruction of the Wi-Fi fingerprint map mainly focuses on three directions: firstly, establishing a signal propagation model: defining a propagation model (such as a logarithmic distance path loss model, an attenuation factor model and the like) to predict position information, wherein the model is easy to fail when being subjected to external interference and scene change; secondly, adopting a crowdsourcing mode to update: the fingerprint map is updated by artificially acquiring data, the acquisition workload of the method is large, and invalid data are increased due to different data densities of all areas of the fingerprint map; thirdly, based on deep learning: the mapping relation from the sparse map to the dense map is learned through the neural network, the workload of fingerprint data acquisition is reduced, and the technology also belongs to Wi-Fi fingerprint map reconstruction based on deep learning.
At present, the Wi-Fi fingerprint map reconstruction research mainly has the following problems: compared with the original map, the fine-grained map obtained by the existing Wi-Fi fingerprint map reconstruction method is smoother, loses partial details and has longer off-line training time.
Disclosure of Invention
The embodiment of the invention provides a method, a device, a terminal device and a medium for reconstructing a Wi-Fi fingerprint map, which can effectively solve the problem that the reconstructed Wi-Fi fingerprint map is smooth and loses details in the prior art, effectively improve the information content of the reconstructed Wi-Fi fingerprint map, improve the reconstruction effect and greatly reduce the training time.
An embodiment of the present invention provides a Wi-Fi fingerprint map reconstruction method, including:
constructing a scene map, collecting Wi-Fi fingerprint data of collection points on the scene map, and generating a Wi-Fi fingerprint map;
acquiring a signal intensity value corresponding to any AP access point according to the Wi-Fi fingerprint map, and generating a corresponding signal intensity distribution image;
carrying out nearest neighbor downsampling operation on the signal intensity distribution image to obtain a low-resolution image;
training a preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model;
and taking the low-resolution images corresponding to all the AP access points as the input of the super-resolution reconstruction model, and integrating the super-resolution images output by the super-resolution reconstruction model into a reconstructed Wi-Fi fingerprint map.
In some embodiments, the building a scene map and collecting Wi-Fi fingerprint data of collection points on the scene map to generate a Wi-Fi fingerprint map includes:
carrying a laser radar by using a mobile robot to construct a scene map, and acquiring acquisition point coordinates corresponding to an acquisition point set in the scene map;
carrying Wi-Fi signal acquisition equipment by using the mobile robot, and acquiring data of the acquisition points to obtain Wi-Fi fingerprint data of the acquisition points, wherein the Wi-Fi fingerprint data comprises Wi-Fi signal strength, an MAC address, an AP access point and acquisition time;
and summarizing the Wi-Fi fingerprint data of all acquisition points under different AP access points to generate a Wi-Fi fingerprint map.
In some embodiments, the training a preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model includes:
dividing the low-resolution image into an image training sample, an image verification sample and an image test sample according to a preset proportion;
training a preset neural network model by using the image training sample to obtain a super-resolution reconstruction model;
performing parameter tuning on the super-resolution reconstruction model by using the image verification sample;
and testing the optimized super-resolution reconstruction model by using the image test sample.
In some embodiments, the neural network model comprises an input module, a feature extraction module and an output module which are connected in sequence; the input module is an input convolution layer; the feature extraction module is composed of 16 residual blocks, and each residual block comprises 2 convolution layers and 1 ReLu active layer; the output module comprises an up-sampling layer and an output convolution layer which are connected.
In some embodiments, the testing the tuned super-resolution reconstruction model with the image test sample includes:
inputting the image test sample into the super-resolution reconstruction model, outputting a super-resolution test image, and integrating the super-resolution test image into a reconstructed test Wi-Fi fingerprint map;
selecting a test acquisition point, and acquiring Wi-Fi fingerprint data corresponding to the test acquisition point from the reconstructed test Wi-Fi fingerprint map;
calculating the coordinates of the test acquisition points according to the Wi-Fi fingerprint data corresponding to the test acquisition points;
and comparing the calculated coordinates with the coordinates of the acquisition points corresponding to the test acquisition points in the Wi-Fi fingerprint map.
In some embodiments, the method further comprises:
data collection is carried out on the collection points by using the mobile robot regularly so as to update the Wi-Fi fingerprint map;
and inputting the updated low-resolution image samples into the super-resolution reconstruction model according to the time sequence for training.
In some embodiments, the neural network model is trained by the following loss function, in particular equation (1):
wherein the content of the first and second substances,
the average absolute error of the super-resolution image, WH is the length and width of the image matrix,
for high resolution images corresponding to the signal intensity distribution image, ISR x,yIs a super-resolution image.
Another embodiment of the present invention correspondingly provides a device for reconstructing a Wi-Fi fingerprint map, including:
the Wi-Fi fingerprint map generation module is used for constructing a scene map, acquiring Wi-Fi fingerprint data of acquisition points on the scene map and generating a Wi-Fi fingerprint map;
the signal intensity distribution image generation module is used for acquiring a signal intensity value corresponding to any AP access point according to the Wi-Fi fingerprint map and generating a corresponding signal intensity distribution image;
the image processing module is used for carrying out nearest neighbor down-sampling operation on the signal intensity distribution image to obtain a low-resolution image;
the model training module is used for training a preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model;
and the Wi-Fi fingerprint map reconstruction module is used for taking the low-resolution images corresponding to all the AP access points as the input of the super-resolution reconstruction model, and integrating the super-resolution images output by the super-resolution reconstruction model into a reconstructed Wi-Fi fingerprint map.
Another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, the processor implements the Wi-Fi fingerprint map reconstruction method according to the above embodiment of the present invention.
Another embodiment of the present invention provides a storage medium, where the computer-readable storage medium includes a stored computer program, where when the computer program runs, the apparatus where the computer-readable storage medium is located is controlled to execute the Wi-Fi fingerprint map reconstruction method according to the above-described embodiment of the present invention.
Compared with the prior art, the Wi-Fi fingerprint map reconstruction method disclosed by the embodiment of the invention comprises the steps of constructing a scene map, acquiring Wi-Fi fingerprint data of acquisition points on the scene map, generating a Wi-Fi fingerprint map, acquiring a signal intensity value corresponding to any AP access point according to the Wi-Fi fingerprint map, generating a corresponding signal intensity distribution image, performing nearest neighbor down-sampling operation on the signal intensity distribution image to obtain a low-resolution image, training a preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model, taking the low-resolution images corresponding to all AP access points as input of the super-resolution reconstruction model, integrating the super-resolution images output by the super-resolution reconstruction model into a reconstructed Wi-Fi fingerprint map, therefore, the method based on the residual error network super-resolution can effectively solve the problem that the reconstructed Wi-Fi fingerprint map is smooth and loses details in the prior art, effectively improve the information quantity of the reconstructed Wi-Fi fingerprint map, have more detail information, improve the reconstruction effect and greatly reduce the training time.
Drawings
Fig. 1 is a schematic flowchart of a Wi-Fi fingerprint map reconstruction method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a scene map provided by an embodiment of the invention;
FIG. 3 is a schematic diagram of a Wi-Fi fingerprint map provided by an embodiment of the invention;
FIG. 4 is a Wi-Fi signal strength profile of an AP access point according to an embodiment of the present invention;
FIG. 5 is a schematic structural diagram of a neural network model provided in an embodiment of the present invention;
fig. 6 is a schematic structural diagram of a Wi-Fi fingerprint map reconstruction apparatus according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a schematic flowchart of a Wi-Fi fingerprint map reconstruction method according to an embodiment of the present invention is shown, where the method includes steps S101 to S105.
S101, constructing a scene map, collecting Wi-Fi fingerprint data of collection points on the scene map, and generating the Wi-Fi fingerprint map.
In some embodiments, referring to fig. 2, which is a schematic view of a scene map provided in an embodiment of the present invention, step S101 includes:
carrying a laser radar by using a mobile robot to construct a scene map, and acquiring acquisition point coordinates corresponding to an acquisition point set in the scene map;
carrying Wi-Fi signal acquisition equipment by using the mobile robot, and acquiring data of the acquisition points to obtain Wi-Fi fingerprint data of the acquisition points, wherein the Wi-Fi fingerprint data comprises Wi-Fi signal strength, an MAC address, an AP access point and acquisition time;
and summarizing the Wi-Fi fingerprint data of all acquisition points under different AP access points to generate a Wi-Fi fingerprint map.
Specifically, the Wi-Fi signal acquisition equipment is an instrument or equipment with a Wi-Fi signal acquisition function. The mobile robot can be a mobile intelligent monitoring device such as a robot trolley and an unmanned aerial vehicle. For example, the mobile robot is a robot car Turtlebot2, a scene map is constructed by using an open source mapping program package by loading an RPLIDAR-A3 laser radar on the robot car, and the number of acquisition points and the position coordinates thereof are marked on the map, as shown in fig. 2, and 21 in fig. 2 is an acquisition point. Further, the robot trolley carries Wi-Fi signal acquisition equipment, data acquisition is carried out around a field according to the position of an acquisition point marked on a scene map, and the acquired Wi-Fi signal strength, the MAC address, the AP access point (namely the equipment name), the acquisition time, the coordinates of the acquisition point and other data are used for generating Wi-Fi fingerprint data, wherein the Wi-Fi fingerprint data are shown in the following table 1. Referring to fig. 3, which is a schematic view of a Wi-Fi fingerprint map provided in an embodiment of the present invention, a mobile robot collects Wi-Fi signal strengths from different AP Access points at each collection Point, and classifies the Wi-Fi fingerprint map according to different Wi-Fi signal sending devices (Access points, APs), so that Wi-Fi fingerprint data of each collection Point under all AP Access points are collected to obtain a fingerprint database, that is, the Wi-Fi fingerprint map.
| Gathering items | Sample examples |
| MAC address | 08:00:20:0A:8C:6D |
| Device name | Example (c): TP-Link xxx |
| Wi-Fi Signal Strength (RSSI) | Example (c): x dB |
| Coordinates of acquisition points | Example (c): (x, y) |
TABLE 1
Therefore, Wi-Fi signal acquisition and updating are performed through the mobile robot in the embodiment, the map maintenance cost and the labor cost are reduced, and the signal acquisition efficiency is greatly improved.
S102, acquiring a signal intensity value corresponding to any AP access point according to the Wi-Fi fingerprint map, and generating a corresponding signal intensity distribution image.
S103, carrying out nearest neighbor down-sampling operation on the signal intensity distribution image to obtain a low-resolution image.
Specifically, fig. 4 is a Wi-Fi signal strength distribution diagram of an AP access point according to an embodiment of the present invention, fig. 4(a) is a Wi-Fi signal strength digital distribution diagram of the AP access point, and fig. 4(b) is a Wi-Fi signal strength distribution diagram of the AP access point. Acquiring a signal intensity value corresponding to an AP access point, as shown in fig. 4(a), converting the signal intensity into a visual image according to a pixel and a numerical ratio thereof, as shown in fig. 4(b), generating an original HR (High Resolution) data set, performing nearest neighbor downsampling on an image of the HR data set, and generating a corresponding LR (Low Resolution) data set.
And S104, training a preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model.
In some embodiments, step S104 includes:
dividing the low-resolution image into an image training sample, an image verification sample and an image test sample according to a preset proportion;
training a preset neural network model by using the image training sample to obtain a super-resolution reconstruction model;
performing parameter tuning on the super-resolution reconstruction model by using the image verification sample;
and testing the optimized super-resolution reconstruction model by using the image test sample.
Illustratively, the low resolution image is displayed as 8: 1: the scale of 1 is divided into an image training sample, an image verification sample, and an image test sample.
On the basis of the foregoing embodiments, in some embodiments, referring to fig. 5, the present invention is a schematic structural diagram of a neural network model provided in an embodiment of the present invention, where the neural network model includes an input module, a feature extraction module, and an output module, which are connected in sequence; the input module is an input convolutional layer (Conv); the feature extraction module is composed of 16 residual blocks (ResBlock), each of which contains 2 convolutional layers (Conv) and 1 ReLu active layer (ReLu); the output module comprises an upsampling layer (UnSample) and an output convolutional layer (Conv) which are connected.
In the present embodiment, please refer to fig. 5, a LR (Low Resolution) image is used as an input of the neural network model, and an SR (Super Resolution) image is used as an output of the neural network model. Specifically, an LR image is input to an input convolutional layer (Conv), the input convolutional layer is connected to 16 residual blocks (ResBlock), in which a ReLu active layer is connected between two convolutional layers, and the residual blocks are also connected to an upsampling layer (UnSample), which is further connected to an output convolutional layer, thereby outputting an SR image.
Preferably, the neural network model is trained by the following loss function, specifically the following formula (1):
wherein the content of the first and second substances,
the average absolute error of the super-resolution image, WH is the length and width of the image matrix,
for high resolution images corresponding to the signal intensity distribution image, ISR x,yFor super-resolution images, the loss function is specifically the L1 norm loss function.
In some embodiments, the testing the tuned super-resolution reconstruction model with the image test sample includes:
inputting the image test sample into the super-resolution reconstruction model, outputting a super-resolution test image, and integrating the super-resolution test image into a reconstructed test Wi-Fi fingerprint map;
selecting a test acquisition point, and acquiring Wi-Fi fingerprint data corresponding to the test acquisition point from the reconstructed test Wi-Fi fingerprint map;
calculating the coordinates of the test acquisition points according to the Wi-Fi fingerprint data corresponding to the test acquisition points;
and comparing the calculated coordinates with the coordinates of the acquisition points corresponding to the test acquisition points in the Wi-Fi fingerprint map.
In this embodiment, an LR image of an image test sample is input into a super-resolution reconstruction model to obtain an SR image, visual images of all APs are integrated into a reconstructed test Wi-Fi fingerprint map, that is, a test acquisition point can be selected in a scene, a Wi-Fi signal strength, an MAC address, and the like of the point in the reconstructed test Wi-Fi fingerprint map are obtained, coordinates of the point are calculated according to a KNN (K Nearest Neighbors) positioning algorithm, and the coordinates of the point are compared with coordinates of the point on the previously established Wi-Fi fingerprint map to evaluate positioning accuracy of the reconstructed fingerprint map.
And S105, taking the low-resolution images corresponding to all the AP access points as the input of the super-resolution reconstruction model, and integrating the super-resolution images output by the super-resolution reconstruction model into a reconstructed Wi-Fi fingerprint map.
In some embodiments, the method further comprises:
data collection is carried out on the collection points by using the mobile robot regularly so as to update the Wi-Fi fingerprint map;
and inputting the updated low-resolution image samples into the super-resolution reconstruction model according to the time sequence for training.
In the embodiment, the time constraint is introduced into the network, so that the model learns the dynamic change condition of the Wi-Fi signal along with the time, and the real-time effect of the reconstruction of the Wi-Fi fingerprint map is improved.
The Wi-Fi fingerprint map reconstruction method provided by the embodiment of the invention comprises the steps of constructing a scene map, acquiring Wi-Fi fingerprint data of acquisition points on the scene map, generating a Wi-Fi fingerprint map, acquiring a signal intensity value corresponding to any AP access point according to the Wi-Fi fingerprint map, generating a corresponding signal intensity distribution image, performing nearest neighbor down-sampling operation on the signal intensity distribution image to obtain a low-resolution image, training a preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model, taking the low-resolution images corresponding to all AP access points as input of the super-resolution reconstruction model, integrating the super-resolution images output by the super-resolution reconstruction model into a reconstructed Wi-Fi fingerprint map, therefore, the method based on the residual error network super-resolution can effectively solve the problem that the reconstructed Wi-Fi fingerprint map is smooth and loses details in the prior art, effectively improve the information quantity of the reconstructed Wi-Fi fingerprint map, have more detail information, improve the reconstruction effect and greatly reduce the training time. Meanwhile, the migration effect is improved by fine tuning the model by using a small amount of data of the new scene.
Referring to fig. 6, which is a schematic structural diagram of a Wi-Fi fingerprint map reconstruction apparatus according to an embodiment of the present invention, including:
the Wi-Fi fingerprint map generation module 201 is used for constructing a scene map, acquiring Wi-Fi fingerprint data of acquisition points on the scene map, and generating a Wi-Fi fingerprint map;
a signal intensity distribution image generation module 202, configured to obtain a signal intensity value corresponding to any AP access point according to the Wi-Fi fingerprint map, and generate a corresponding signal intensity distribution image;
the image processing module 203 is configured to perform nearest neighbor downsampling on the signal intensity distribution image to obtain a low-resolution image;
the model training module 204 is used for training a preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model;
and the Wi-Fi fingerprint map reconstruction module 205 is configured to use the low-resolution images corresponding to all the AP access points as input of the super-resolution reconstruction model, and integrate the super-resolution images output by the super-resolution reconstruction model into a reconstructed Wi-Fi fingerprint map.
Preferably, the Wi-Fi fingerprint map generation module 201 includes:
the scene map building unit is used for building a scene map by using a mobile robot to carry a laser radar and acquiring the coordinates of acquisition points corresponding to the acquisition points set in the scene map;
the Wi-Fi fingerprint data acquisition unit is used for carrying Wi-Fi signal acquisition equipment by using the mobile robot, acquiring data of the acquisition points and obtaining Wi-Fi fingerprint data of the acquisition points, wherein the Wi-Fi fingerprint data comprise Wi-Fi signal strength, MAC addresses, AP access points and acquisition time;
and the data summarizing unit is used for summarizing the Wi-Fi fingerprint data of all the acquisition points under different AP access points to generate a Wi-Fi fingerprint map.
Preferably, the model training module 204 includes:
the sample dividing unit is used for dividing the low-resolution image into an image training sample, an image verification sample and an image test sample according to a preset proportion;
the training unit is used for training a preset neural network model by adopting the image training sample to obtain a super-resolution reconstruction model;
the verification unit is used for performing parameter tuning on the super-resolution reconstruction model by adopting the image verification sample;
and the test unit is used for testing the adjusted super-resolution reconstruction model by adopting the image test sample.
Preferably, the neural network model comprises an input module, a feature extraction module and an output module which are connected in sequence; the input module is an input convolution layer; the feature extraction module is composed of 16 residual blocks, and each residual block comprises 2 convolution layers and 1 ReLu active layer; the output module comprises an up-sampling layer and an output convolution layer which are connected.
Preferably, the test unit includes:
the reconstructed test Wi-Fi fingerprint map construction unit is used for inputting the image test sample into the super-resolution reconstruction model, outputting a super-resolution test image and integrating the super-resolution test image into a reconstructed test Wi-Fi fingerprint map;
the test acquisition point information acquisition unit is used for selecting test acquisition points and acquiring Wi-Fi fingerprint data corresponding to the test acquisition points from the reconstructed test Wi-Fi fingerprint map;
the test acquisition point coordinate calculation unit is used for calculating the coordinates of the test acquisition points according to the Wi-Fi fingerprint data corresponding to the test acquisition points;
and the comparison unit is used for comparing the calculated coordinates with the coordinates of the acquisition points corresponding to the test acquisition points in the Wi-Fi fingerprint map.
Preferably, the apparatus further comprises:
the regular updating unit is used for regularly using the mobile robot to acquire data of the acquisition points so as to update the Wi-Fi fingerprint map;
and the model training and updating unit is used for inputting the updated low-resolution image samples into the super-resolution reconstruction model according to the time sequence for training.
Preferably, the model training module 204 includes:
a loss function calculation unit, configured to train the neural network model through the following loss function, specifically the following formula (1):
wherein the content of the first and second substances,
the average absolute error of the super-resolution image, WH is the length and width of the image matrix,
for high resolution images corresponding to the signal intensity distribution image, ISR x,yIs a super-resolution image.
The Wi-Fi fingerprint map reconstruction device provided by the embodiment of the invention generates a Wi-Fi fingerprint map by constructing a scene map and acquiring Wi-Fi fingerprint data of acquisition points on the scene map, acquires a signal intensity value corresponding to any AP access point according to the Wi-Fi fingerprint map, generates a corresponding signal intensity distribution image, performs nearest neighbor down-sampling operation on the signal intensity distribution image to obtain a low-resolution image, trains a preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model, thereby taking the low-resolution images corresponding to all AP access points as the input of the super-resolution reconstruction model, integrates the super-resolution images output by the super-resolution reconstruction model into a reconstructed Wi-Fi fingerprint map, therefore, the method based on the residual error network super-resolution can effectively solve the problem that the reconstructed Wi-Fi fingerprint map is smooth and loses details in the prior art, effectively improve the information quantity of the reconstructed Wi-Fi fingerprint map, have more detail information, improve the reconstruction effect and greatly reduce the training time. Meanwhile, the migration effect is improved by fine tuning the model by using a small amount of data of the new scene. The terminal device of this embodiment includes: a processor, a memory, and a computer program, such as a Wi-Fi fingerprint reconstruction program, stored in the memory and executable on the processor. The processor implements the steps in the various Wi-Fi fingerprint map reconstruction method embodiments described above when executing the computer program. Alternatively, the processor implements the functions of the modules/units in the above device embodiments when executing the computer program.
Illustratively, the computer program may be partitioned into one or more modules/units that are stored in the memory and executed by the processor to implement the invention. The one or more modules/units may be a series of computer program instruction segments capable of performing specific functions, which are used for describing the execution process of the computer program in the terminal device.
The terminal device can be a desktop computer, a notebook, a palm computer, a cloud server and other computing devices. The terminal device may include, but is not limited to, a processor, a memory.
The Processor may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may be any conventional processor or the like, which is the control center of the terminal device and connects the various parts of the whole terminal device using various interfaces and lines.
The memory may be used for storing the computer programs and/or modules, and the processor may implement various functions of the terminal device by executing or executing the computer programs and/or modules stored in the memory and calling data stored in the memory. The memory may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required by at least one function (such as a sound playing function, an image playing function, etc.), and the like; the storage data area may store data (such as audio data, a phonebook, etc.) created according to the use of the cellular phone, and the like. In addition, the memory may include high speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), at least one magnetic disk storage device, a Flash memory device, or other volatile solid state storage device.
Wherein, the terminal device integrated module/unit can be stored in a computer readable storage medium if it is implemented in the form of software functional unit and sold or used as a stand-alone product. Based on such understanding, all or part of the flow of the method according to the embodiments of the present invention may also be implemented by a computer program, which may be stored in a computer-readable storage medium, and when the computer program is executed by a processor, the steps of the method embodiments may be implemented. Wherein the computer program comprises computer program code, which may be in the form of source code, object code, an executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, usb disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM), Random Access Memory (RAM), electrical carrier wave signals, telecommunications signals, software distribution medium, and the like. It should be noted that the computer readable medium may contain content that is subject to appropriate increase or decrease as required by legislation and patent practice in jurisdictions, for example, in some jurisdictions, computer readable media does not include electrical carrier signals and telecommunications signals as is required by legislation and patent practice.
It should be noted that the above-described device embodiments are merely illustrative, where the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. In addition, in the drawings of the embodiment of the apparatus provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, and may be specifically implemented as one or more communication buses or signal lines. One of ordinary skill in the art can understand and implement it without inventive effort.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.
Claims (10)
1. A Wi-Fi fingerprint map reconstruction method is characterized by comprising the following steps:
constructing a scene map, collecting Wi-Fi fingerprint data of collection points on the scene map, and generating a Wi-Fi fingerprint map;
acquiring a signal intensity value corresponding to any AP access point according to the Wi-Fi fingerprint map, and generating a corresponding signal intensity distribution image;
carrying out nearest neighbor downsampling operation on the signal intensity distribution image to obtain a low-resolution image;
training a preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model;
and taking the low-resolution images corresponding to all the AP access points as the input of the super-resolution reconstruction model, and integrating the super-resolution images output by the super-resolution reconstruction model into a reconstructed Wi-Fi fingerprint map.
2. The Wi-Fi fingerprint map reconstruction method of claim 1, wherein the constructing a scene map and acquiring Wi-Fi fingerprint data of acquisition points on the scene map to generate a Wi-Fi fingerprint map comprises:
carrying a laser radar by using a mobile robot to construct a scene map, and acquiring acquisition point coordinates corresponding to an acquisition point set in the scene map;
carrying Wi-Fi signal acquisition equipment by using the mobile robot, and acquiring data of the acquisition points to obtain Wi-Fi fingerprint data of the acquisition points, wherein the Wi-Fi fingerprint data comprises Wi-Fi signal strength, an MAC address, an AP access point and acquisition time;
and summarizing the Wi-Fi fingerprint data of all acquisition points under different AP access points to generate a Wi-Fi fingerprint map.
3. The Wi-Fi fingerprint map reconstruction method of claim 2, wherein the training of the preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model comprises:
dividing the low-resolution image into an image training sample, an image verification sample and an image test sample according to a preset proportion;
training a preset neural network model by using the image training sample to obtain a super-resolution reconstruction model;
performing parameter tuning on the super-resolution reconstruction model by using the image verification sample;
and testing the optimized super-resolution reconstruction model by using the image test sample.
4. The Wi-Fi fingerprint map reconstruction method of claim 3, wherein the neural network model comprises an input module, a feature extraction module, and an output module connected in sequence; the input module is an input convolution layer; the feature extraction module is composed of 16 residual blocks, and each residual block comprises 2 convolution layers and 1 ReLu active layer; the output module comprises an up-sampling layer and an output convolution layer which are connected.
5. The Wi-Fi fingerprint map reconstruction method of claim 3, wherein the testing the optimized super-resolution reconstruction model with the image test samples comprises:
inputting the image test sample into the super-resolution reconstruction model, outputting a super-resolution test image, and integrating the super-resolution test image into a reconstructed test Wi-Fi fingerprint map;
selecting a test acquisition point, and acquiring Wi-Fi fingerprint data corresponding to the test acquisition point from the reconstructed test Wi-Fi fingerprint map;
calculating the coordinates of the test acquisition points according to the Wi-Fi fingerprint data corresponding to the test acquisition points;
and comparing the calculated coordinates with the coordinates of the acquisition points corresponding to the test acquisition points in the Wi-Fi fingerprint map.
6. The Wi-Fi fingerprint map reconstruction method of claim 1, wherein the method further comprises:
data collection is carried out on the collection points by using the mobile robot regularly so as to update the Wi-Fi fingerprint map;
and inputting the updated low-resolution image samples into the super-resolution reconstruction model according to the time sequence for training.
7. The Wi-Fi fingerprint map reconstruction method of claim 4, wherein the neural network model is trained with a loss function, in particular the following equation (1):
wherein the content of the first and second substances,
the average absolute error of the super-resolution image, WH is the length and width of the image matrix,
for high resolution images corresponding to the signal intensity distribution image, ISR x,yIs a super-resolution image.
8. A Wi-Fi fingerprint map reconstruction device, comprising:
the Wi-Fi fingerprint map generation module is used for constructing a scene map, acquiring Wi-Fi fingerprint data of acquisition points on the scene map and generating a Wi-Fi fingerprint map;
the signal intensity distribution image generation module is used for acquiring a signal intensity value corresponding to any AP access point according to the Wi-Fi fingerprint map and generating a corresponding signal intensity distribution image;
the image processing module is used for carrying out nearest neighbor down-sampling operation on the signal intensity distribution image to obtain a low-resolution image;
the model training module is used for training a preset neural network model according to the low-resolution image to obtain a super-resolution reconstruction model;
and the Wi-Fi fingerprint map reconstruction module is used for taking the low-resolution images corresponding to all the AP access points as the input of the super-resolution reconstruction model, and integrating the super-resolution images output by the super-resolution reconstruction model into a reconstructed Wi-Fi fingerprint map.
9. A terminal device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, the processor when executing the computer program implementing the Wi-Fi fingerprint map reconstruction method of any one of claims 1 to 7.
10. A computer-readable storage medium, comprising a stored computer program, wherein the computer program, when executed, controls an apparatus on which the computer-readable storage medium is located to perform the Wi-Fi fingerprint map reconstruction method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110289719.1A CN113108792A (en) | 2021-03-16 | 2021-03-16 | Wi-Fi fingerprint map reconstruction method and device, terminal equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110289719.1A CN113108792A (en) | 2021-03-16 | 2021-03-16 | Wi-Fi fingerprint map reconstruction method and device, terminal equipment and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113108792A true CN113108792A (en) | 2021-07-13 |
Family
ID=76711877
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110289719.1A Pending CN113108792A (en) | 2021-03-16 | 2021-03-16 | Wi-Fi fingerprint map reconstruction method and device, terminal equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113108792A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113556680A (en) * | 2021-07-20 | 2021-10-26 | 汤恩智能科技(上海)有限公司 | Fingerprint data processing method, medium and mobile robot |
| CN114326721A (en) * | 2021-12-20 | 2022-04-12 | 达闼机器人有限公司 | Drawing establishing method and device, cloud server and robot |
| CN114710831A (en) * | 2022-03-10 | 2022-07-05 | 南京市地铁交通设施保护办公室 | RFID label positioning system based on deep learning |
| CN114916059A (en) * | 2022-04-29 | 2022-08-16 | 湖南大学 | WiFi fingerprint sparse map extension method based on interval random logarithm shadow model |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107958246A (en) * | 2018-01-17 | 2018-04-24 | 深圳市唯特视科技有限公司 | A kind of image alignment method based on new end-to-end human face super-resolution network |
| CN108462992A (en) * | 2018-03-05 | 2018-08-28 | 中山大学 | A kind of indoor orientation method based on super-resolution rebuilding Wi-Fi fingerprint maps |
| CN109308725A (en) * | 2018-08-29 | 2019-02-05 | 华南理工大学 | A kind of system that expression interest figure in mobile terminal generates |
| CN110223224A (en) * | 2019-04-29 | 2019-09-10 | 杰创智能科技股份有限公司 | A kind of Image Super-resolution realization algorithm based on information filtering network |
| CN110300370A (en) * | 2019-07-02 | 2019-10-01 | 广州纳斯威尔信息技术有限公司 | A kind of reconstruction wifi fingerprint map indoor orientation method |
| CN110705699A (en) * | 2019-10-18 | 2020-01-17 | 厦门美图之家科技有限公司 | Super-resolution reconstruction method and device, electronic equipment and readable storage medium |
| CN111833251A (en) * | 2020-07-13 | 2020-10-27 | 北京安德医智科技有限公司 | Three-dimensional medical image super-resolution reconstruction method and device |
| CN111983635A (en) * | 2020-08-17 | 2020-11-24 | 浙江商汤科技开发有限公司 | Pose determination method and device, electronic equipment and storage medium |
-
2021
- 2021-03-16 CN CN202110289719.1A patent/CN113108792A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107958246A (en) * | 2018-01-17 | 2018-04-24 | 深圳市唯特视科技有限公司 | A kind of image alignment method based on new end-to-end human face super-resolution network |
| CN108462992A (en) * | 2018-03-05 | 2018-08-28 | 中山大学 | A kind of indoor orientation method based on super-resolution rebuilding Wi-Fi fingerprint maps |
| CN109308725A (en) * | 2018-08-29 | 2019-02-05 | 华南理工大学 | A kind of system that expression interest figure in mobile terminal generates |
| CN110223224A (en) * | 2019-04-29 | 2019-09-10 | 杰创智能科技股份有限公司 | A kind of Image Super-resolution realization algorithm based on information filtering network |
| CN110300370A (en) * | 2019-07-02 | 2019-10-01 | 广州纳斯威尔信息技术有限公司 | A kind of reconstruction wifi fingerprint map indoor orientation method |
| CN110705699A (en) * | 2019-10-18 | 2020-01-17 | 厦门美图之家科技有限公司 | Super-resolution reconstruction method and device, electronic equipment and readable storage medium |
| CN111833251A (en) * | 2020-07-13 | 2020-10-27 | 北京安德医智科技有限公司 | Three-dimensional medical image super-resolution reconstruction method and device |
| CN111983635A (en) * | 2020-08-17 | 2020-11-24 | 浙江商汤科技开发有限公司 | Pose determination method and device, electronic equipment and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| 史振威等: "图像超分辨重建算法综述", 数据采集与处理 * |
| 陈赛健等: "基于生成对抗网络的文本图像联合超分辨率与去模糊方法", 计算机应用 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113556680A (en) * | 2021-07-20 | 2021-10-26 | 汤恩智能科技(上海)有限公司 | Fingerprint data processing method, medium and mobile robot |
| CN113556680B (en) * | 2021-07-20 | 2023-09-29 | 汤恩智能科技(上海)有限公司 | Fingerprint data processing method, medium and mobile robot |
| CN114326721A (en) * | 2021-12-20 | 2022-04-12 | 达闼机器人有限公司 | Drawing establishing method and device, cloud server and robot |
| CN114710831A (en) * | 2022-03-10 | 2022-07-05 | 南京市地铁交通设施保护办公室 | RFID label positioning system based on deep learning |
| CN114710831B (en) * | 2022-03-10 | 2023-12-08 | 南京市地铁交通设施保护办公室 | RFID label positioning system based on deep learning |
| CN114916059A (en) * | 2022-04-29 | 2022-08-16 | 湖南大学 | WiFi fingerprint sparse map extension method based on interval random logarithm shadow model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113108792A (en) | Wi-Fi fingerprint map reconstruction method and device, terminal equipment and medium | |
| CN105933294B (en) | Network user's localization method, device and terminal | |
| CN110536245B (en) | Deep learning-based indoor wireless positioning method and system | |
| CN110232584B (en) | Parking lot site selection method and device, computer readable storage medium and terminal equipment | |
| CN113344291B (en) | Urban inland inundation range forecasting method, device, medium and equipment | |
| CN110139359B (en) | Interference source positioning processing method and device | |
| CN114205831B (en) | Method, device, storage medium and equipment for determining optimization scheme | |
| CN115561408A (en) | Air pollution early warning method and device, electronic equipment and storage medium | |
| WO2021103027A1 (en) | Base station positioning based on convolutional neural networks | |
| KR101694521B1 (en) | Apparatus and method for generating radio fingerprint map | |
| CN116415652A (en) | Data generation method and device, readable storage medium and terminal equipment | |
| CN115731560A (en) | Slot line identification method and device based on deep learning, storage medium and terminal | |
| CN112508938B (en) | Optical satellite image geometric quality evaluation method, device, equipment and storage medium | |
| CN113532424B (en) | Integrated equipment for acquiring multidimensional information and cooperative measurement method | |
| CN112561817B (en) | Remote sensing image cloud removing method, device, equipment and storage medium based on AM-GAN | |
| CN111767357B (en) | Regional mining complete evaluation method and equipment, electronic equipment and storage medium | |
| CN115942231A (en) | RSS-based 5G outdoor positioning method | |
| CN113141570B (en) | Underground scene positioning method, device, computing equipment and computer storage medium | |
| US8026914B2 (en) | Numerical analysis mesh generation apparatus, numerical analysis mesh generation method, and numerical analysis generation program | |
| CN113936106A (en) | Three-dimensional visualization method and system of monitoring map and related equipment | |
| CN113938895A (en) | Method and device for predicting railway wireless signal, electronic equipment and storage medium | |
| CN112188388A (en) | Hybrid indoor positioning method based on machine learning | |
| CN111563134A (en) | Fingerprint database clustering method, system, equipment and storage medium of positioning system | |
| CN111554089A (en) | Deep learning-based traffic state prediction method and device | |
| CN112218233B (en) | Position fingerprint generation method, position determination device and computer equipment |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Liu Ning Inventor after: Wu Di Inventor after: Niu Qun Inventor after: Jiang Weina Inventor before: Liu Ning Inventor before: Wu Di |
|
| CB03 | Change of inventor or designer information | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210713 |