CN111241881B - Method, apparatus, device and medium for region identification - Google Patents
Method, apparatus, device and medium for region identification Download PDFInfo
- Publication number
- CN111241881B CN111241881B CN201811444648.2A CN201811444648A CN111241881B CN 111241881 B CN111241881 B CN 111241881B CN 201811444648 A CN201811444648 A CN 201811444648A CN 111241881 B CN111241881 B CN 111241881B
- Authority
- CN
- China
- Prior art keywords
- region
- area
- picture
- target
- identification
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/255—Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
Abstract
The application discloses a method, a device, equipment and a medium for region identification. The method comprises the following steps: acquiring a satellite picture comprising known region information; preprocessing a satellite picture comprising known region information to obtain a target picture; taking a satellite picture comprising known region information as a training set, taking a target picture as target data, and training a region recognition model to obtain a trained region recognition model; and determining a target region in the satellite picture needing region identification by using the region identification model. According to the technical scheme provided by the embodiment of the invention, the region identification result obtained by the region identification model can be more accurate through the region identification model, and meanwhile, the region identification efficiency can be improved.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a method, an apparatus, a device, and a medium for region identification.
Background
The network quality evaluation is an important component of wireless network optimization work, and is an important means and basis for operators to develop network operation maintenance, optimization adjustment and engineering construction. When network quality evaluation is performed on a certain area, the specific position of the certain area needs to be determined, and the network quality can be accurately evaluated based on the determined area position.
Currently, the network quality of a certain area is usually evaluated and optimized according to data information in a Measurement Report (MR). However, the MR data cannot accurately represent the geographical location of the user, and therefore, the accuracy of the subsequent evaluation of the network quality of a certain area is affected.
Therefore, there is a technical problem that the region cannot be accurately identified.
Disclosure of Invention
The embodiment of the invention provides a method, a device, equipment and a medium for identifying a region, which can accurately identify the region.
In one aspect of the embodiments of the present invention, a method for identifying a region is provided, where the method includes:
acquiring a satellite picture comprising known area information;
preprocessing a satellite picture including known region information to obtain a target picture;
taking a satellite picture comprising known region information as a training set, taking a target picture as target data, and training a region recognition model to obtain a trained region recognition model;
and determining a target region in the satellite picture needing region identification by using the trained region identification model.
In another aspect of the embodiments of the present invention, an apparatus for identifying an area is provided, where the apparatus includes:
the image acquisition module is used for acquiring a satellite image comprising known region information;
the image processing module is used for preprocessing the satellite image comprising the known region information to obtain a target image;
the model creating module is used for training the region recognition model by taking the satellite picture comprising the known region information as a training set and taking the target picture as target data to obtain a trained region recognition model;
and the region identification module is used for determining a target region in the satellite picture needing region identification by using the trained region identification model.
According to another aspect of the embodiments of the present invention, there is provided an apparatus for area identification, the apparatus including:
a processor and a memory storing computer program instructions;
the processor, when executing the computer program instructions, implements the method of region identification as provided in any aspect of the embodiments of the present invention described above.
According to another aspect of embodiments of the present invention, there is provided a computer storage medium having computer program instructions stored thereon, the computer program instructions when executed by a processor implement the method of region identification as provided in any one of the aspects of embodiments of the present invention.
The embodiment of the invention provides a method, a device, equipment and a medium for identifying a region. Based on the satellite picture with known region information and the target picture obtained after preprocessing, a region recognition model is trained, the region recognition model can be more accurate by adjusting the hyper-parameters, and therefore the region recognition result obtained through the region recognition model is more accurate. Meanwhile, the trained region identification model is used for carrying out region identification on the satellite picture needing region identification, and the efficiency and the accuracy of the region identification are improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required to be used in the embodiments of the present invention will be briefly described below, and for those skilled in the art, other drawings may be obtained according to the drawings without creative efforts.
FIG. 1 is a schematic diagram of a method of region identification according to an embodiment of the invention;
FIG. 2 is a schematic diagram of a satellite picture according to an embodiment of the invention;
FIG. 3 illustrates a schematic diagram of a convolutional neural network in accordance with an embodiment of the present invention;
FIG. 4 shows a flow diagram of a method of region identification of another embodiment of the present invention;
FIG. 5 is a schematic diagram of a village area according to an embodiment of the present invention;
FIG. 6a is a schematic Manhattan distance diagram of an embodiment of the present invention;
FIG. 6b illustrates Manhattan distances between pixel points according to an embodiment of the invention;
FIG. 7 illustrates a flow diagram of a method of clustering in accordance with an embodiment of the invention;
FIG. 8 is a diagram illustrating a clustering result according to an embodiment of the present invention;
FIG. 9 illustrates a schematic diagram of the definition of an 8-pass detection contour tracking algorithm according to an embodiment of the present invention;
FIG. 10 is a diagram illustrating connected neighbor pixel points according to an embodiment of the present invention;
FIG. 11 shows a flow diagram of a method of region identification of yet another embodiment of the present invention;
FIG. 12 is a schematic diagram of a device for area identification according to an embodiment of the present invention;
FIG. 13 is a block diagram illustrating an exemplary hardware architecture of a computing device capable of implementing the method and apparatus for region identification according to embodiments of the present invention.
Detailed Description
Features and exemplary embodiments of various aspects of the present invention will be described in detail below, and in order to make objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not to be construed as limiting the invention. It will be apparent to one skilled in the art that the present invention may be practiced without some of these specific details. The following description of the embodiments is merely intended to provide a better understanding of the present invention by illustrating examples of the present invention.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising 8230; \8230;" comprises 8230; "does not exclude the presence of additional like elements in a process, method, article, or apparatus that comprises the element.
A method, an apparatus, a device, and a medium for region identification according to embodiments of the present invention are described in detail below with reference to the accompanying drawings. It should be noted that these examples are not intended to limit the scope of the present disclosure.
The method of area identification according to an embodiment of the present invention is described in detail below with reference to fig. 1 to 11.
In one embodiment of the present invention, as shown in fig. 1, fig. 1 is a schematic diagram illustrating a method of area identification according to an embodiment of the present invention. The area identification method in the embodiment of the present invention is described in detail below by taking a rural area as an example of an area to be identified, i.e., a target area.
As shown in fig. 1, a method 100 for identifying a region in an embodiment of the present invention includes the following steps:
and S110, acquiring a satellite picture comprising the known area information.
Specifically, as shown in fig. 2, fig. 2 is a schematic diagram illustrating a satellite picture in an embodiment of the present invention. Firstly, a preset number of satellite pictures including rural areas are obtained.
And S120, preprocessing the satellite picture including the known region information to obtain a target picture.
In one embodiment of the invention, reference is continued to FIG. 2. And marking out the rural area in the acquired satellite picture. For example, the pixel value of the rural area in the acquired satellite picture may be set to 0, that is, the rural area in the satellite picture is set to black, so as to obtain the target picture. The target picture comprises a rural area and a non-rural area of a black area.
And S130, taking the satellite picture comprising the known region information as a training set, taking the target picture as target data, and training the region identification model to obtain the trained region identification model.
In an embodiment of the present invention, the obtained satellite pictures are used as a training set of the region identification model. Wherein the training set is used to estimate the model. And taking the target picture as target data of the area identification model, namely taking the target picture as a label of the area identification model.
Firstly, dividing the satellite pictures including the known region information in the training set according to a preset specification. For example, a satellite picture including known region information in the training set and a target picture in the target data may be divided according to an 8 × 8 preset specification to obtain a model grid picture. It should be noted that the training set of region recognition models and the target data of the region recognition models together constitute the data set of the region recognition models.
Next, the target data, i.e., the labels of the region identification model, are converted into discrete labels, e.g., label [0,1] represents a non-rural region and label [1,0] represents a rural region.
And finally, performing model training on the model grid picture and the discrete label through a multilayer convolutional neural network model, and adjusting the hyper-parameters to enable the iterated model to be converged so as to obtain a trained region identification model. It should be understood that the hyper-parameter refers to a parameter set for training a model before starting model training, and during the model training, the hyper-parameter is continuously adjusted, so that the training result of the model is closer to the target data, and the performance of the training model is improved.
As shown in fig. 3, fig. 3 is a schematic diagram illustrating the principle of a convolutional neural network in the embodiment of the present invention.
The convolutional neural network is mainly used for simulating human recognition and cognition on objects. As can be seen in fig. 3, the features at the lowermost layer are substantially similar for different objects. The more upward, some features of such objects can be extracted, such as: wheels, eyes, and torso, etc. To the uppermost layer, different high-level features are finally combined into corresponding images, thereby enabling a human to accurately distinguish different objects.
The convolutional neural network is to identify the primary image features through the lower layer network first. Second, a further layer of features is composed of a number of underlying features. Through the combination of multiple levels, classification is finally made at the top level. The convolutional network successfully reduces the dimension of the image recognition problem with huge data volume continuously through a series of methods, and finally enables the image recognition problem to be trained.
In the embodiment of the invention, model training is carried out based on the convolutional neural network to obtain the trained area identification model, and the technical problem of area identification can be converted into the two classification problems based on the convolutional neural network. I.e. the result of the area identification is divided into a target area and a non-target area. The regions in the satellite pictures can be identified more accurately.
And S140, determining a target region in the satellite picture needing region identification by using the trained region identification model.
In an embodiment of the present invention, first, a satellite picture that needs to be subjected to region identification is divided according to a preset specification in a trained region identification model, for example, the satellite picture that needs to be subjected to region identification is divided according to an 8 × 8 preset specification, so as to obtain a grid picture to be identified.
Secondly, training the trained region identification model of the grid picture to be identified to obtain a training result. And obtaining a training result of the grid picture to be recognized as [0,1] or [1,0] based on the discrete label obtained by converting the target data of the trained region recognition model.
And when the training result of the grid picture to be recognized is [1,0], marking the grid picture to be recognized as a first identifier. For example, the grid picture to be recognized may be set to black. And when the training result of the grid picture to be recognized is [0,1], marking the grid picture to be recognized as a second identifier. For example, the grid picture to be recognized may be set to white.
And finally, splicing the grid picture to be recognized with the first identification and the grid picture to be recognized with the second identification together to be used as a result picture. It should be noted that, in order to prevent noise, the obtained training result may also be subjected to multiple times of training of the region identification model.
The resulting picture may then be scaled up to match the size of the original satellite picture. And obtaining the outline of the grid picture to be recognized with the first identifier in the expanded result picture through an Open Source Computer Vision Library (OpenCV), so as to determine a target area in the satellite picture which needs area recognition. Wherein, the image processing can be realized by OpenCV.
By the region identification method in the embodiment, the region identification model is trained by the convolutional neural network, and the technical problem of region identification can be converted into the two classification problems based on the convolutional neural network. I.e. the result of the area identification is divided into a target area and a non-target area. Training a trained region identification model of a satellite picture needing region identification and processing an OpenCV image can accurately identify the region needing region identification in the satellite picture.
For ease of understanding, fig. 4 shows a flowchart of a method of area identification according to another embodiment of the present invention. The steps in fig. 4 that are the same as in fig. 1 are given the same reference numerals.
As shown in fig. 4, the steps of the method 400 for identifying a region are the same as those of the method 100 for identifying a region shown in fig. 1, and are not described herein again. The method 400 of region identification in the embodiment of the present invention further includes the following steps:
s410, clustering the target area according to the Point of Interest (POI) points of the area, and determining the target area corresponding to the POI points of the area.
Specifically, based on the rural areas identified on the satellite map, the area of a particular village may also be specifically identified. It should be understood that in a geographic information system, a local POI may be a house, a business, a mailbox, a bus station, or the like.
In one embodiment of the present invention, the POI point is taken as the center point of the village. As shown in fig. 5, fig. 5 is a schematic view showing village areas in the embodiment of the present invention. And marking the POI points as black, marking rural areas identified in the satellite picture as gray, and marking non-rural areas in the satellite picture as white. When the area where a specific village is located needs to be specifically identified, it needs to be noted which gray areas belong to the current POI point.
As shown in fig. 6a, fig. 6a is a schematic diagram illustrating manhattan distance in an embodiment of the invention. The three paths a, b and c in FIG. 6a, all represent Manhattan paths, and their lengths are Manhattan distance values. In a plane, e.g. coordinate (x) i ,y i ) And the coordinate (x) j ,y j ) The manhattan distance of (c) can be calculated according to expression (1).
D(i,j)＝|x i -x j |+|y i -y j | (1)
Wherein D (i, j) represents the coordinate (x) i ,y i ) And the coordinate (x) j ,y j ) Manhattan distance of.
As shown in fig. 6b, fig. 6b illustrates manhattan distances between pixels according to an embodiment of the present invention. For example, a point having a manhattan distance of 1 from point a is B, and a point having a manhattan distance of 2 from point a is C.
As a specific example, as shown in fig. 7, fig. 7 is a schematic diagram illustrating a clustering process flow in the embodiment of the present invention. When clustering a target region, i.e., a rural region, according to a POI point, a maximum distance value D of manhattan distances may be set first max Setting the initial value of the Manhattan distance Dist to be 1, and setting all POI points P i Corresponding contour boundaryIs Edge i ＝{(x i ,y i ) }, setting all POI points P i (x i ,y i ) I =1,2, \8230, the set of regions corresponding to Np is area = { (x) i ,y i ) Wherein i is a positive integer.
Next, the initial value of the initialization boundary change number NUM is 0, i is 1. And confirming a pixel point with the Manhattan distance Dist between the pixel point and the current Pi in the set of the contour boundary Edgei = { (xi, yi) } corresponding to the current Pi. When the pixel point (x) is obtained n ,y n ) Has a pixel value of 0.5, i.e. the pixel point is a gray point, (x) will be n ,y n ) Joining Edge i And Area i And (x) is n ,y n ) Is set to 0, i.e. a pixel point (x) n ,y n ) Set to black dots. It should be understood that n is a positive integer.
When Edge i When a change occurs, NUM = NUM +1 is set, and Edge is paired i And (4) screening the pixel points in the set, and if the periphery of one pixel point is surrounded by the black point, screening the Edge from the outline boundary i And (5) removing.
Traversing Edge in sequence according to the method i Each pixel point in the set until Dist = D is satisfied max 。
As shown in fig. 8, fig. 8 is a schematic diagram illustrating a clustering result in an embodiment of the present invention. As can be seen, starting from the POI point, the manhattan distance is 1 and continuously extends outward until the connected domain, i.e. the region not blocked by the white region, is completely allocated as shown in fig. 8 d. Then increasing the Manhattan distance until crossing the white area to another unallocated area as shown in FIG. 8e, then repeating the Manhattan distance of 1 and extending outward, and finally when the Manhattan distance reaches the predetermined maximum threshold D max And if no new unallocated area is found, the whole clustering process is completed, and the obtained whole black area is the target area corresponding to the POI point.
In the embodiment of the invention, as for the distances between different pixel points, the Manhattan distance is adopted, and only the addition and subtraction method needs to be calculated, so that the operation efficiency is greatly improved, and no error is ensured. The regions belonging to a particular village can be accurately identified by the clustering process. So that the identification of the region is more accurate.
And S420, obtaining the area outline of the target area corresponding to the POI point through an outline tracking algorithm.
Specifically, the contour tracking algorithm may adopt an 8-connectivity detection contour tracking algorithm. In an embodiment of the invention, a binary image is input, i.e. the pixel values or grey values are only 0 and 1. Where 0 represents black and 1 represents white.
As shown in fig. 9, fig. 9 is a definition diagram illustrating an 8-connectivity detection contour tracking algorithm in one embodiment of the invention.
The frame is a rectangular frame surrounded by a layer of pixel points at the outermost periphery of the picture, namely two columns of the uppermost and the lowermost rows, and two columns of the leftmost and the rightmost rows of the picture.
The background S1 is a pixel point at the same level as the frame and having a pixel value of 0.
Connected blocks S2 and S5 are a block of regions made up of connected white point pixels.
The hole S3 is a region composed of connected black dots except for the background.
The outer outlines B1, B3 and B4 refer to the outermost white points wrapping a communicating block.
The inner contour B2 is a white point at the outermost periphery of a hole or a white point at the innermost periphery of a communicating block.
As shown in fig. 10, fig. 10 is a schematic diagram illustrating connected neighbor pixel points in an embodiment of the present invention. For the target pixel point A, the condition of 4 communication means that only the point B directly connected in the four directions of up, down, left and right calculate the neighbor pixel point. And the points B and C which are communicated by the point 8 and added into four corners of the upper left corner, the lower left corner, the upper right corner and the lower right corner calculate the neighbor pixel points.
In one embodiment of the present invention, the input image is denoted as F = { F (i, j) }, the sequence number of the current boundary is denoted by NBD, the NBD is initialized to 1, the frame of F is denoted, and the rest of the boundaries are labeled starting from 2. Traversing each pixel point in the picture according to the sequence from top to bottom (i is increased) and from left to right (j is increased), resetting the variable LNBD to 1 each time a new row is scanned, and executing the following steps when f (i, j) is not equal to 0:
step1, if f (i, j) =1 and f (i, j-1) =0, then the pixel (i, j) is considered as a starting point of the outer contour, let NBD = NBD +1 and (i, j-1) 2 ,j 2 ) And (i, j-1). If f (i, j) ≧ 1 and f (i, j + 1) =0, then pixel (i, j) is considered as a starting point of the inner contour, let NBD = NBD +1 and (i, j) be the starting point of the inner contour 2 ,j 2 ) = (i, j + 1), if f (i, j)>1 LNBD = f (i, j). Otherwise, jump to Step4.
And Step2, determining the parent node of the current contour according to the newly found contour with the type and sequence number of the LNBD contour, namely the contour meeting in the previous line. The newly found decision rule of the parent contour of contour B is shown in table 1.
TABLE 1
Step3, the detected boundary is traced from the start point (i, j).
In an embodiment of the present invention, step3 specifically includes the following four steps:
first step of starting from (i) 2 ,j 2 ) Starting, checking the neighbor pixel points of (i, j) in a clockwise direction, finding a non-0 pixel point, and recording the first non-0 pixel point as (i) 1 ,j 1 ). If not found, -NBD is assigned to f (i, j) and Step4 is skipped.
A second step of reacting (i) 2 ,j 2 )＝(i 1 ,j 1 )，(i 3 ,j 3 )＝(i,j)。
A third step of starting from (i) 2 ,j 2 ) Starting with the next pixel in the counter-clockwise direction, the current pixel is detected in the counter-clockwise direction (i) 3 ,j 3 ) Finding the first non-0 pixel as (i) 4 ,j 4 )。
A fourth step of, if (i) 3 ,j 3 + 1) is a 0 pixel, thenLet f (i) 3 ,j 3 ) = -NBD. If f (i) 3 ,j 3 ) =1 and (i) 3 ,j 3 + 1) is a non-0 pixel, then let f (i) 3 ,j 3 ) = NBD. Otherwise, f (i) 3 ,j 3 ) No change occurs.
A fifth step of, if (i) 4 ,j 4 ) And (i) = (i, j) and (i) 3 ,j 3 )＝(i 1 ,j 1 ) I.e. back to the starting point, step4 is jumped to. Otherwise, let (i) 2 ,j 2 )＝(i 3 ,j 3 )，(i 3 ,j 3 )＝(i 4 ,j 4 ) And returning to the third step.
Step4, if f (i, j) ≠ 1, let LNBD = | f (i, j) |. And continuing to restore traversal from the pixel point (i, j + 1), and when the image is traversed to the lower right corner, terminating the 8-connected detection contour tracking algorithm.
In the embodiment of the invention, the boundary is tracked by finding out the edge points according to the set sequence through the 8-communication detection contour tracking algorithm, so that the boundary points of the region can be accurately identified, and the contour of the region is further obtained.
In another embodiment of the invention, when the area outline of the target village needs to be specifically identified, the image with the target area needs to be spliced and screened because one satellite image may not cover one village area.
Firstly, finding out a contour picture of a target village point, taking the contour picture as a center, finding out 8 surrounding pictures for splicing, and obtaining a spliced picture.
Secondly, since there may be a plurality of contours in the stitched picture, the contour closest to the target village is first found, and if there is only one village point in the contour region, the contour is the region contour of the target village. If there are multiple village points in the contour area, i.e., there may be connected villages, then distance determination is made for the contour. The village point closest to the outline area is the village area to which the outline area belongs.
And S430, performing network quality evaluation on the target area corresponding to the area POI in the area outline of the target area corresponding to the area POI.
In an embodiment of the present invention, the network quality evaluation is performed by using OTT sampling points in Measurement Report (MR) data of a Global System for mobile Communication (GSM) network within the area outline. The MR measurement period of a Long Term Evolution (LTE) network is implemented by an Evolved Node B (eNodeB) or a User Equipment (UE) as required, the measurement period may be 120 milliseconds (ms), 240ms, 640ms, 1024ms, 2048ms, 5120ms, or 10240ms, and the measured data may be used for network evaluation and optimization.
In one embodiment of the invention, network coverage may be evaluated. Specifically, the weak coverage area may be counted according to the average field strength of each contour area or according to the proportion of the weak coverage sampling points, so as to evaluate the network coverage. The average field intensity threshold of each contour region may be set to-105 decibels (dB), the proportion of the weak coverage sampling point may be set to be greater than 10%, and a region smaller than the average field intensity threshold may be referred to as a weak coverage region.
In one embodiment of the present invention, the communication Quality condition of each region can be evaluated by the average uplink Signal-to-Noise Ratio (Signal to Interference plus Noise Ratio UL, sinrll) and the Reference Signal Receiving Quality (Reference Signal Receiving Quality, RSRQ) of each profile region. In another embodiment of the present invention, the quality difference region may also be counted according to the average sinrll and the average RSRQ of the contour region or according to the weak quality sampling point ratio.
In an embodiment of the present invention, the switching success rate of each region may also be counted by counting the switching data of the XDR format files of the Uu interface and the X2 interface. Here, handover success rate (%) = (number of successful handovers × 100)/number of handover requests. And counting the area with the switching failure rate exceeding a specified threshold, and finding a problem outline area with the switching failure.
In an embodiment of the present invention, the disconnection event refers to an occurrence of an abnormal Radio Resource Control (RRC) connection release on the Uu interface.
And finding out a dropped-line problem area by counting the contour area with the dropped-line rate exceeding a specified threshold.
In one embodiment of the present invention, the sampling point overlapping coverage refers to the number of cells (including the main cell) whose field intensity difference with the main cell is within 6dB of the absolute value. The statistical range comprises pilot frequency adjacent cells and adjacent cells without pilot frequency.
The area average overlapping coverage refers to calculating an average value of the overlapping coverage of the sampling points in the area, and represents the average overlapping coverage of the area.
And evaluating the overlapping coverage condition of an area by averaging the overlapping coverage and the high overlapping coverage proportion, and finding out the overlapping coverage problem area by counting that the high overlapping coverage proportion exceeds a specified threshold.
As shown in fig. 11, fig. 11 is a flowchart illustrating a method of area identification according to still another embodiment of the present invention. The method 1100 of region identification comprises:
and S1110, processing sample data of the satellite picture in the rural residential area.
And S1120, modeling a rural residential area model.
And S1130, marking the rural residential area as a whole.
And S1140, clustering residential areas corresponding to the rural POI information points.
S1150, acquiring the residential area contour corresponding to the rural POI information point.
And S1160, evaluating the quality of the wireless network in the rural residential area.
Hereinafter, a region identifying apparatus according to an embodiment of the present invention, which corresponds to the region identifying method, will be described in detail with reference to fig. 12.
Fig. 12 is a schematic structural diagram of an apparatus for area identification according to an embodiment of the present invention.
As shown in fig. 12, the apparatus 1200 for area identification includes:
a picture acquiring module 1210 for acquiring a satellite picture including information of a known area.
The picture processing module 1220 is configured to pre-process a satellite picture including known region information to obtain a target picture.
A model creating module 1230, configured to train the region identification model with the satellite pictures including the known region information as a training set and the target picture as target data.
And the region identification module 1240 is used for determining a target region in the satellite picture which needs region identification by using the trained region identification model.
By the area recognition device in the embodiment of the present invention, based on the image acquisition module 1210, the image processing module 1220, and the model creation module 1230, the area recognition model is trained, the area recognition model may be trained based on a satellite image with known area information and a target image obtained after preprocessing, and the area recognition model may be more accurate by adjusting the hyper-parameter, so that the area recognition result obtained by the area recognition model is more accurate. Through the area identification module 1240, the trained area identification model is used for carrying out area identification on the satellite picture needing area identification, so that the efficiency and the accuracy of the area identification are improved.
In an embodiment of the present invention, the image processing module 1220 is specifically configured to label a target region in a satellite image including known region information, and use the satellite image including the labeled target region as the target image.
In an embodiment of the present invention, the model creating module 1230 is specifically configured to divide a satellite picture including known area information in a training set and a target picture in target data based on a preset specification to obtain a model grid picture, and train an area identification model by using a convolutional neural network and the model grid picture.
In an embodiment of the present invention, the area identification module 1240 is specifically configured to divide a satellite picture that needs to be subjected to area identification according to a preset specification, so as to obtain a grid picture to be identified.
And performing model training on the to-be-recognized grid picture based on the trained region recognition model to obtain a training result.
And marking the grid picture to be recognized with the training result as the target area as a first mark, and marking the grid picture to be recognized with the training result as the non-target area as a second mark.
Splicing the grid picture to be recognized with the first identification and the grid picture to be recognized with the second identification to obtain a result picture, acquiring the outline of the grid picture to be recognized with the first identification in the enlarged result picture, and determining a target area in the satellite picture to be subjected to area recognition.
In another embodiment of the present invention, the apparatus 1200 for area identification further comprises:
the clustering module 1250 is configured to perform clustering on the target area according to the POI points in the area, and determine a target area corresponding to the POI points in the area.
The contour tracking module 1260 is configured to obtain an area contour of the target area corresponding to the area POI point by using a contour tracking algorithm.
A quality evaluation module 1270, configured to perform network quality evaluation on the target area corresponding to the area POI point within the area contour of the target area corresponding to the area POI point.
In an embodiment of the present invention, the cluster processing module 1250 is specifically configured to set an initial value of the manhattan distance and set a maximum value of the manhattan distance.
And increasing the initial value of the Manhattan distance according to the preset increment to obtain the current Manhattan distance value.
And determining a region boundary point with the distance between the region POI point and the target region as the current Manhattan distance value by taking the region POI point as a starting point.
And when the current Manhattan distance value is the maximum value of the Manhattan distance, taking the area formed by the obtained area boundary points as the target area corresponding to the area POI point.
In one embodiment of the present invention, the contour tracing module 1260 is specifically configured for contour tracing using an 8-connectivity detection contour tracing algorithm.
In an embodiment of the present invention, the quality evaluation module 1270 is specifically configured to perform a quality evaluation by using OTT sampling points in the measurement report MR data within the area profile, and the quality evaluation includes at least one of the following: network coverage assessment, network quality assessment, handover assessment, dropped call assessment, and repeated coverage assessment.
FIG. 13 is a block diagram illustrating an exemplary hardware architecture of a computing device capable of implementing the method and apparatus for region identification according to embodiments of the present invention.
As shown in fig. 13, computing device 1300 includes an input device 1301, an input interface 1302, a central processor 1303, a memory 1304, an output interface 1305, and an output device 1306. The input interface 1302, the central processor 1303, the memory 1304, and the output interface 1305 are connected to each other through a bus 1310, and the input device 1301 and the output device 1306 are connected to the bus 1310 through the input interface 1302 and the output interface 1305, respectively, and further connected to other components of the computing device 1300.
Specifically, the input device 1301 receives input information from the outside, and transmits the input information to the central processor 1303 through the input interface 1302; the central processor 1303 processes input information based on computer-executable instructions stored in the memory 1304 to generate output information, stores the output information in the memory 1304 temporarily or permanently, and then transmits the output information to the output device 1306 through the output interface 1305; output device 1306 outputs output information to the exterior of computing device 1300 for use by a user.
That is, the computing device shown in fig. 13 may also be implemented as a device for region identification, which may include: a memory storing computer executable instructions; and a processor which, when executing computer executable instructions, may implement the method and apparatus for region identification described in connection with fig. 1-12.
An embodiment of the present invention further provides a computer-readable storage medium, where the computer-readable storage medium stores computer program instructions; the computer program instructions, when executed by a processor, implement region identification provided by embodiments of the present invention.
It is to be understood that the invention is not limited to the specific arrangements and instrumentality described above and shown in the drawings. A detailed description of known methods is omitted herein for the sake of brevity. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present invention are not limited to the specific steps described and illustrated, and those skilled in the art can make various changes, modifications and additions or change the order between the steps after comprehending the spirit of the present invention. The functional blocks shown in the above-described structural block diagrams may be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, it may be, for example, an electronic circuit, an Application Specific Integrated Circuit (ASIC), suitable firmware, plug-in, function card, or the like. When implemented in software, the elements of the invention are the programs or code segments used to perform the required tasks. The program or code segments can be stored in a machine-readable medium or transmitted by a data signal carried in a carrier wave over a transmission medium or a communication link. A "machine-readable medium" may include any medium that can store or transfer information. Examples of a machine-readable medium include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio Frequency (RF) links, and so forth. The code segments may be downloaded via computer networks such as the internet, intranets, etc.
The present invention may be embodied in other specific forms without departing from its spirit or essential characteristics. For example, the algorithms described in the specific embodiments may be modified without departing from the basic spirit of the invention. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.
Claims (10)
1. A method of region identification, comprising:
acquiring a satellite picture comprising known area information;
preprocessing the satellite picture including the known region information to obtain a target picture;
taking the satellite picture comprising the known region information as a training set, taking the target picture as target data, and training a region recognition model to obtain a trained region recognition model;
determining a target region in a satellite picture needing region identification by using the trained region identification model;
clustering the target area according to the POI points of interest of the area, and determining the target area corresponding to the POI points of the area;
obtaining the area contour of the target area corresponding to the area POI points by utilizing a contour tracking algorithm;
and performing network quality evaluation on the target area corresponding to the area POI point in the area outline of the target area corresponding to the area POI point.
2. The method for identifying the region according to claim 1, wherein the preprocessing the satellite picture including the known region information to obtain a target picture comprises:
labeling the target region in the satellite picture comprising the known region information;
and taking the satellite picture comprising the marked target area as a target picture.
3. The method of claim 1, wherein the training a region recognition model with the satellite picture including known region information as a training set and the target picture as target data to obtain a trained region recognition model comprises:
dividing satellite pictures including known region information in the training set and target pictures in the target data based on a preset specification to obtain model grid pictures;
and training the region recognition model by using the convolutional neural network and the model grid picture to obtain the trained region recognition model.
4. The method of claim 3, wherein the determining, by using the trained region recognition model, a target region in a satellite picture that needs region recognition comprises:
dividing the satellite picture needing region identification according to the preset specification to obtain a grid picture to be identified;
performing model training on the grid picture to be recognized based on the trained region recognition model to obtain a training result;
marking the grid picture to be recognized with the training result as a target area as a first mark, and marking the grid picture to be recognized with the training result as a non-target area as a second mark;
splicing the grid picture to be recognized with the first identification and the grid picture to be recognized with the second identification to obtain a result picture, acquiring the outline of the grid picture to be recognized with the first identification in the enlarged result picture, and determining a target area in the satellite picture to be subjected to area recognition.
5. The method for identifying the area according to claim 1, wherein the clustering the target area according to the POI points of interest of the area to determine the target area corresponding to the POI points of the area comprises:
setting an initial value of a manhattan distance and setting a maximum value of the manhattan distance;
increasing the initial value of the Manhattan distance according to a preset increment to obtain a current Manhattan distance value;
determining a region boundary point, which takes the distance between the region POI point and the target region as a current Manhattan distance value, by taking the region POI point as a starting point;
and when the current Manhattan distance value is the maximum value of the Manhattan distance, taking the area formed by the obtained area boundary points as a target area corresponding to the area POI point.
6. The method of region identification according to claim 1, wherein the contour tracking algorithm comprises: 8-connectivity detection contour tracking algorithm.
7. The method for area identification according to claim 1, wherein the quality evaluation based on the area contour of the target area corresponding to the area POI point comprises:
using the OTT sampling points in the measurement report MR data in the area outline to perform quality evaluation, wherein the quality evaluation at least comprises one of the following steps: network coverage assessment, network quality assessment, handover assessment, dropped call assessment, and repeated coverage assessment.
8. An apparatus for region identification, comprising:
the image acquisition module is used for acquiring a satellite image comprising known area information;
the image processing module is used for preprocessing the satellite image comprising the known region information to obtain a target image;
the model creating module is used for taking the satellite picture comprising the known region information as a training set, taking the target picture as target data, training a region recognition model and obtaining the trained region recognition model;
the area identification module is used for determining a target area in the satellite picture needing area identification by using the trained area identification model;
the clustering processing module is used for clustering the target area according to the POI points of the area interest points and determining the target area corresponding to the POI points of the area;
the contour tracking module is used for acquiring the region contour of the target region corresponding to the region POI point by using a contour tracking algorithm;
and the quality evaluation module is used for carrying out network quality evaluation on the target area corresponding to the area POI in the area outline of the target area corresponding to the area POI.
9. An apparatus for region identification, the apparatus comprising: a processor and a memory storing computer program instructions;
the processor, when executing the computer program instructions, implements a method of region identification as claimed in any of claims 1-7.
10. A computer storage medium having computer program instructions stored thereon, which when executed by a processor, implement a method of region identification as claimed in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811444648.2A CN111241881B (en) | 2018-11-29 | 2018-11-29 | Method, apparatus, device and medium for region identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811444648.2A CN111241881B (en) | 2018-11-29 | 2018-11-29 | Method, apparatus, device and medium for region identification |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111241881A CN111241881A (en) | 2020-06-05 |
| CN111241881B true CN111241881B (en) | 2022-12-27 |
Family
ID=70865512
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811444648.2A Active CN111241881B (en) | 2018-11-29 | 2018-11-29 | Method, apparatus, device and medium for region identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111241881B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111818557B (en) * | 2020-08-04 | 2023-02-28 | 中国联合网络通信集团有限公司 | Network coverage problem identification method, device and system |
| CN113192229B (en) * | 2021-05-12 | 2022-05-10 | 西安图迹信息科技有限公司 | Power plant inspection method based on wireless Bluetooth equipment |
| CN113723405A (en) * | 2021-08-31 | 2021-11-30 | 北京百度网讯科技有限公司 | Method and device for determining area outline and electronic equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6850497B1 (en) * | 1995-09-19 | 2005-02-01 | Mobile Satellite Ventures, Lp | Satellite trunked radio service system |
| CN105095838A (en) * | 2014-05-20 | 2015-11-25 | 中国移动通信集团广东有限公司 | Target detection method and device |
| CN107067003A (en) * | 2017-03-09 | 2017-08-18 | 百度在线网络技术（北京）有限公司 | Extracting method, device, equipment and the computer-readable storage medium of region of interest border |
| CN107527109A (en) * | 2016-06-21 | 2017-12-29 | 中国辐射防护研究院 | Dose of radiation caused by gaseous state radionuclide and the evaluation method of safeguard procedures suggestion |
| CN108052876A (en) * | 2017-11-28 | 2018-05-18 | 广东数相智能科技有限公司 | Regional development appraisal procedure and device based on image identification |
| CN108124279A (en) * | 2017-12-12 | 2018-06-05 | 中国联合网络通信集团有限公司 | The appraisal procedure and device of network coverage quality |
| WO2018209057A1 (en) * | 2017-05-11 | 2018-11-15 | The Research Foundation For The State University Of New York | System and method associated with predicting segmentation quality of objects in analysis of copious image data |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104378635B (en) * | 2014-10-28 | 2017-12-05 | 西交利物浦大学 | The coding method of video interested region based on microphone array auxiliary |
| WO2016154320A1 (en) * | 2015-03-24 | 2016-09-29 | Carrier Corporation | System and method for determining rf sensor performance relative to a floor plan |
| CN108399454A (en) * | 2018-03-05 | 2018-08-14 | 山东领能电子科技有限公司 | A kind of completely new sectional convolution neural network target recognition |
-
2018
- 2018-11-29 CN CN201811444648.2A patent/CN111241881B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6850497B1 (en) * | 1995-09-19 | 2005-02-01 | Mobile Satellite Ventures, Lp | Satellite trunked radio service system |
| CN105095838A (en) * | 2014-05-20 | 2015-11-25 | 中国移动通信集团广东有限公司 | Target detection method and device |
| CN107527109A (en) * | 2016-06-21 | 2017-12-29 | 中国辐射防护研究院 | Dose of radiation caused by gaseous state radionuclide and the evaluation method of safeguard procedures suggestion |
| CN107067003A (en) * | 2017-03-09 | 2017-08-18 | 百度在线网络技术（北京）有限公司 | Extracting method, device, equipment and the computer-readable storage medium of region of interest border |
| WO2018209057A1 (en) * | 2017-05-11 | 2018-11-15 | The Research Foundation For The State University Of New York | System and method associated with predicting segmentation quality of objects in analysis of copious image data |
| CN108052876A (en) * | 2017-11-28 | 2018-05-18 | 广东数相智能科技有限公司 | Regional development appraisal procedure and device based on image identification |
| CN108124279A (en) * | 2017-12-12 | 2018-06-05 | 中国联合网络通信集团有限公司 | The appraisal procedure and device of network coverage quality |
Non-Patent Citations (2)
| Title |
|---|
| An automatic method for road extraction in rural and semi-urban areas starting from high resolution satellite imagery;J.B.Mena .etc;《Pattern Recognition Letters》;20050731;第26卷(第9期);1201-1220 * |
| 基于多数据源关联的市网分析产品终端战略地图;黄莎;《电信工程技术与标准化》;20160930(第9期);1-6 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111241881A (en) | 2020-06-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111241881B (en) | Method, apparatus, device and medium for region identification | |
| CN110557716B (en) | Indoor positioning method based on lognormal model | |
| CN106454747B (en) | Wireless positioning method of mobile phone terminal | |
| CN112887897B (en) | Terminal positioning method, device and computer readable storage medium | |
| CN110798804B (en) | Indoor positioning method and device | |
| CN113591967A (en) | Image processing method, device and equipment and computer storage medium | |
| CN110390261A (en) | Object detection method, device, computer readable storage medium and electronic equipment | |
| CN105335952A (en) | Matching cost calculation method and apparatus, and parallax value calculation method and equipment | |
| CN114722944A (en) | Point cloud precision determination method, electronic device and computer storage medium | |
| CN114828026A (en) | Base station planning method, device, equipment, storage medium and program product | |
| Mihara et al. | Positioning for user equipment of a mmWave system using RSSI and stereo camera images | |
| CN108401222B (en) | Positioning method and device | |
| CN114339603A (en) | Switching method, device and equipment of positioning area and storage medium | |
| CN112543411B (en) | Interference positioning method, device and system of wireless communication system | |
| US20230036577A1 (en) | Swapped Section Detection and Azimuth Prediction | |
| CN114611635B (en) | Object identification method and device, storage medium and electronic device | |
| Ji et al. | Accurate Long‐Term Evolution/Wi‐Fi hybrid positioning technology for emergency rescue | |
| CN115830342A (en) | Method and device for determining detection frame, storage medium and electronic device | |
| CN111356152B (en) | Base station position correction method, device, equipment and medium | |
| CN113034613B (en) | External parameter calibration method and related device for camera | |
| CN111210471B (en) | Positioning method, device and system | |
| CN114782496A (en) | Object tracking method and device, storage medium and electronic device | |
| CN113141570B (en) | Underground scene positioning method, device, computing equipment and computer storage medium | |
| CN111372309B (en) | Positioning method and device based on LTE signal and readable storage medium | |
| CN115731458A (en) | Processing method and device for remote sensing image and electronic 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |