Street lamp nameplate recognition positioning method and system

Abstract

This application relates to a method and system for identifying and locating streetlight nameplates. By receiving images of streetlight nameplates and extracting the information attached to those images, the system obtains streetlight information and latitude / longitude coordinates. This simplifies the process of recording and locating streetlight information by simply taking a photo, significantly reducing the workload of data collection personnel and improving efficiency. During the extraction of information from streetlight nameplate images, to prevent blurry images, the system prioritizes checking for blurriness. If the image is blurry, the system sharpens it, increasing the utilization rate and efficiency of the captured images. Furthermore, this method has high applicability; the system can extract text from the information attached to streetlight nameplate images according to predetermined text inclusion rules.

Description

Technical Field

[0001] This application relates to the field of intelligent street light technology, and in particular to a method and system for identifying and locating street light nameplates. Background Technology

[0002] With the development of IoT technology, streetlights, as electronic terminals, need to be connected to the IoT network for unified management. However, due to the large number of streetlights, the workload for data entry and location tracking is substantial. Traditional methods involve manually copying the streetlight nameplates and simultaneously recording the streetlight's location information, then uploading the collected information to the management system. This workflow is cumbersome and inefficient due to the large workload. Therefore, it is necessary to propose a streetlight nameplate identification and location method and system to address the inefficiency of traditional streetlight management methods. Summary of the Invention

[0003] Therefore, it is necessary to propose a street light nameplate identification and positioning method and system to address the problem of low efficiency in traditional street light management methods.

[0004] This application provides a method for identifying and locating street light nameplates, including:

[0005] Receive images of street light nameplates and determine if the images are blurry;

[0006] If the image of the street light sign is blurry, the image blurring algorithm is used to sharpen the image of the street light sign to form the first image, and it is then determined whether the first image is blurry.

[0007] If the first image is blurry, a prompt message is sent, and the image of the street light sign is returned to determine whether the image of the street light sign is blurry.

[0008] If the first image is not blurry, then extract the text information from the image of the street lamp nameplate;

[0009] If the image of the street light sign is clear, then extract the text information from the image of the street light sign; during the process of extracting the text information from the image of the street light sign, preserve the text arrangement format;

[0010] Receive the pre-defined text inclusion rules, and extract the street light information contained in the text information of the street light nameplate image according to the pre-defined text inclusion rules;

[0011] Analyze the image of the street light sign to obtain its latitude and longitude information;

[0012] The system uploads the street light information and latitude and longitude information of the street light nameplate image to the IoT management system, and returns the received street light nameplate image to determine if the image is blurry, until no more street light nameplate images are received.

[0013] This application also provides a street light nameplate identification and positioning system, including:

[0014] A mobile terminal is used to execute a street light sign identification and positioning method; the mobile terminal includes a camera module, a positioning module, and a storage module.

[0015] The Internet of Things (IoT) management system communicates with the mobile terminal.

[0016] This application relates to a method and system for identifying and locating streetlight nameplates. By receiving images of streetlight nameplates and extracting the information attached to those images, the system obtains streetlight information and latitude / longitude coordinates. This simplifies the process of recording and locating streetlight information by simply taking a photo, significantly reducing the workload of data collection personnel and improving efficiency. During the extraction of information from the streetlight nameplate images, to prevent blurry images, the mobile terminal prioritizes determining if the image is blurry. If the image is blurry, the mobile terminal performs image sharpening, increasing the utilization rate and efficiency of the captured images. Furthermore, this method has high applicability; the mobile terminal can extract text from the information attached to the streetlight nameplate images according to predetermined text inclusion rules. Attached Figure Description

[0017] The accompanying drawings, which form part of this application, are used to provide a further understanding of the application and to make other features, objects, and advantages of the application more apparent. The illustrative embodiments and descriptions of this application are used to explain the application and do not constitute an undue limitation of the application.

[0018] Figure 1 This is a flowchart of a method for identifying and locating street light nameplates, provided as an embodiment of this application.

[0019] Figure 2 This is a module connection diagram of a street light nameplate identification and positioning system provided in one embodiment of this application.

[0020] Figure label:

[0021] 100 - Mobile terminal; 110 - Camera module; 120 - Positioning module; 130 - Storage module;

[0022] 200-Internet of Things Management System. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0024] This application provides a method for identifying and locating street light nameplates.

[0025] like Figure 1 As shown, in one embodiment of this application, a street light nameplate identification and positioning method includes:

[0026] S100 receives images of street light nameplates and determines whether the images are blurry.

[0027] S200: If the image of the street light sign is blurry, the image blurring algorithm is used to sharpen the image of the street light sign to form a first image, and it is determined whether the first image is blurry.

[0028] S300, if the first image is blurry, a prompt message is sent, and the image of the received street light nameplate is returned to determine whether the image of the street light nameplate is blurry.

[0029] Specifically, in actual operation, before receiving the image of the streetlight sign, the mobile terminal will give the operator the option to send the information in real time. The purpose of this option is that if the operator takes and uploads the image in real time at the location of the streetlight, and the image of the streetlight sign is blurry, the mobile terminal will directly provide a "blurry photo" message and return to the photo-taking step to retake the photo. If the operator only takes photos of the streetlight sign information, and all streetlight sign photos are taken before the information is processed and uploaded, then if the streetlight sign images are blurry, the only solution is to use an image blurring algorithm to sharpen the images.

[0030] S400, if the first image is not blurry, then extract the text information of the image of the street light sign.

[0031] S500, if the image of the street light sign is clear, then extract the text information of the image of the street light sign, and preserve the text arrangement format during the extraction of the text information of the image of the street light sign.

[0032] S600 receives a pre-defined text inclusion rule and extracts the street light information contained in the text information of the street light nameplate image according to the pre-defined text inclusion rule.

[0033] S700 analyzes images of street light nameplates to obtain their latitude and longitude information.

[0034] The S800 uploads the street light information and latitude and longitude information of the street light nameplate image to the IoT management mobile terminal, and returns the received street light nameplate image to determine whether the street light nameplate image is blurry, until no more street light nameplate images are received.

[0035] Specifically, obtaining street light information and obtaining latitude and longitude information are two relatively independent steps, and their execution processes can be sequential or executed in parallel.

[0036] In fact, the underlying logic for giving operators the option to "send information in real time" is as follows:

[0037] Receive instructions from the operator regarding whether to send information in real time.

[0038] If the instruction is to send information in real time, before processing the image of the street light sign, the system directly returns the received image of the street light sign, determines whether the image of the street light sign is blurry, and interrupts the process of processing the image of the street light sign using the image blurring algorithm to form the first image, and then determines whether the first image is blurry.

[0039] If the instruction is not to send information in real time, then the step of receiving the image of the street light nameplate and determining whether the image of the street light nameplate is blurry is executed.

[0040] The core of these steps is to determine whether the operator is sending information in real time.

[0041] This embodiment relates to a street light sign identification and positioning method and a mobile terminal. By receiving images of street light signs and extracting the information attached to the images, street light information and latitude and longitude information can be obtained. This simplifies the street light information entry and positioning work, requiring only taking photos, greatly reducing the workload of personnel collecting information and improving work efficiency. During the extraction of information attached to the street light sign images, to prevent blurry images, the mobile terminal prioritizes determining whether the image is blurry. If the image is blurry, the mobile terminal will sharpen it, improving the utilization rate of the captured street light sign images and increasing efficiency. Furthermore, this method has high applicability; the mobile terminal can extract text from the information attached to the street light sign images according to predetermined text inclusion rules.

[0042] In one embodiment of this application, S100 includes:

[0043] S111, Receive a sample image defined as a clear street light sign.

[0044] S112, use the DB algorithm to binarize all pixels of the selected street lamp nameplate sample image to form a sample binarized image of the street lamp nameplate sample image.

[0045] S113, Calculate the proportion of binary distribution in the binary image of the sample, and generate a threshold standard for the binary distribution proportion based on the proportion of binary distribution in the sample binary image. The threshold standard for the binary distribution proportion is the percentage of the number of pixels judged as black out of the total number of pixels.

[0046] Specifically, the expression for the DB algorithm is:

[0047]

[0048] For an approximate binarized image, Pi,j represents the pixel points on the sample binarized image, and T... i，j The grayscale threshold is a preset value that is constant at any pixel coordinate. K is the magnification factor, and (i, j) are the pixel coordinates.

[0049] It is worth mentioning that the binary distribution ratio threshold standard is the result of the ratio of the number of black points (represented by a value of 1) to the total number of points in the set.

[0050] This embodiment relates to a method for obtaining a binary distribution ratio threshold standard for images of streetlight nameplates. Determining the clarity of a streetlight nameplate image requires multiple images of clearly defined streetlight nameplates as samples. This method extracts a binary distribution ratio threshold standard from the binarized images of the sample streetlight nameplates as the basis for judgment. It is worth noting that this embodiment only demonstrates a method for obtaining the binary distribution ratio threshold standard for a single streetlight nameplate image. In fact, each clearly defined streetlight nameplate image will yield a binary distribution ratio threshold standard. Due to differences in text information within the image, these binary distribution ratio threshold standards will exhibit some discrete distribution patterns. Therefore, it is necessary to determine the variance and mean of the binary distribution ratio threshold standard based on this discreteness. By using the variance and mean of the binary distribution ratio threshold standard to determine the floating threshold of the binary distribution ratio threshold standard, a usable binary distribution ratio threshold standard can be obtained.

[0051] In one embodiment of this application, S100 further includes:

[0052] S121, receives images of street light nameplates.

[0053] S122, Select the entire area of ​​the image for the street light sign.

[0054] S123, use the DB algorithm to binarize the global pixels of the selected street lamp nameplate image to form a binarized image of the street lamp nameplate image.

[0055] S124, Calculate the proportion of binary distribution in the binarized image of the street light sign.

[0056] S125, if the binary distribution ratio of the image of the street light sign is equal to or greater than the binary distribution ratio threshold standard, then the image of the street light sign is determined to be clear.

[0057] S126, if the binary distribution ratio in the binarized image of the street light sign is less than the binary distribution ratio threshold standard, then the image of the street light sign is determined to be blurry.

[0058] Specifically, when using the binary distribution ratio threshold standard, the received street light sign image first needs to be binarized using the DB algorithm. Based on the binarized image of the street light sign, the binary distribution ratio in the binarized image of the street light sign is statistically analyzed. Based on the binary distribution ratio threshold standard, it can be determined whether the image of the street light sign is blurry.

[0059] This embodiment relates to a method for determining whether an image of a street light sign is blurry. The DB algorithm for binarizing street light sign images is a mature algorithm with high fault tolerance and small code size, which can improve the efficiency of determining whether a street light sign image is blurry.

[0060] In one embodiment of this application, S200 includes:

[0061] S211: The blurry street light sign image is used as a coordinate point to form a set of points to be processed.

[0062] S212 uses Gaussian noise to degenerate the set of points to be processed, forming the initial iteration function.

[0063] S213 receives the number of iterations and uses an iterative algorithm to iterate the initial iterative function to obtain the final iterative function.

[0064] S214 uses the non-blind deconvolution algorithm Richardson-Lucy to restore the final iterative function to the first image.

[0065] Specifically, the initial iteration function is:

[0066] g(x,y)=f(x,y)×h(x,y)+Δn(x,y)

[0067] f(x,y) represents the initial blurred image, h(x,y) represents the point spread function of the initial blurred image, Δn(x,y) represents the Gaussian noise function, g(x,y) represents the noisy blurred image, and (x,y) represents the coordinates of the pixel.

[0068] The iteration function at the Kth iteration:

[0069]

[0070]

[0071] in, This represents the restored image obtained in the k-th iteration. This represents the restored image obtained in the (k-1)th iteration. Let represent the point spread function obtained in the (k-1)th iteration. for transpose, Let represent the point spread function obtained in the k-th iteration. for The transpose of , × is the convolution operation, i represents the restored image or point spread function of the (k-1)th iteration, and i+1 represents the restored image or point spread function of the kth iteration.

[0072] This embodiment relates to a method for sharpening images of street light signboards. The convolutional iterative algorithm can quickly sharpen the images of street light signboards. It is worth mentioning that, due to the high deblurring capability of the convolutional iterative algorithm, the probability of blurring the first image is low in actual use, which also improves the efficiency of collecting street light information and positioning.

[0073] In one embodiment of this application, S200 further includes:

[0074] S221 smooths the iterative function for each iteration.

[0075] S222, calculate the local gradient and edge direction of the coordinate points on the smoothed iterative function.

[0076] S223, receives the definition of local gradient maximum point as 1 and local gradient non-maximum point as 0, and uses the local gradient of the coordinate point and the edge direction to obtain a binarized matrix, and assigns the binarized matrix to each pixel to weight the iterative function.

[0077] Specifically, when smoothing the iterative function for each iteration, a Gaussian filter with a known standard deviation needs to be specified.

[0078] Calculate the local gradient for each pixel. Its expression is:

[0079]

[0080] The expression for calculating the local gradient θ(x,y) at each pixel is as follows:

[0081]

[0082] in: This represents the gradient in the x-direction. This represents the gradient in the y-direction.

[0083] This embodiment relates to a method for reconstructing a first image using an iterative function. Optimizing and weighting the iterative function can improve the rationality of the pixel distribution in the first image, thereby improving its clarity.

[0084] In one embodiment of this application, S600 includes:

[0085] S611, receives a database of commonly used characters on nameplates;

[0086] S612, retrieves the binarized image of the street light nameplate.

[0087] S613 uses an edge detection algorithm to divide the binarized image of the street light sign into multiple text division regions.

[0088] S614, Select a text partition.

[0089] S615, determine whether the suspected text formed by all pixels of the text division area matches the database of commonly used characters for nameplates.

[0090] Specifically, using the suspected text as an index, the suspected text is searched sequentially in the table of commonly used Chinese characters, the table of English alphabets, and the table of Arabic numerals. If the suspected text is found in at least one of the three tables, it is considered that the suspected text formed by all pixels of the text division area matches the database of commonly used characters for nameplates.

[0091] S616, if the suspected text formed by all pixels of the text division area matches the common text database of nameplates, then the matching text in the common text database of nameplates is mapped to the text division area.

[0092] S617 If the suspected text formed by all pixels of the text division area does not match the database of commonly used characters for nameplates, then the text division area is ignored.

[0093] S618 returns the selected text partition, continuing until all text partitions have been traversed.

[0094] Specifically, the database of commonly used characters for nameplates includes a table of commonly used Chinese characters, an English alphabet, and Arabic numerals. When using the binarized image of a streetlight nameplate, the mobile terminal cannot directly extract the text. It can only use edge detection algorithms to divide the binarized image of the streetlight nameplate to obtain segmented regions. These segmented regions are suspected text, so it is necessary to use the database of commonly used characters for nameplates to determine the similarity between the point set within the segmented region and the text in the database.

[0095] This embodiment involves the collection of street light information. A database of commonly used characters on nameplates can be used to extract text information from images of street light nameplates. This significantly reduces the workload for operators.

[0096] In one embodiment of this application, S600 further includes:

[0097] S621, Generate a blank table for text arrangement based on the arrangement of the text division areas in the binarized image of the street light sign.

[0098] S622: Import the matched text from the text database mapped by the text division area into a blank text arrangement table to generate a text arrangement table for the image of the street lamp nameplate.

[0099] S623 receives the pre-defined text inclusion rules, extracts text information from the text sorting list, and obtains the street light information contained in the text information.

[0100] This embodiment relates to a method for including predetermined text. Since the extraction of text from streetlight sign images utilizes the binarized image of the streetlight sign and edge detection algorithms, a blank text arrangement table can be generated simply by analyzing the arrangement of text regions within the binarized image of the streetlight sign image. Since some text information in the text arrangement table of the streetlight sign image is not the target text, receiving predetermined text inclusion rules allows for the inclusion of predetermined text from the streetlight information.

[0101] In another embodiment of this application, S600 further includes:

[0102] S631, Establish a keyword extraction model. The keyword extraction model includes rules for incorporating text.

[0103] S632, receives standard text arrangement samples.

[0104] S633, using a keyword extraction model, keywords are extracted from the standard text arrangement samples, and the standard text arrangement samples are sorted according to keyword weights. The text inclusion rule is the standard result of the keyword weight sorting.

[0105] S634, based on the keyword weight ranking result, adjusts the keyword weight and number of keywords, and returns the standard arrangement sample of the received text until the keyword weight ranking result is more than 95% similar to the rules of the included text, and generates the trained keyword extraction model.

[0106] S635, calls the text sorting list.

[0107] S636 uses a trained keyword extraction model to extract text information from a text sorting list in order to obtain street light information contained in the text information.

[0108] This embodiment relates to a method for recording predetermined text. Because different nameplates have different text arrangement rules, the arrangement of key information, i.e., keywords, varies. The keyword extraction model can handle different nameplate engraving text arrangement rules.

[0109] In one embodiment of this application, S700 includes:

[0110] S711, retrieve the image storage directory of the street light nameplate.

[0111] S712 uses the image storage directory of the street light nameplate to lock multiple information storage addresses that form a mapping relationship with the image storage directory of the street light nameplate, forming a set of addresses to be filtered.

[0112] S713 receives latitude and longitude information from images of street light nameplates, including the format and keywords for saving the information.

[0113] S714: Using the latitude and longitude information storage format and keywords of the street light nameplate image, traverse each stored information in the address set to be filtered to obtain the latitude and longitude information of the street light nameplate image.

[0114] This embodiment relates to a method for obtaining the latitude and longitude information of images of streetlight nameplates. A mobile terminal is equipped with a positioning module. After the positioning module obtains geographic location information, the mobile terminal needs to store this information and establish a mapping relationship with the images of streetlight nameplates. When retrieving this geographic location information, it is only necessary to lock multiple information storage addresses that are mapped to the storage directory of the streetlight nameplate images, forming a set of addresses to be filtered. Based on the latitude and longitude information storage format and keywords of the streetlight nameplate images, the latitude and longitude information of the streetlight nameplate images can be retrieved. This greatly improves the efficiency of location information acquisition.

[0115] In one embodiment of this application, S700 further includes:

[0116] S721, call the set of addresses to be filtered.

[0117] S722, Select an address to filter.

[0118] S723, retrieve the stored data within the address to be filtered.

[0119] S724: Using the latitude and longitude information storage format of the street light nameplate image, screen the stored data in the address to be screened, and determine whether the stored data in the address to be screened conforms to the latitude and longitude information storage format of the street light nameplate image.

[0120] S725, if the stored data in the address to be filtered does not conform to the latitude and longitude information saving format of the street light nameplate image, then return to select an address to be filtered.

[0121] S726, if the stored data in the address to be filtered matches the latitude and longitude information saving format of the street lamp nameplate image, then call the keywords of the latitude and longitude information of the street lamp nameplate image to lock the latitude and longitude information of the street lamp nameplate image, so as to obtain the latitude and longitude information of the street lamp nameplate image.

[0122] This embodiment relates to a method for traversing the latitude and longitude information storage addresses of street light sign images. In practice, determining the latitude and longitude information storage address of a street light sign image from multiple data storage addresses, and identifying the location information within that address, requires the storage format and keywords of the street light sign image's latitude and longitude information. These two directional rules can significantly shorten the traversal time and improve efficiency. Simultaneously, the formation of the address set to be filtered also reduces the range of addresses to be traversed.

[0123] This application also provides a mobile terminal for identifying and locating street light nameplates.

[0124] like Figure 2 As shown in one embodiment of this application, a street light sign identification and positioning mobile terminal includes a mobile terminal 100 and an Internet of Things management system 200.

[0125] The mobile terminal 100 is used to execute a street light nameplate identification and positioning method. The mobile terminal includes a camera module 110, a positioning module 120 and a storage module 130.

[0126] The Internet of Things management system 200 is communicatively connected to the mobile terminal 100.

[0127] This embodiment relates to a mobile terminal for identifying and locating street light signage. The mobile terminal 100 includes a camera module, a positioning module, and a storage module. The camera module 110 is capable of taking pictures of the street light signage, and the positioning module 120 is capable of locating the street light signage. The storage module 130 stores photos of the street light signage and its location information.

[0128] The technical features of the embodiments described above can be combined arbitrarily, and the execution order of the method steps is not limited. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification. The embodiments described above only illustrate several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A street light nameplate recognition positioning method, characterized in that, include: The S100 receives images of street light nameplates and determines whether the images are blurry. If the image of the street lamp nameplate is blurry, the image blurring algorithm is used to clarify the image of the street lamp nameplate to form a first image, and it is determined whether the first image is blurry. S300, if the first image is blurry, a prompt message is sent, and the received image of the street light nameplate is returned to determine whether the image of the street light nameplate is blurry; S400 If the first image is not blurry, then extract the text information of the image of the street lamp nameplate; S500 If the image of the street lamp nameplate is clear, then extract the text information of the image of the street lamp nameplate; in the process of extracting the text information of the image of the street lamp nameplate, the text arrangement format is preserved; The S600 receives a pre-defined text inclusion rule and extracts the street light information contained in the text information of the street light nameplate image according to the pre-defined text inclusion rule. The S700 analyzes the image of the street light sign and obtains the latitude and longitude information of the image. The S800 uploads the street light information and latitude and longitude information of the street light nameplate image to the IoT management system, and returns the received street light nameplate image to determine whether the street light nameplate image is blurry, until no more street light nameplate images are received; Among them: S200 includes: S211, use pixels as coordinates to form a set of points to be processed from the blurry image of the street lamp sign; S212, using Gaussian noise to degenerate the set of points to be processed, forming the initial iteration function; S213 receives the number of iterations and uses the iterative algorithm to iterate the initial iterative function to obtain the final iterative function; S214 uses the non-blind deconvolution algorithm Richardson-Lucy to restore the final iterative function to the first image; Specifically, the initial iteration function is: This represents the initial blurred image. The point spread function represents the initial blurred image. Represents the Gaussian noise function. This indicates a noisy and blurred image. Represents the coordinates of a pixel; The iteration function at the Kth iteration: in, This represents the restored image obtained in the k-th iteration. This represents the restored image obtained in the (k-1)th iteration. Let represent the point spread function obtained in the (k-1)th iteration. for transpose, Let represent the point spread function obtained in the k-th iteration. for transpose, For convolution operations, This represents the restored image or point spread function for the (k-1)th iteration. Represents the restored image or point spread function for the k-th iteration; The S200 also includes: S221, smooth the iteration function for each iteration; S222, calculate the local gradient and edge direction of the coordinate points on the smoothed iterative function; S223, Receive the definition that the local gradient maximum point is 1 and the local gradient non-maximum point is 0, and use the local gradient of the coordinate point and the edge direction to obtain the binarization matrix, and assign the binarization matrix to each pixel to weight the iterative function; Specifically, when smoothing the iterative function for each iteration, a Gaussian filter with a known standard deviation needs to be specified. Calculate the local gradient for each pixel. Its expression is: Calculate the local gradient for each pixel. Its expression is: in: This represents the gradient in the x-direction. This represents the gradient in the y-direction.

2. The street light nameplate identification and positioning method according to claim 1, characterized in that, The process of receiving an image of a street light sign and determining whether the image is blurry includes: Receive a sample image of a clearly defined street light sign. The DB algorithm is used to binarize all pixels of the sample image of the street lamp nameplate, forming a sample binarized image of the sample image of the street lamp nameplate. The proportion of binary distribution in the binary image of the sample is statistically analyzed, and a threshold standard for the binary distribution proportion is generated based on the proportion of binary distribution in the binary image of the sample; the threshold standard for the binary distribution proportion is the percentage of the number of black pixels out of the total number of pixels.

3. The street light nameplate identification and positioning method according to claim 2, characterized in that, The process of receiving an image of a street light sign and determining whether the image is blurry also includes: Receive images of street light nameplates; Select the entire image of the street light sign; The DB algorithm is used to binarize the global pixels of the selected street lamp nameplate image, forming a binarized image of the street lamp nameplate image. Statistical analysis of the binary distribution ratio in the binarized image of the street light sign; If the binarized image of the street light sign is equal to or greater than the binarized image threshold, then the image of the street light sign is considered clear. If the proportion of the binary distribution in the binarized image of the street light sign is less than the threshold standard for the proportion of the binary distribution, then the image of the street light sign is determined to be blurry.

4. The street light nameplate identification and positioning method according to claim 3, characterized in that, If the image of the street light sign is blurry, an image blurring algorithm is used to sharpen the image of the street light sign, forming a first image, and it is determined whether the first image is blurry, including: The blurry street light sign image is used as a coordinate point to form a set of points to be processed; Gaussian noise is used to degenerate the set of points to be processed, forming an initial iteration function; Receive the number of iterations and use the iterative algorithm to iterate the initial iterative function to obtain the final iterative function; The Richardson-Lucy algorithm, a non-blind deconvolution algorithm, is used to restore the final iterative function to the first image.

5. The street light nameplate identification and positioning method according to claim 4, characterized in that, The step of using the non-blind deconvolution algorithm Richardson-Lucy to restore the final iterative function to the base first image includes: Smooth the iteration function for each iteration; Calculate the local gradient and edge direction of the coordinate points on the smoothed iterative function; The function receives a local gradient maximum point defined as 1 and a local gradient non-maximum point defined as 0. It uses the local gradient of the coordinate point and the edge direction to obtain a binarized matrix, and assigns the binarized matrix to each pixel to weight the iterative function.

6. The street light nameplate identification and positioning method according to claim 5, characterized in that, The receiving of the predetermined text inclusion rules involves selecting text information from the street light nameplate image based on these rules to obtain street light information, including: Receive a database of commonly used characters for nameplates; the database includes a table of commonly used Chinese characters, an English alphabet, and an Arabic numeral table. Retrieve the binarized image of the street light nameplate; The binarized image of the street light sign image is divided using an edge detection algorithm to obtain multiple text segmentation regions; Select a text area; Determine whether the suspected text formed by all pixels in the text segmentation area matches the database of commonly used characters for nameplates; If the suspected text formed by all pixels of the text division area matches the database of commonly used characters for nameplates, then the matching text in the database of commonly used characters for nameplates will be mapped to the text division area. If the suspected text formed by all pixels in the text division area does not match the database of commonly used characters for nameplates, then the text division area is ignored. Return to the selected text partition and continue until all text partitions have been traversed.

7. The street light nameplate identification and positioning method according to claim 6, characterized in that, The method of receiving predetermined text inclusion rules, selecting text information from images of street light nameplates according to predetermined text inclusion rules to obtain street light information, also includes: Generate a blank table of text arrangement based on the arrangement of text division areas in the binarized image of the street lamp nameplate; Import the matched text from the text database mapped by the text division area into a blank text arrangement table to generate a text arrangement list for the image of the street lamp nameplate. Receive the pre-defined text inclusion rules, extract text information from the text list, and obtain the street light information contained in the text information.

8. The street light nameplate identification and positioning method according to claim 7, characterized in that, The process of analyzing the image of the street light sign to obtain its latitude and longitude information includes: Retrieve the image storage directory of the street light nameplates; By utilizing the image storage directory of the street light sign, multiple information storage addresses that form a mapping relationship with the image storage directory of the street light sign are locked to form a set of addresses to be filtered; The format and keywords for saving the latitude and longitude information of images received from street light nameplates; By using the latitude and longitude information storage format and keywords of the street light sign images, the stored information in the address set to be filtered is traversed to obtain the latitude and longitude information of the street light sign images.

9. A method for identifying and locating street light nameplates according to claim 8, characterized in that, The method of using the latitude and longitude information storage format and keywords of the street light sign image to traverse each stored information in the address set to be filtered, and obtaining the latitude and longitude information of the street light sign image, includes: Call the set of addresses to be filtered; Select an address to filter; Retrieve the stored data within the address to be filtered; Using the latitude and longitude information storage format of the street light nameplate image, screen the stored data in the address to be screened, and determine whether the stored data in the address to be screened conforms to the latitude and longitude information storage format of the street light nameplate image; If the stored data in the address to be filtered does not conform to the latitude and longitude information saving format of the street light nameplate image, then return to select an address to be filtered; If the stored data in the address to be filtered matches the latitude and longitude information storage format of the street light sign image, then the keywords of the latitude and longitude information of the street light sign image are called to lock the latitude and longitude information of the street light sign image, so as to obtain the latitude and longitude information of the street light sign image.

10. A street light nameplate identification and positioning system, characterized in that, include: A mobile terminal is used to execute a street light nameplate identification and positioning method as described in any one of claims 1 to 9; The mobile terminal includes a camera module, a positioning module, and a storage module; an Internet of Things (IoT) management system is communicatively connected to the mobile terminal.

Images