WO2018086208A1

WO2018086208A1 - Scene image-based advertising method and system

Info

Publication number: WO2018086208A1
Application number: PCT/CN2016/112061
Authority: WO
Inventors: 曾昭兴
Original assignee: 广州市沃希信息科技有限公司
Priority date: 2016-11-10
Filing date: 2016-12-26
Publication date: 2018-05-17
Also published as: CN106530008B; CN106530008A

Abstract

A scene image-based advertising method and system. The method comprises the following steps: obtain a rough position of a user; obtain, according to the rough position, a selected position in a first preset range, and obtain first scene images corresponding to the selected positions; the user photographs a currently viewed second scene image and uploads same to a server; obtain the current position of the user according to similarity comparison between the first scene images and the second scene image, the first scene images being panoramic images photographed at the selected positions by rotating a circle; and display advertisement information to the user when uploading the second scene image to the server. On one hand, a user can obtain more accurate positioning coordinates. On the other hand, an advertiser can carry out advertising by means of a taxi taking platform.

Description

Description: A method and system for advertising based on scene pictures

[0001] The present invention relates to the field of urban traffic advertising, and in particular, to a method and system for advertising based on scene pictures.

Background technique

[0002] Taxi software has changed the traditional way of taxiing, and established and cultivated the modern mobile travel mode that users led by the big mobile Internet. Compared with traditional telephone recruiting and roadside recruiting, the birth of taxi software has changed the traditional taxi market pattern, subverting the concept of roadside blocking, using the characteristics of mobile Internet, integrating online and offline, from taxi From the initial stage to getting off the bus, paying the fare on the line, changing the traditional taxi driver and other passenger modes, allowing the driver to "take orders" according to the passenger's destination, saving the driver and passenger communication costs, reducing the vacancy rate, and maximizing the savings of both sides. Resources and time.

technical problem

[0003] After a passenger places an order, how quickly the driver picks up the driver poses a great challenge. At present, according to the intelligent navigation driver, the approximate location of the passenger can be quickly known, but the location is not accurate enough. For example, if a passenger takes a taxi in a strange place, he does not know how to describe his specific location. Therefore, he is on the left side of a street. Because the satellite positioning error is generally within 10 meters, the passengers are on the map. It is on the right side of the street. This time, the driver often needs to go around a long way. Therefore, it is difficult for a driver to find a passenger with a simple address after positioning. In some cases, even the driver is very close to the passenger but they can't find each other. Even if the driver confirms the phone, it may not be able to find the passenger accurately, so that the driver may circle around the destination to find the passenger after arriving at the destination. In order to improve efficiency, how to quickly and accurately pick up the driver becomes a problem to be solved. At present, according to the intelligent navigation driver, the approximate location of the passenger can be quickly known, but the location is not accurate enough. At the same time, as merchandise advertisements bring huge profits to the merchants, most merchants are constantly searching for suitable platforms for advertising. A kind of advertising method that allows the user to obtain more accurate positioning coordinates, and also enables the merchant to advertise the product in the appropriate scene becomes a new user demand. Problem solution Technical solution

[0004] Based on the above situation, the present invention provides a method for advertising based on a scene picture, in which a plurality of sets of panoramic pictures rotated one week at a selected position are pre-stored in the server, and the first scene picture synthesized by the user is used. The scene picture taken by the location and the uploaded picture are compared with the first scene picture for similarity; the user can obtain the more accurate coordinates, and then obtain the user's best boarding position. Use the user to upload the image to show the appropriate ads to the user. On the one hand, the user can get more accurate positioning coordinates, on the other hand, the advertiser can use the taxi platform to carry out appropriate advertising.

[0005] The present invention provides an advertisement method based on a scene picture, including the steps of: obtaining a rough position of a user; acquiring a selected position in the first preset range according to the coarse position, and acquiring a corresponding position of each selected position a scene picture; the user captures the currently viewed second scene picture and uploads it to the server; and obtains the current position of the user according to the similarity comparison between the first scene picture and the second scene picture; the first scene picture A panoramic picture taken for one rotation at each selected position; after uploading the second scene picture to the server, the advertisement information is displayed to the user.

[0006] The second scene picture is composed of two or more pictures taken by the user at different positions using the different shooting angles at the current position; the different shooting angles include at least a first shooting angle and a second shooting angle.

[0007] The method for obtaining the optimal boarding position of the user by comparing the similarity of the first scene picture and the second scene picture comprises: obtaining the first by comparing the similarity between the first scene picture and the second scene picture a first photographing direction of the photographing angle image; obtaining a second photographing direction of the second photographing angle image by comparing the similarity between the first scene image and the second scene image; according to the first photographing direction and the The intersection of the two shooting directions to determine the current position of the user.

[0008] The user uploads a picture taken by the third shooting angle and acquires a third shooting direction; determines whether the first shooting direction, the second shooting direction, and the third shooting direction intersect at one point, and if so; uses the point as the current coordinate of the user; If not, the intersection points of the first shooting direction, the second shooting direction, and the third shooting direction may form a triangle, and the center of gravity of the triangle is used as the current coordinate of the user.

[0009] acquiring the merchant information in the second preset range of the coarse position; and embedding the related advertisement of the merchant information in the scene picture.

[0010] The present invention further provides an advertisement system based on a scene picture, including a user location acquisition module, a scene picture information acquisition module, a scene picture information processing module, and a network connection of each module; The module is configured to obtain a rough location of the user; the scene image information obtaining module is configured to obtain a selected location in the preset range according to the coarse location, and obtain a first scene image corresponding to each selected location; the scene image information processing module is configured to use The second scene picture taken by the user is compared with the first scene picture to obtain the current position of the user; the first scene picture here is a panoramic picture rotated one week at each selected position.

Advantageous effects of the invention

Beneficial effect

[0011] Compared with the prior art, the present invention is simple in operation, and obtains a first scene picture corresponding to each selected position by acquiring a rough position of the user; the user captures the currently seen second scene picture and uploads it to The server obtains a current location of the user according to the comparison between the first scene picture and the second scene picture. In this way, the pick-up passenger can obtain a more precise position of the passenger, thereby avoiding the problem of the detour caused by the positioning error, and saving the waiting time of the passenger and the driver. At the same time, advertisers can also conduct appropriate advertising, while taxi platforms can increase the revenue of advertising.

Brief description of the drawing

DRAWINGS

1 is a schematic flow chart of an embodiment of a method for advertising based on a scene picture according to the present invention;

2 is a schematic diagram of an embodiment of a method for acquiring a current location according to the present invention;

3 is a schematic structural diagram of an embodiment of a scene picture based advertisement system according to the present invention.

BEST MODE FOR CARRYING OUT THE INVENTION

[0015] The solution of the present invention will be described in detail below in conjunction with the preferred embodiments thereof.

[0016] As shown in FIG. 1, the method in this embodiment includes the following steps:

[0017] S101: Obtain a rough location of the user.

[0018] The user snores the taxi software, the taxi software obtains the rough position of the user according to the positioning system (GPS, Beidou), and can automatically generate the corresponding address name, because the range indicated by an address name may be very large (such as a bus) The name of the station, most of which does not know which side of the bus stop), may cause the driver to travel a long way when picking up passengers. After obtaining the current rough position of the user, the system In order to obtain selected locations within a predetermined range, these selected locations are generally empirical values derived by the passenger after multiple rides. In most cases, passengers are unlikely to be very familiar with the name of the boarding and disembarking location and the scene pictures they can see. This results in the inability to accurately reach the destination even after the user knows the name of the best boarding location. Ground. It is also possible that the user sees the corresponding scene but cannot get the corresponding location name.

[0019] acquiring a selected location in the first preset range according to the coarse location, and acquiring a first scene picture corresponding to each selected location.

[0020] After the system acquires the selected location within the first preset range, the user can see the corresponding location name of the corresponding road segment on the map. In general, after GPS positioning, the system will provide a number of locations around its location as the selected location (or recommended on-board location). The corresponding scene picture can be preset at the selected position on the system. In this way, when the selected location is obtained, it can acquire the scene picture corresponding to each selected location and display it to the user for selection. The scene picture here is used by the panorama image used at the selected position for one rotation at each selected position (of course, it is also possible to use a smaller rotation angle because generally not all angles have markers). The selected position is a position in the range of one location to upload the panoramic picture. The panoramic picture taken by the one-time rotation of the selected position may have a different effect depending on the distance and the change of the light. Therefore, the number of selected locations should be as many as possible, and different panoramas should be set in the middle. This can reduce the poor matching effect caused by the increased distance between the selected position and the actual position and the change in the day. In addition, the panoramic picture can be displayed as a rectangular image on the display or mobile phone, and the length of the rectangle divided by 360, you can get a partial picture corresponding to each angle. The scene picture corresponding to each selected location (that is, the first scene picture) is uploaded to the server, where the uploaded picture needs to be angle-marked (ie, which position of the picture corresponds to how many degrees).

[0021] S103: The user captures the currently viewed second scene image and uploads it to the server, and obtains the current location of the user according to the comparison between the first scene image and the second scene image; The server 吋 displays the advertisement information to the user.

[0022] pre-storing a scene picture of a plurality of places in the server, and acquiring a plurality of groups of pre-stored locations (ie, selected locations) within a predetermined range of the user's rough location (eg, within a range of 100 meters) through the current rough location information of the user.

) The first scene picture. The user can take a picture of the second scene that he currently sees and upload it to the service. The server compares the first scene picture with the second scene picture system to know the current selected location of the user. In some places where the traffic environment is simple, a scene picture can clearly locate the user, and the similarity between the partial picture of the first scene picture and the second scene picture is more than 99%. It is possible to determine that the user's current location is almost coincident with the system's selected location. However, in a more complicated traffic environment, the positioning effect of an image will be greatly weakened.

[0023] As a better example of implementation, the second scene picture here is a scene picture taken using different angles. Due to the certain error of satellite positioning accuracy, most mobile phone positioning can only locate one street. It is difficult to accurately locate which side of the street, which increases the difficulty of pick-up and drop-off for taxi drivers. The scene picture is compared with the pre-stored first scene picture, which can accurately know the location of the user and solve the inconvenience caused by the traditional satellite positioning. In general, the shooting direction of the picture can be obtained by comparing the first scene picture and the shooting angle picture, and the connection direction of the camera and the reference object is the shooting direction (the direction from the camera to the reference object from the plane).

[0024] The method for obtaining a current position of a user by using a scene picture includes: acquiring a first shooting direction of a first shooting angle picture (the first shooting direction may also be a first shooting direction line; other directions may also be deduced by analogy), Obtaining a second shooting direction of the second shooting angle picture; determining a current position of the user according to an intersection of the first shooting direction and the second shooting direction. Since the distance of the position is not far, the similarity between the scene picture taken at the selected position and the scene picture taken by the user's current position is very high, so the current position of the user can be determined according to the intersection of the first shooting direction and the second shooting direction. (The more the number of selected positions, the more accurate the current position is.) The peers can also easily get the best boarding position from the selected location based on the most recent selection of the current location. The first shooting angle picture and the second shooting angle picture here are uploaded second scene pictures.

[0025] As shown in FIG. 2, the first user A is at the bottom of the street, and the second user B is at the other side of the street; A is photographed by taking reference object A, reference B, and reference C and uploading to the server, through the system. The first shooting direction of the reference object A ( can be obtained separately (the direction of the pointing object A, the embodiment is the direction of the straight line, and so on), and the second of the reference object B 拍摄The shooting direction, the current position of A can be determined by the intersection of the various directions (ie, the intersection of the reverse extension lines of the camera direction). Similarly, the current position of B can also be obtained by this method. Here, in theory, the current position of A can be determined by two directions. The third shooting direction of the photographing reference C吋 can be used for the school. Quasi. The third shooting direction can be obtained by the user taking the reference C at the current position and uploading it to the server. The method of calibrating with the third shooting direction calibration is: determining whether the first shooting direction, the second shooting direction, and the third shooting direction intersect at one point, and if so; using the point as the precise coordinate of the current A (or B). If not, the intersection of the first shooting direction, the second shooting direction and the third method may form a triangle, and the center of gravity of the triangle (according to the actual situation, the center of gravity here may also be replaced by the outer center and the inner core of the triangle) , center of mind or side heart) as the coordinates of the current A (or B). In the case of a triangle, it indicates that the current position of the user is deviated from the shooting position of the pre-existing server. After the user selects the pre-stored scene picture, it can be proved that the position is relatively close to the precise coordinate in the pre-existing system, and the triangle is taken. The center of gravity can be combined with the actual coordinates of the user and the precise coordinates in the pre-existing system, so that the calibrated coordinates are closer to the true coordinates. By such a more precise positioning, after receiving the rough position information of the passenger, the taxi locator can quickly know which side of the street the passenger is on, thereby saving the time for finding the passenger.

[0026] As shown in FIG. 3, A is in the selected position and uses the first scene picture (the same first scene picture can be selected in the vicinity of the selected position), the first shooting angle after shooting A The picture and the second shooting angle picture are uploaded to the server. The system compares the similarity between the first scene picture and the first shooting angle picture, and can compare the position of the shooting position in the first scene picture, and then can convert the first shooting direction; similarly, the system according to the The second photographing direction can be converted by comparing the similarity between the scene image and the second photographing angle image. The panoramic picture corresponding to the first scene picture is displayed on the display or the mobile phone as a rectangular image, and the length of the rectangle is divided by 360, and the partial picture corresponding to each angle; and the shooting device used by the user also has a corresponding lens angle ( For example, 150°); the first shooting angle picture is also rectangular after shooting, and the length of the rectangle is divided by 150, and a part of the picture corresponding to each angle can be obtained; thus, 1° and the first scene in the first shooting angle picture are extracted. Comparing 1° in the picture, the similarity can be determined relatively quickly; as a test, the picture of the first shooting angle picture is compared with the first scene picture multiple times, and finally the first shooting angle picture is taken 75°. An intermediate value is used as the corresponding direction of the first shooting angle direction, and only 1° of data can be taken. Since the entire captured picture is a parallel plane, the directions of the points are parallel, and the intersecting points are shown as planes, and from the three-dimensional view, the two parallel faces intersect.

[0027] In order to make the user, the taxi platform and the merchant get better economic benefits. The user is using the above solution After obtaining the location information, you can use both the ad mode and the ad-free mode. In the advertisement mode, the merchant information in the second preset range of the user's end position is obtained; after the user uploads the second scene image to the server, the advertisement of the merchant information is displayed to the user. Simultaneously, obtaining the business information in the second preset range of the user end position (or coarse position); and displaying the corresponding business advertisement to the user, the user can further understand the business around the destination (or current location) , to achieve the purpose of accurate advertising. The user's end position is the information of the starting point and the ending point of the user's taxi input. When the user goes to the destination, he or she will usually shop or do the corresponding business there. At this time, there is an advertisement of the corresponding merchant, which can help the user. Better understand the situation of the destination, reduce the user's corresponding travel cost according to the length of the advertisement (that is, the longer the travel time is reduced, the more the ride cost), the peer can also be based on the frequency of the advertisement. The advertiser is charged a corresponding fee (ie, the more frequently the advertisement is paid, the more the fee is charged to the advertiser). In this mode, the user can reduce part of the ride cost, the advertiser can also carry out appropriate advertising, and the taxi platform can increase the advertising revenue.

[0028] FIG. 4 is a schematic structural diagram of an embodiment of a scene picture based advertisement method system according to the present invention.

[0029] An advertisement system based on a scene picture, comprising: a user location acquisition module, a scene picture information acquisition module, a scene picture information processing module, and a network connection of each module; a user location acquisition module is configured to acquire a rough location of the user; The information acquiring module is configured to obtain a first scene picture corresponding to each selected position according to the selected position in the preset range according to the rough location; the scene picture information processing module is configured to use the second scene picture and the first scene picture that are captured by the user The comparison is to obtain the current location of the user; the first scene picture here is a panoramic picture rotated one week at each selected position, and after uploading the second scene picture to the server, the advertisement information is displayed to the user.

[0030] Pre-storing a scene picture of a plurality of places in the server, and acquiring, by the user's current rough position information, a first scene picture of a plurality of groups of pre-stored locations within a predetermined range of the user's rough position (eg, within a range of 100 meters). The user can capture the second scene picture that he currently sees and upload it to the server. By comparing the first scene picture with the second scene picture system, the user can generally know the current selected position of the user. In some places where the traffic environment is simple, a scene picture can clearly locate the user.

There is also a similarity between the partial picture of the first scene picture and the second picture picture, which is more than 99%. This time, it can be judged that the current position of the user and the selected position of the system almost coincide. But in the more complex In a complicated traffic environment, the positioning effect of a picture will be greatly weakened.

[0031] As a better example of implementation, the second scene picture here is a scene picture taken using different angles. Due to the certain error of satellite positioning accuracy, most mobile phone positioning can only locate one street. It is difficult to accurately locate which side of the street, which increases the difficulty of pick-up and drop-off for taxi drivers. The scene picture is compared with the pre-stored first scene picture, which can accurately know the location of the user and solve the inconvenience caused by the traditional satellite positioning. In general, the shooting direction of the picture can be obtained by comparing the first scene picture and the shooting angle picture, and the connection direction of the camera and the reference object is the shooting direction (the direction from the camera to the reference object from the plane).

[0032] The method for obtaining a current location of the user, comprising: acquiring a first shooting direction of the first shooting angle picture, and acquiring a second shooting direction of the second shooting angle picture; according to the first shooting direction and the second shooting direction The intersection of the user to determine the current location. Since the distance of the position is not far, the similarity between the scene picture taken at the selected position and the scene picture taken by the user's current position is very high, so the current position of the user can be determined according to the intersection of the first shooting direction and the second shooting direction. (The more the number of selected positions, the more accurate the current position is.) The peer can also easily get the best boarding position from the selected location based on the most recent selected location of the current location. The first shooting angle picture and the second shooting angle picture here are uploaded second scene pictures.

[0033] As a better example, the system further includes a positioning calibration module, configured to obtain a third shooting direction according to the third shooting angle picture uploaded by the user station at the current position, and then determine the first shooting direction and the second shooting. Whether the direction and the third shooting direction intersect at one point, if yes; use this point as the exact coordinate of the current A (or B). If not, the intersection points of the first shooting direction, the second shooting direction and the third shooting direction may form a triangle, and the center of gravity of the triangle (according to the actual situation, the center of gravity here may also be replaced by the outer center of the triangle, Inner, heart or side) as the coordinates of the current A (or B).

[0034] In order to make the user, the taxi platform and the merchant get better economic benefits. The system further includes an advertisement module, configured to display the advertisement to the user after the user uploads the second scene image to the server. Simultaneously, obtaining the business information in the second preset range of the user end position (or coarse position); and displaying the corresponding business advertisement to the user, the user can further understand the business around the destination (or current location) , to achieve the purpose of accurate advertising. The user's end position is the information of the starting point and the ending point of the time when the user takes a taxi. When the user goes to the destination, he or she will usually shop or do the corresponding business there; The advertisement of the merchant can help the user to better understand some situations of the destination. In this mode, users can reduce part of the ride cost, advertisers can carry out appropriate advertising, and the taxi platform can increase the revenue of the advertisement, and make higher discounts for the user to improve the user's use of the software system.

Embodiments of the invention

[0035] The present invention provides an advertisement method based on a scene picture, including the steps of: obtaining a rough position of the user; acquiring a selected position in the first preset range according to the coarse position, and acquiring a corresponding position of each selected position a scene picture; the user captures the currently viewed second scene picture and uploads it to the server; and obtains the current position of the user according to the similarity comparison between the first scene picture and the second scene picture; the first scene picture A panoramic picture taken for one rotation at each selected position; after uploading the second scene picture to the server, the advertisement information is displayed to the user. The second scene picture is composed of two or more pictures taken by the user at different positions using the different shooting angles at the current position; the different shooting angles include at least the first shooting angle and the second shooting angle. Through this implementation example, the user can obtain more accurate positioning by manual assistance under the condition that the satellite positioning accuracy is low, which can effectively avoid the driver from detouring. At the same time, advertisers can also place appropriate advertising, and taxi platforms can increase advertising revenue.

[0036] The method for obtaining the optimal boarding position of the user by comparing the similarity of the first scene picture and the second scene picture comprises: obtaining the first by comparing the similarity between the first scene picture and the second scene picture a first photographing direction of the photographing angle image; obtaining a second photographing direction of the second photographing angle image by comparing the similarity between the first scene image and the second scene image; according to the first photographing direction and the The intersection of the two shooting directions to determine the current position of the user. The user uploads a picture taken by the third shooting angle and obtains a third shooting direction; determines whether the first shooting direction, the second shooting direction, and the third shooting direction intersect at one point, and if so; uses the point as the current coordinate of the user; if not, Then, the intersection points of the first shooting direction, the second shooting direction, and the third shooting direction may form a triangle, and the center of gravity of the triangle is taken as the current coordinate of the user. Obtaining the business information in the second preset range of the rough position; and embedding the related advertisement of the business information in the scene picture.

[0037] The above described embodiments merely express embodiments of the present invention, the description of which is more specific and detailed, but not It can be understood as a limitation of the scope of the invention. It should be noted that a number of variations and modifications may be made by those skilled in the art without departing from the spirit and scope of the invention. Therefore, the scope of the invention should be determined by the appended claims.

Industrial applicability

The invention connects the network of each module by the user location acquisition module, the scene picture information acquisition module, the scene picture information processing module, and the respective module; and obtains the scene picture corresponding to each possible boarding position by acquiring the rough position of the user; Select each scene picture to be seen according to the scene that is currently seen; finally obtain the best boarding position of the user. It can allow the passengers to pick up passengers to obtain a more precise position of the passengers, thus avoiding the problem of bypass caused by positioning errors, saving passengers and drivers waiting for the day. At the same time, advertisers can also perform appropriate advertising, while taxi platforms can increase advertising revenue. The traditional mode of simple passenger positioning has been improved, and it has industrial applicability.

Claims

Claim

An advertisement method based on a scene picture, comprising the steps of: obtaining a rough position of a user; acquiring a selected position in the first preset range according to the coarse position, and acquiring a first scene picture corresponding to each selected position The user captures the currently viewed second scene image and uploads it to the server; acquires the current location of the user according to the similarity comparison between the first scene image and the second scene image; the first scene image is in each The selected position is rotated to rotate the panoramic picture of one week; after uploading the second scene picture to the server, the advertisement information is displayed to the user.

The advertising method according to claim 1, wherein the merchant information in the second preset range of the user end position is acquired; and the advertisement information of the merchant information is embedded in the scene picture.

The advertising method according to claim 1, comprising: determining, according to the record of obtaining the end position of the user, whether the user first arrives in the third preset range of the location, and if so, using the longer advertisement to advertise If not, select a shorter ad to advertise.

The advertising method according to any one of claims 1 to 3, wherein the second scene picture is composed of two or more pictures taken by the user at different positions at different positions in the current position; The angle includes at least a first shooting angle and a second shooting angle. The method for obtaining the current position of the user by comparing the similarity between the first scene picture and the second scene picture comprises: using the first scene picture and the second Acquiring a similarity of the scene picture to obtain a first shooting direction of the first shooting angle picture; obtaining a second shooting direction of the second shooting angle picture by comparing the similarity between the first scene picture and the second scene picture; The intersection of the first shooting direction and the second shooting direction is used to determine the current position of the user.

The advertising method according to claim 4, wherein the user uploads a picture taken by the third shooting angle and acquires a third shooting direction; and determines whether the first shooting direction, the second shooting direction, and the third shooting direction intersect at a point , if yes; the point is taken as the current coordinate of the user; if not, the two directions of the first shooting direction, the second shooting direction and the third shooting direction The intersection point may form a triangle, and the center of gravity of the triangle is taken as the current coordinate of the user.

An advertisement system based on a scene picture, comprising: a user location acquisition module, a scene picture information acquisition module, a scene picture information processing module, and a network connection of each module; the user location acquisition module is configured to obtain a rough location of the user; The scene picture information obtaining module is configured to obtain a first position image corresponding to each selected position according to the selected position in the first preset range according to the coarse position; the scene picture information processing module is configured to shoot according to a user And uploading a similarity between the second scene picture and the first scene picture to obtain a recently selected location of the user; the first scene picture is a panoramic picture rotated one week at each selected position, and the second picture is uploaded The scene picture is sent to the server, and the advertisement information is displayed to the user.

The advertising system according to claim 6, further comprising an advertisement module, configured to acquire merchant information in a second preset range of the user end position; and embedding the merchant information in the scene image Advertising information.

The advertising system according to claim 6, wherein the method for playing the advertisement information comprises: determining, according to the record of obtaining the end position of the user, whether the user first arrives in the third preset range of the location, and if so, Use longer advertising time to advertise; if not, select shorter advertising to advertise.

The advertising system according to any one of claims 6-8, wherein the second scene picture is composed of two or more pictures taken by the user at different positions at different positions in the current position; The angle includes at least a first shooting angle and a second shooting angle. The method for obtaining the optimal boarding position of the user by comparing the similarity between the first scene picture and the second scene picture comprises: using the first scene picture and the Obtaining a first photographing direction of the first photographing angle image by comparing the similarity of the second scene image; obtaining a second photographing direction of the second photographing angle image by comparing the similarity between the first scene image and the second scene image And determining a current location of the user according to an intersection of the first shooting direction and the second shooting direction.

The advertising system according to claim 9, further comprising a positioning calibration module, configured to obtain a third beat according to the third shooting angle picture uploaded by the user station at the current position. In the shooting direction, determine whether the first shooting direction, the second shooting direction, and the third shooting direction intersect at one point, if yes; use the point as the current coordinate of the user; if not, the first shooting direction, the second shooting direction, and the third The two or two intersection points of the shooting direction may form a triangle, and the center of gravity of the triangle is taken as the current coordinate of the user.