In-game advertising visibility tracking method based on ray casting

Abstract

The present invention discloses a method for tracking the visibility of in-game advertisements based on ray projection, and relates to the technical field of image analysis. The technical key points are as follows: using the rate of change of pixel points within a specific time, an image is divided into a non-projection area and a projection area, so that the area with the least impact on consumers is determined before advertisement projection, thereby facilitating the subsequent placement of advertisements in the best viewing position for consumers as much as possible without affecting the consumer's gaming experience, and providing a basis for improving the observability of advertisements; in addition, the color difference comparison value between the image to be advertised and the projection area is compared with the adjustment threshold interval to obtain an advertisement suitable for the current projection game scene, thereby reducing the color conflict between the advertisement image and the game projection image while ensuring the clear display of the advertisement image, thereby further improving the observability of the advertisement image, and at the same time will not affect the normal observation of consumers.

Description

Technical Field

[0001] The present invention relates to the field of image analysis technology, and in particular to a method for tracking in-game advertisement visibility based on ray casting. Background Art

[0002] Ad visibility is an indicator that measures whether an ad is actually seen by users. It is used to measure the number of visible exposures on the user's screen. For brand ads that focus on exposure, ad visibility determines whether the ad exposure is effective and is a prerequisite for generating brand effects.

[0003] At present, the Chinese patent with the existing patent application number "CN202310382475.0" discloses an advertising exposure analysis method and related devices. The method can obtain multiple indicators that affect the exposure in the advertisement, find the K indicators that have the greatest impact on the exposure from these multiple indicators, and modify the plane layout of the advertisement when these K indicators include the plane layout indicator. The exposure of the modified advertisement is greater than the exposure before the modification. It can be seen that the method introduces the relevant indicators involved in the plane layout of the advertisement to analyze the exposure of the advertisement, making the analysis of the exposure of the advertisement more complete and accurate, and when the indicators involved in the plane layout are included in the indicators that have the greatest impact on the exposure, the plane layout is modified in time to help advertisers increase the exposure of the advertisement as much as possible. Although it plays a certain role in facilitating the exposure of the advertisement, both the advertisement itself and the image to be occupied have certain colors, which will lead to unclear color distinction or color conflict when the advertisement is placed.

[0004] Single advertising location: The existing advertising exposure methods mainly focus on the duration and volume of advertising, but ignore the most important visibility, resulting in many ads reaching the specified volume and time, but not being effectively "seen" by consumers. For example, when browsing the web, we often see pop-up ads in the lower right corner of the web page. Although they pop up every time the web page is opened, they are located at the very edge of the web page, while the best observation area is in the center of the web page, which makes it impossible for consumers to effectively observe the ads. On the contrary, if the pop-up ads are directly placed in the center of the web page, although the ads are placed in the best observation area for consumers, since the main purpose of consumers logging in to the web page is not because of the pop-up ads, the ads will be closed as soon as they pop up, which will also affect consumers' comfort in using the web page. Summary of the invention

[0005] 1. Technical issues to be resolved

[0006] In view of the deficiencies of the prior art, the present invention provides an in-game advertisement visibility tracking method based on ray casting, which solves the problems raised in the background technology.

[0007] (II) Technical solution

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0009] The in-game ad visibility tracking method based on ray casting. The specific steps of the tracking method are as follows:

[0010] S1. Use a four-point measurement method to detect the image scaling ratio and projection image distance, and determine the flatness of the projected image based on whether the sizes of each square pixel block after projection are consistent. The four-point measurement method is to insert a square pixel block at each of the four corners of the projected image, and use image recognition technology to collect image data of the square pixel blocks after projection;

[0011] S2. Collect pixel points in the projection image during the current game projection duration, and divide them into low change points and high change points according to the change rate of the pixel points. The area covered by the low change points is recorded as the projection area. Conversely, the area covered by the high change points is recorded as the non-projection area.

[0012] S3. Compare the color difference comparison value between the image to be advertised and the projection area with the adjustment threshold interval, and adjust the corresponding projection adjustment strategy according to the comparison result. The adjustment threshold interval includes an adjustment upper limit and an adjustment lower limit. When the adjustment lower limit ≤ the color difference comparison value ≤ the adjustment upper limit, no response is made, and the advertisement image corresponding to the color difference comparison value is recorded as an advertisement that can be directly projected. On the contrary, when the color difference comparison value > the adjustment upper limit or the adjustment lower limit > the color difference comparison value, a color difference correction instruction is issued, and the correction strategy is executed. After adjustment, the color difference comparison value is recalculated and compared with the adjustment threshold interval again. When the preset number of adjustments is reached, the advertisement image that is still not in the adjustment threshold area is recorded as an advertisement that cannot be projected.

[0013] S4, establishing a two-dimensional coordinate system in the projected image, and inputting the original size of the projected advertisement into the two-dimensional coordinate system to form a projection area frame, then obtaining all projection units that meet the projection area frame size requirements in the projection image, calculating the proportion of the projectable screen area blocked by the advertisement image to the advertisement image, and comparing the occupancy ratio with a preset qualified threshold of the proportion, when the proportion of the projectable screen area is less than the qualified threshold of the proportion, an unqualified instruction is issued, a scaling strategy is executed, the projection advertisement size is scaled, and re-detection is performed; otherwise, a qualified instruction is issued;

[0014] S5. After receiving the qualified instruction, the visibility evaluation value of each projection unit in the projection unit set that meets the projection requirements is calculated based on the color difference adaptation value between the projectable screen area and the image to be advertised, the projection units are sorted according to the size of the visibility evaluation value, and the advertisements are placed in order.

[0015] Furthermore, in step S1, the zoom ratio and projection image distance of the image are detected by using the four-point measurement method. The specific process is as follows:

[0016] Set the side length of the initial square pixel block, denoted as ;

[0017] The image after the pixel block projection is identified by image recognition technology, and the side length of the square pixel block after projection is obtained, which is recorded as ;

[0018] By analyzing the formula Calculate the scaling factor of a pixel block and projection image distance , where Indicates the object distance between the display chip and the lens;

[0019] Similarly, the scaling ratio and projection image distance of each square pixel block after projection are obtained and compared with each other. When the scaling ratio or projection image distance of the square pixel block after projection are inconsistent, it is recorded as an uneven projection image; conversely, when the scaling ratio and projection image distance of the square pixel block after projection are consistent, it is recorded as a flat projection image.

[0020] Furthermore, in step S2, the analysis process of the pixel change rate is as follows:

[0021] Set the color difference response threshold, total detection time, and detection interval time, respectively. , , ;

[0022] by The acquisition detection interval of the image is the length of time. The color value of each pixel in the projected image is collected, and the color difference between the color values ​​of two adjacent pixels is compared with the color difference response threshold. When the color difference of the same pixel is greater than or equal to the color difference response threshold, When , the pixel point is recorded as having changed, otherwise, the pixel point is recorded as having not changed;

[0023] By analyzing the formula Calculate the color difference change rate of the pixel , where Indicates the total detection time The number of times the inner pixel changes;

[0024] By analyzing the formula Calculate the classification threshold of the pixel , where , Represent the color difference response threshold and the weight of the number of image detections, respectively, and , e represents a natural constant;

[0025] To get the classification threshold of the pixel As the dividing line, the color difference change rate The corresponding pixels are classified. ≤ When , the color difference change rate The corresponding pixel points are recorded as low change points, otherwise, they are recorded as high change points.

[0026] Furthermore, in step S3, the color difference comparison value analysis process between the image to be advertised and the projection area is as follows:

[0027] Set the color value selection distance D of the advertising image, obtain the color value of the advertising image pixel within the range D from the edge of the advertising image, and calculate the average color value of the edge of the advertising image, which are recorded as , , ;

[0028] Get the color value of the pixel in the current projectable area, recorded as , , ;

[0029] By analyzing the formula Calculate the color difference comparison value of the mth advertisement image with edge distance D , where , , represent the average R value, average G value, and average B value of the mth advertisement image, respectively. , , Represents the R value, G value, and B value of the i-th pixel in the projectable area. , , Respectively represent the weights corresponding to the color value R value, G value, and B value of the color difference pixel point, and , , ;

[0030] According to the color difference comparison value The values ​​are compared with the adjustment threshold interval in order from large to small to obtain advertisements that can be projected directly.

[0031] Furthermore, in step S3, the calculation formulas for adjusting the upper limit and the lower limit are as follows:

[0032]

[0033] In the formula, Indicates the upper limit of the adjustment threshold range. Indicates the lower limit of the adjustment threshold range. Indicates the maximum R value of the pixel point in the projectable area. Indicates the maximum G value of the pixel point in the projectable area. Indicates the maximum B value of the pixel in the projectable area. Indicates the minimum R value of the pixel point in the projectable area. Indicates the minimum G value of the pixel point in the projectable area. Indicates the minimum B value of the pixel in the projectable area. , They respectively represent the upper limit adjustment ratio coefficient and the lower limit adjustment ratio coefficient of the color value of the projectable area, and , .

[0034] Further, in step S3, when the color difference comparison value>the adjustment upper limit or the adjustment lower limit>the color difference comparison value, a color difference correction instruction is issued and the correction strategy is executed. The specific analysis process of executing the correction strategy is as follows:

[0035] Get the average color value in the projection area;

[0036] By analyzing the formula Calculate the average color value of the mth advertising image after adjustment , , , where , , Respectively represent the preset adjustment ratios of the R value, G value, and B value in the average color value of the advertisement image;

[0037] Will , , Import color difference comparison values The calculation formula is used and compared with the adjustment threshold interval again until the preset adjustment times are reached, and the advertising images that are still not within the adjustment threshold area are recorded as non-projectable advertisements.

[0038] Further, in step S4, the specific process of obtaining the frame size that meets the projection area is as follows:

[0039] Establish a two-dimensional coordinate system, with the endpoint of the lower left corner of the projected image as the coordinate origin, the direction to the right of the origin as the X axis, and the direction upward from the origin as the Y axis;

[0040] Get the length and width of the projected image and the original length and width of the projected advertisement, and record them as , , , ;

[0041] Set the lower left corner of the ad image as the translation coordinate point ( , ), insert the advertising image into the projection image, and obtain the translation coordinate points of the advertising image ( , )exist ,and All images within the range are recorded as projection units;

[0042] The qualified threshold of proportion and the number of zooming times are set. When the proportion of the screen-projectable area in the advertising image in the projection unit is greater than the qualified threshold of proportion, it is recorded as an image that meets the projection requirements; otherwise, the size of all projected advertising images is scaled according to the preset ratio, and the proportion of the advertising image in the projection unit to the screen-projectable area is re-obtained until the occupancy ratio is greater than the qualified threshold of proportion or the number of zooming times is reached. In this case, the image with a proportion ≤ the qualified threshold of proportion after the number of zooming times is reached is recorded as an unprojectable advertisement.

[0043] Furthermore, in step S5, the calculation formula of the color difference adaptation value is as follows:

[0044]

[0045] In the formula, represents the color difference adaptation value of the mth advertising image, represents the color difference comparison value of the mth advertising image as a whole, , They respectively represent the weights of the color difference comparison value of the mth advertisement image with an edge distance D and the color difference comparison value of the entire advertisement image, and e represents a natural constant.

[0046] Furthermore, in step S5, the calculation formula of the visibility evaluation value is as follows:

[0047]

[0048] In the formula, represents the visibility evaluation value of the mth advertisement image, Indicates the total duration of the mth advertisement. , Respectively represent the length and width of the mth advertising image, represents the distance from the mth advertisement image to the center point of the projection image, , , They respectively represent the weights of the total delivery duration, advertisement size, and the distance from the advertisement image to the center point of the projected image, and e represents a natural constant.

[0049] Further, in step S5, the order of advertisement delivery: setting the advertisement delivery ranking gradient value C, obtaining the advertisement delivery ranking gradient interval [0, C), [C, 2C), [2C, 3C), ..., [n-1C, nC); the visibility evaluation value of each advertisement image is calculated based on the advertisement delivery ranking gradient interval Classify the advertisements, and place the advertisements in the same advertisement ranking gradient interval according to the remaining proportions from large to small;

[0050] The remaining delivery ratio refers to the ratio of the remaining duration of each delivered advertisement image to the total duration required for delivery.

[0051] (III) Beneficial effects

[0052] The present invention provides a method for tracking in-game advertisement visibility based on ray casting, which has the following beneficial effects:

[0053] By utilizing the rate of change of each pixel in the projected image within a specific time during game projection, the image is divided into a non-projectable area and a projectable area. In this way, the area with the least impact on consumers is determined before advertising is projected, thereby facilitating the subsequent placement of advertisements in the best viewing position for consumers as much as possible without affecting the consumer's gaming experience, thereby providing a basis for improving the observability of advertisements. In addition, by obtaining the color difference comparison value between the image of the advertisement to be placed and the projectable area, the degree of color conflict between the advertisement to be placed and the current projectable area is understood, and then compared with the adjustment threshold interval corresponding to the current projectable area to obtain advertisements suitable for the current projected game scene. In this way, while ensuring the clear display of the advertisement image, the color conflict between the advertisement image and the game projection image is reduced, thereby further improving the observability of the advertisement image without affecting the normal observation of consumers.

[0054] When the color difference comparison value between the image to be advertised and the projectable area cannot meet the needs of direct delivery, a correction strategy is adopted to adjust the color value of the advertising image toward the color value of the projectable area, so that the advertising image meets the needs of direct projection, thereby improving the availability and compliance rate of the advertising image.

[0055] All projection units that meet the projection area border size requirements are obtained in the projection image, and then the preferred projection position is obtained according to the proportion of the projection image to the projectable screen area, so as to place the advertisement in the best viewing position for the consumer as much as possible without affecting the consumer's gaming experience. In addition, when the size of the advertisement image cannot meet the demand, the projection needs are met by scaling the projection advertisement size, thereby improving the applicability of the advertisement image and increasing the effective exposure time of the advertisement image.

[0056] By combining the color difference adaptation value between the projectable screen area and the image to be advertised, a visibility evaluation value is calculated to intuitively evaluate whether the advertisement is effectively observed, so as to ensure that the advertisement image is truly and effectively observed according to the visibility evaluation value, and the advertisement images are graded according to the gradient interval of the advertisement ranking. Then, the advertisements located in the same gradient interval of the advertisement ranking are placed in descending order according to the remaining proportion of the placement, so as to meet the time requirements of different advertisement projections and avoid the situation where the same advertisement image is placed for a long time. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a flowchart of the overall structure of the present invention;

[0058] Figure 2 It is a schematic diagram of the detection by the four-point determination method of the present invention;

[0059] Figure 3 It is a schematic diagram of obtaining the frame size of the projection area in accordance with the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] Projection flatness analysis

[0062] In order to ensure the flatness of the projected image and avoid the unevenness of the projection screen surface, and also to avoid affecting subsequent calculations and tests, it is necessary to test the flatness of the projected image. The flatness of the image is affected by many factors, such as the tilt angle of the projection device, the scaling ratio, etc. The game projection device we developed has a camera built in to facilitate the acquisition of images in the projection state. In addition to obtaining the images required for subsequent tests, it can also be connected to the background for remote debugging and guidance. Based on the camera, we use the four-point measurement method to detect the image condition;

[0063] Four-point measurement method: insert a square pixel block at each of the four corners of the projected image, and use image recognition technology to collect the image data of the square pixel block after projection, such as Figure 2 As shown, the side length of an initial square pixel block is set in the image to be projected, and then the image after projection is recognized by using image recognition technology, and the projected image is obtained by shooting with a camera on the projection device. and projection image distance , obtained using the following formula:

[0064]

[0065] In the formula, represents the side length of the initial square pixel block, represents the side length of the initial square pixel block after projection, Indicates the object distance between the display chip and the lens.

[0066] When the projected image is stable, the size of each part of the image will be scaled proportionally after the image is projected. Therefore, by obtaining the size change of a square pixel block before and after scaling, the scaling ratio of the projection device can be obtained. Similarly, the ratio between the image distance and the object distance is equal to the scaling ratio, that is, , thereby obtaining the distance from the projection device to the location of the projected image.

[0067] Getting the image scale and projection image distance After that, the scaling ratio and projection image distance of each square pixel block after projection are obtained and compared with each other. When the scaling ratio or projection image distance of the square pixel block after projection is inconsistent, it is recorded as an uneven projection image; on the contrary, when the scaling ratio and projection image distance of the square pixel block after projection are consistent, it is recorded as a flat projection image. That is, Figure 2 As shown, when the scaling ratio or projection image distance of the square pixel block in the upper left corner is consistent with that of the square pixel block in the lower left corner, the projected image is flat from top to bottom. Similarly, when the scaling ratio and projection image distance of the square pixel blocks after projection are consistent, it is only necessary to detect whether the projected sizes are consistent to judge the flatness. At the same time, in order to avoid the deviation caused by image tilt, the projected image can be directly shot to compare the sizes of the square pixel blocks after projection, thereby ignoring the impact of projection tilt.

[0068] Projection area selection

[0069] When projecting a game, the center of the projected graphics is the main range of consumer activities (i.e., the game area). At the same time, the scene in the game is also mainly reflected in the position. The image in this area will change continuously with the control of the game, so the pixel points at this position will change greatly in a short time. The farther away from this area, the less impact the image has on the game. In addition, in order to reduce memory and capacity, games often use modular images outside the game area. The most common one is the tile system of Super Mario, which divides the game scene into multiple small blocks (tiles), each tile is 8×8 pixels in size. This reduces the actual storage requirements of each scene element in the memory. When needed, the game scene can be built through these tiles. During the running of the game, the background does not change much, and the range that the game can affect is very small. This can be used as the game projection area. Based on this, we can determine the area required by the game by detecting the changes in each pixel in the image, and the area not required by the game is the area that can be projected.

[0070] In order to detect the corresponding situation of the image, it is necessary to set the color difference response threshold before detection. , Total detection time And the detection interval ,by The acquisition detection interval of the image is the length of time. The color value of each pixel in the projected image is collected, and the color difference between the color values ​​of two adjacent pixels is compared with the color difference response threshold. When the color difference of the same pixel is greater than or equal to the color difference response threshold, If , the pixel point is recorded as having changed once, otherwise, the pixel point is recorded as not having changed, so that the number of times the pixel point has changed within a specified time can be obtained.

[0071]

[0072] In the formula, Indicates the rate of change of color difference of pixel points. Indicates the total detection time The number of times the inner pixel changes can determine the game area and the non-game area, which provides a basis for placing the advertisement in the best viewing position for consumers as much as possible without affecting the consumer's gaming experience, so as to improve the observability of the advertisement. Since the changes of different game images are different, a classification threshold that can change with the game image is required. The classification threshold of the pixel is based on the following formula:

[0073]

[0074] In the formula, Represents the classification threshold of the pixel, , Represent the color difference response threshold and the weight of the number of image detections, respectively, and , e represents a natural constant.

[0075] To get the classification threshold of the pixel As the dividing line, the color difference change rate The corresponding pixels are classified. ≤ When , the color difference change rate The corresponding pixel points are recorded as low change points, and vice versa, they are recorded as high change points. The area covered by the low change points is recorded as the projection area, and vice versa, the area covered by the high change points is recorded as the non-projection area.

[0076] Advertising image color difference analysis

[0077] Both the advertisement itself and the image to be occupied have certain colors, which will lead to unclear color distinction or color conflict when the advertisement is placed. Taking the web search interface as an example, the web search interface defaults to white, so if the main color of the pop-up advertisement is also white or light-colored, it cannot be observed by consumers at the first time. On the contrary, when the operation interface is set to black, the advertisement with white as the main color is very obvious. In order to avoid this problem, it is necessary to select an image that is compatible with the color of the current screen-projectable area from many alternative advertising images. It is only necessary to perform a color difference adaptability test between the screen-projectable area and the image to be advertised, and then according to the degree of color difference comparison between the two, it can be understood whether the advertising image is suitable for insertion into the game.

[0078] Since the visibility of the advertising image depends on the color difference between the color of the advertising image border and the color of the screen projection area, rather than the color difference between the entire advertisement and the screen projection area, it is necessary to obtain the image of the advertising image border before comparison. By obtaining the color value of the advertising image pixel points within a specific distance range from the edge of the advertising image, and calculating the average color value of the advertising image edge, the average color value represents the color value of the advertising image. The specific distance range can be set in advance, and the specific distance range is set to D. The color value of the advertising image pixel points within a range of D from the edge of the advertising image is obtained, and the average color value of the advertising image edge is calculated, which are recorded as , , , the calculation formula of color difference comparison value is as follows:

[0079]

[0080] In the formula, represents the color difference comparison value of the mth advertisement image with edge distance D, Indicates the R value of the pixel point in the current projectable area. Indicates the G value of the pixel point in the current projectable area. Indicates the B value of the pixel in the current projectable area. represents the average R value of the mth advertising image, represents the average G value of the mth advertisement image, represents the average B value of the mth advertisement image, Represents the R value of the i-th pixel in the projectable area, Represents the G value of the i-th pixel in the projectable area, Represents the B value of the i-th pixel in the projectable area, , , Respectively represent the weights corresponding to the color value R value, G value, and B value of the color difference pixel point, and , , .

[0081] According to the color difference comparison value The values ​​are compared with the adjustment threshold interval in order from large to small, and the adjustment threshold interval includes an adjustment upper limit and an adjustment lower limit.

[0082] Since different projectable areas have different color values, their adaptive color ranges are also different. Therefore, it is necessary to find the corresponding adjustment threshold interval according to the specific color value in the projectable area. The calculation formula for the adjustment threshold interval is as follows:

[0083]

[0084] In the formula, Indicates the upper limit of the adjustment threshold range. Indicates the lower limit of the adjustment threshold range. Indicates the maximum R value of the pixel point in the projectable area. Indicates the maximum G value of the pixel point in the projectable area. Indicates the maximum B value of the pixel in the projectable area. Indicates the minimum R value of the pixel point in the projectable area. Indicates the minimum G value of the pixel point in the projectable area. Indicates the minimum B value of the pixel point in the projectable area. , They respectively represent the upper limit adjustment ratio coefficient and the lower limit adjustment ratio coefficient of the color value of the projectable area, and , .

[0085] When the adjustment lower limit ≤ color difference comparison value ≤ adjustment upper limit, no response is made, and the advertisement image corresponding to the color difference comparison value is recorded as an advertisement that can be directly projected, and the advertisement located in this area can be directly projected;

[0086] On the contrary, when the color difference comparison value is greater than the adjustment upper limit or the adjustment lower limit is greater than the color difference comparison value, a color difference correction instruction is issued and the correction strategy is executed.

[0087] Execute correction strategy: obtain the average color value in the projection area; calculate the average color value of the mth advertising image after adjustment , , The specific formula is as follows:

[0088]

[0089] In the formula, , , They respectively represent the preset adjustment ratios of the R value, G value, and B value in the average color value of the advertisement image.

[0090] Adjusted by the average color value of the mth advertising image For example, is the average R value of the mth advertising image. Taking this as the contact, the difference between it and the average R value in the projectable and screen-castable area is obtained, that is, , and then use the difference as the adjustment range and set the adjustment ratio , by multiplying the adjustment ratio by the adjustment range, we can get the required adjustment value, that is, , and then adjust the value by the average R value of the mth advertising image to obtain the parameter that makes the advertising image approach the average color value in the projection area.

[0091] For example, the average R value of the mth advertising image before adjustment is 100, and the average R value in the projection area is 150. The adjustment ratio of R is set to 10%. Therefore, the difference between the average R value of the mth advertising image before adjustment and the average R value in the projection area is 150-100=50. By multiplying the adjustment ratio by the adjustment range, the required adjustment value is obtained, that is, (100-150)×10%. Then, based on this value as the range, the average R value of the mth advertising image before adjustment is adjusted, that is, 100-(100-150)×10%=105.

[0092] Then , , Import color difference comparison values The calculation formula is used and compared with the adjustment threshold interval again until the preset adjustment times are reached. The advertising images that are still not in the adjustment threshold area are recorded as non-projectable advertisements. Of course, the adjustment times can also be set as needed. When the adjustment times are reached, the results are output and the adjustment stops.

[0093] Ad size analysis

[0094] In order to detect and adjust the size of the advertising image, such as Figure 2 As shown, a two-dimensional coordinate system is established in the projected image, and the original size of the projected advertisement is input into the two-dimensional coordinate system to form a projection area frame, and then all projection units that meet the projection area frame size requirements are obtained in the projection image. The specific process of obtaining the projection area frame size that meets the projection area frame size is as follows:

[0095] S101, establishing a two-dimensional coordinate system, with the endpoint of the lower left corner of the projected image as the coordinate origin, the direction to the right of the origin as the X axis, and the direction upward from the origin as the Y axis;

[0096] S102, retrieve the length and width of the projected image and the original length and width of the projected advertisement, and record them as , , , ;

[0097] S103, setting the lower left corner of the advertisement image as the translation coordinate point ( , ), insert the advertising image into the projection image, and obtain the translation coordinate points of the advertising image ( , )exist ,and All images within the range are recorded as projection units, such as Figure 3 As shown;

[0098] S104, setting a qualified proportion threshold Q and a zooming number S. When the proportion of the screen-projectable area in the advertising image in the projection unit is greater than the qualified proportion threshold, it is recorded as an image that meets the projection requirements. Otherwise, the size of the projected advertising image is scaled Fs according to the preset ratio, and the proportion of the advertising image in the screen-projectable area in the projection unit is re-obtained until the occupancy ratio is greater than the qualified proportion threshold or the zooming number S is reached. The image with a proportion ≤ the qualified proportion threshold after the zooming number is reached is recorded as an unprojectable advertisement.

[0099] For example, the projected image size (length × width) is 150mm × 100mm and the original size of the projected advertisement (length × width) is 10mm × 12mm. Then the coordinate point of the lower left corner of the advertisement image is translated ( , ) is the sampling point, and the ,and All images within the range are projection units.

[0100] Occupancy ratio: refers to the area of ​​the screen that can be projected within the area covered by the advertising image, as a percentage of the advertising image. For example, the area of ​​the screen that can be projected within the area covered by the advertising image in a projection unit is 8cm 2 , and the area of ​​the advertising image is 8.5cm 2 , the occupancy ratio is 8 / 8.5≈94%.

[0101] Set the qualified threshold of proportion to 98%, the number of zooms to 3 times, and the zoom ratio to 10%. Assume that during the first comparison, the proportion of the projectable area in the advertising image in the projection unit is 98.2%, that is, 98.2%> the set qualified threshold of proportion 98%, so the location of the advertising image meets the requirements and can be projected directly.

[0102] On the contrary, when the first comparison is made, the maximum value of the proportion of the projectable area in the advertising image in all projection units is 97.5%, that is, 97.5% < the set qualified threshold of 98%, so the position of the advertising image does not meet the requirements and needs to be scaled. The advertising image in the projection unit is reduced according to the set scaling ratio of 10%, and the size of the advertising image is 10mm×12mm=120mm, 120mm×(100%-10%)=108mm. Then, the advertising image is re-imported into the projection image, and the proportion of the projectable area in the advertising image is calculated until it meets the requirements or the advertising image is adjusted three times and the adjustment is stopped.

[0103] Comprehensive analysis of advertising visibility

[0104] After finding a suitable location for advertising, in order to effectively understand the observability of the advertisement, the color difference adaptation value between the projectable screen area and the image to be advertised is combined to calculate a visibility evaluation value that can intuitively evaluate whether the advertisement is effectively observed, so as to ensure that the advertisement image is truly and effectively observed based on the visibility evaluation value. The calculation formula of the color difference adaptation value is as follows:

[0105]

[0106] In the formula, represents the color difference adaptation value of the mth advertising image, represents the color difference comparison value of the mth advertising image as a whole, , They respectively represent the weights of the color difference comparison value of the mth advertisement image with an edge distance D and the color difference comparison value of the entire advertisement image, and e represents a natural constant.

[0107] Due to the chromatic aberration adaptation value The essence is an exponential function with a constant less than 1. Therefore, when the color difference comparison value is still When it increases, the color difference adaptation value This is because the color difference is too obvious and can easily cause eye discomfort to consumers. Therefore, an exponential function with a constant less than 1 is needed to represent it.

[0108]

[0109] In the formula, represents the visibility evaluation value of the mth advertisement image, Indicates the total duration of the mth advertisement. , Respectively represent the length and width of the mth advertising image, represents the distance from the mth advertisement image to the center point of the projection image, , , They respectively represent the weights of the total delivery duration, advertisement size, and the distance from the advertisement image to the center point of the projected image, and e represents a natural constant.

[0110] The distance from the advertising image to the center of the projected image , using image recognition technology, obtain the coordinates of the center point of the mth advertising image in the two-dimensional coordinate system ( , ) and the coordinates of the center point of the projected image ( , ),but Because the closer to the center of the projected image, the easier it is for consumers to observe. In addition, sufficient delivery time and the size ratio of the advertisement are the basic factors that determine whether the advertisement can be effectively observed.

[0111] Advertisement placement order: In order to avoid the situation where the same ad image is placed for a long time, the placement market needs to be taken into account in the ranking, and the visibility of the ad image also needs to be taken into account. For this purpose, a placement ad ranking gradient value C is set, and the placement ad ranking gradient interval is obtained [0, C), [C, 2C), [2C, 3C), ..., [n-1C, nC); the visibility evaluation value of each ad image is evaluated according to the placement ad ranking gradient interval Classify the advertisements, and place the advertisements in the same advertisement ranking gradient interval according to the remaining proportions from large to small;

[0112] The remaining delivery ratio refers to the ratio of the remaining duration of each delivered advertisement image to the total duration required for delivery.

[0113] Simply put, the advertising images are classified according to the advertising ranking gradient interval, and then the advertising images in the same advertising ranking gradient interval are placed according to the remaining proportion from large to small, so as to meet the duration requirements of different advertising projections. For example, after classification according to the advertising ranking gradient interval, Advertisement 1, Advertisement 2, and Advertisement 3 are in the same advertising ranking gradient interval, and then the remaining delivery time of each advertisement is detected. The remaining delivery time of Advertisement 1 is 10h (the set total duration is 20h), the remaining delivery time of Advertisement 2 is 12h (the set total duration is 30h), and the remaining delivery time of Advertisement 3 is 20h (the set total duration is 30h). Then the remaining delivery time ratio between Advertisement 1, Advertisement 2, and Advertisement 3 is , it can be seen that the remaining duration of Ad Three accounts for more than Ad Two or Ad One. Therefore, Ad Three is placed before Ad Two and Ad One, that is, it is played in the order of Ad Three, Ad One, and Ad Two. When the values ​​are the same, one is randomly selected as the priority ad. For example, when the remaining duration of Ad Three and Ad One is equal, one is randomly selected to play first, and the other is arranged after it. Assuming that Ad Three is randomly selected, the playback order is Ad Three, Ad One.

[0114] The weight coefficient is determined by the coefficient of variation method, which is a method of weighting each indicator according to the degree of variation between the current value of each evaluation indicator and the target value. If the numerical difference of an indicator is large, it can clearly distinguish the evaluated objects, indicating that the indicator has rich discrimination information, and thus the indicator should be given a larger weight. On the contrary, if the numerical difference of each evaluated object on a certain indicator is small, then the ability of this indicator to distinguish the evaluation objects is weak, and thus the indicator should be given a smaller weight. This method directly uses the information contained in each indicator and obtains the weight of the indicator through calculation, so it is objective.

[0115] In the application, the several formulas involved are all calculated by taking their numerical values ​​after removing the dimensions, and the formulas are established by collecting a large amount of data and performing software simulation to obtain a formula for the most recent actual situation. Some coefficients or weights in the formulas are set by technical personnel in this field according to actual conditions, so they will not be elaborated here.

[0116] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. A person of ordinary skill in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0117] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0118] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for tracking in-game advertising visibility based on ray casting, characterized in that: The specific steps of this tracking method are as follows: S1. Use a four-point measurement method to detect the image scaling ratio and projection image distance, and determine the flatness of the projected image based on whether the sizes of each square pixel block after projection are consistent. The four-point measurement method is to insert a square pixel block at each of the four corners of the projected image, and use image recognition technology to collect image data of the square pixel blocks after projection; S2. Collect pixel points in the projection image during the current game projection duration, and divide them into low change points and high change points according to the change rate of the pixel points. The area covered by the low change points is recorded as the projection area. Conversely, the area covered by the high change points is recorded as the non-projection area. S3. Compare the color difference comparison value between the image to be advertised and the projection area with the adjustment threshold interval, and adjust the corresponding projection adjustment strategy according to the comparison result. The adjustment threshold interval includes an adjustment upper limit and an adjustment lower limit. When the adjustment lower limit ≤ the color difference comparison value ≤ the adjustment upper limit, no response is made, and the advertisement image corresponding to the color difference comparison value is recorded as an advertisement that can be directly projected. On the contrary, when the color difference comparison value is greater than the upper adjustment limit or the lower adjustment limit is greater than the color difference comparison value, a color difference correction instruction is issued, and the correction strategy is executed. After adjustment, the color difference comparison value is recalculated and compared with the adjustment threshold interval again. When the preset adjustment times are reached, the advertising image that is still not within the adjustment threshold area is recorded as a non-projectable advertisement. S4, establishing a two-dimensional coordinate system in the projected image, and inputting the original size of the projected advertisement into the two-dimensional coordinate system to form a projection area frame, then obtaining all projection units that meet the projection area frame size requirements in the projection image, calculating the proportion of the projectable screen area blocked by the advertisement image to the advertisement image, and comparing the occupancy ratio with a preset qualified threshold of the proportion, when the proportion of the projectable screen area is less than the qualified threshold of the proportion, an unqualified instruction is issued, a scaling strategy is executed, the projection advertisement size is scaled, and re-detection is performed; otherwise, a qualified instruction is issued; S5. After receiving the qualified instruction, the visibility evaluation value of each projection unit in the projection unit set that meets the projection requirements is calculated based on the color difference adaptation value between the projection screen area and the image to be advertised, the projection units are sorted according to the visibility evaluation value, and advertisements are placed in order; The calculation formula of the color difference adaptation value is as follows: ； In the formula, represents the color difference adaptation value of the mth advertising image, represents the color difference comparison value of the mth advertising image as a whole, , denote the weights of the color difference comparison value of the mth advertisement image with edge distance D and the color difference comparison value of the entire advertisement image, respectively, and e denotes a natural constant; The calculation formula of visibility evaluation value is as follows: ； In the formula, represents the visibility evaluation value of the mth advertisement image, Indicates the total duration of the mth advertisement. , Respectively represent the length and width of the mth advertising image, , Respectively represent the length and width of the projected image, represents the distance from the mth advertisement image to the center point of the projection image, , , They represent the weights of the total duration of the advertisement, the size of the advertisement, and the distance from the center of the image advertisement to the projected image, respectively, and e represents a natural constant; Advertisement placement order: Set the advertisement placement ranking gradient value C, and obtain the advertisement placement ranking gradient interval [0, C), [C, 2C), [2C, 3C), ..., [n-1C, nC); evaluate the visibility of each advertisement image based on the advertisement placement ranking gradient interval Classify the advertisements, and place the advertisements in the same advertisement ranking gradient interval according to the remaining proportions from large to small; The remaining delivery ratio refers to the ratio of the remaining duration of each delivered advertisement image to the total duration required for delivery.

2. The in-game advertisement visibility tracking method based on ray casting according to claim 1, characterized in that: In step S1, the four-point measurement method is used to detect the image scaling ratio and projection image distance. The specific process is as follows: Set the side length of the initial square pixel block, denoted as ; The image after the pixel block projection is identified by image recognition technology, and the side length of the square pixel block after projection is obtained, which is recorded as ; By analyzing the formula Calculate the scaling factor of a pixel block and projection image distance , where Indicates the object distance between the display chip and the lens; Similarly, the scaling ratio and projection image distance of each square pixel block after projection are obtained and compared with each other. When the scaling ratio or projection image distance of the square pixel block after projection are inconsistent, it is recorded as an uneven projection image; conversely, when the scaling ratio and projection image distance of the square pixel block after projection are consistent, it is recorded as a flat projection image.

3. The in-game advertisement visibility tracking method based on ray casting according to claim 1, characterized in that: In step S2, the analysis process of the pixel change rate is as follows: Set the color difference response threshold, total detection time, and detection interval time, respectively. , , ; by The acquisition detection interval of the image is the length of time. The color value of each pixel in the projected image is collected, and the color difference between the color values ​​of two adjacent pixels is compared with the color difference response threshold. When the color difference of the same pixel is greater than or equal to the color difference response threshold, When , the pixel point is recorded as having changed, otherwise, the pixel point is recorded as having not changed; By analyzing the formula Calculate the color difference change rate of the pixel , where Indicates the total detection time The number of times the inner pixel changes; By analyzing the formula Calculate the classification threshold of the pixel , where , Represent the color difference response threshold and the weight of the number of image detections, respectively, and , e represents a natural constant; To get the classification threshold of the pixel As the dividing line, the color difference change rate The corresponding pixels are classified. When , the color difference change rate The corresponding pixel points are recorded as low change points, otherwise, they are recorded as high change points.

4. The in-game advertisement visibility tracking method based on ray casting according to claim 3, characterized in that: In step S3, the color difference comparison value analysis process between the image to be advertised and the projection area is as follows: Set the color value selection distance D of the advertising image, obtain the color value of the advertising image pixel within the range D from the edge of the advertising image, and calculate the average color value of the edge of the advertising image, which are recorded as , , ; Get the color value of the pixel in the current projectable area, recorded as , , ; By analyzing the formula Calculate the color difference comparison value of the mth advertisement image with edge distance D , where , , represent the average R value, average G value, and average B value of the mth advertisement image, respectively. , , Represents the R value, G value, and B value of the i-th pixel in the projectable area. , , Respectively represent the weights corresponding to the color value R value, G value, and B value of the color difference pixel point, and , , ; According to the color difference comparison value The values ​​are compared with the adjustment threshold interval in order from large to small to obtain advertisements that can be projected directly.

5. The in-game advertisement visibility tracking method based on ray casting according to claim 4, characterized in that: In step S3, the calculation formulas for adjusting the upper limit and the lower limit are as follows: ； In the formula, Indicates the upper limit of the adjustment threshold range. Indicates the lower limit of the adjustment threshold range. Indicates the maximum R value of the pixel point in the projectable area. Indicates the maximum G value of the pixel point in the projectable area. Indicates the maximum B value of the pixel in the projectable area. Indicates the minimum R value of the pixel point in the projectable area. Indicates the minimum G value of the pixel point in the projectable area. Indicates the minimum B value of the pixel point in the projectable area. , They respectively represent the upper limit adjustment ratio coefficient and the lower limit adjustment ratio coefficient of the color value of the projectable area, and , .

6. The in-game advertisement visibility tracking method based on ray casting according to claim 5, characterized in that: In step S3, when the color difference comparison value>the adjustment upper limit or the adjustment lower limit>the color difference comparison value, a color difference correction instruction is issued and the correction strategy is executed. The specific analysis process of executing the correction strategy is as follows: Get the average color value in the projection area; By analyzing the formula Calculate the average color value of the mth advertising image after adjustment , , , where , , Respectively represent the preset adjustment ratios of the R value, G value, and B value in the average color value of the advertisement image; Will , , Import color difference comparison values The calculation formula is used and compared with the adjustment threshold interval again until the preset adjustment times are reached, and the advertising images that are still not within the adjustment threshold area are recorded as non-projectable advertisements.

7. The in-game advertisement visibility tracking method based on ray casting according to claim 6, characterized in that: In step S4, the specific process of obtaining the frame size that meets the projection area is as follows: Establish a two-dimensional coordinate system, with the endpoint of the lower left corner of the projected image as the coordinate origin, the direction to the right of the origin as the X axis, and the direction upward from the origin as the Y axis; Retrieve the original length and width of the projection advertisement and record them as , ; Set the lower left corner of the ad image as the translation coordinate point , insert the advertising image into the projection image, and obtain the translation coordinate points of the advertising image exist ,and All images within the range are recorded as projection units; The qualified threshold of proportion and the number of zooming times are set. When the proportion of the screen-projectable area in the advertising image in the projection unit is greater than the qualified threshold of proportion, it is recorded as an image that meets the projection requirements; otherwise, the size of all projected advertising images is scaled according to the preset ratio, and the proportion of the advertising image in the projection unit to the screen-projectable area is re-obtained until the occupancy ratio is greater than the qualified threshold of proportion or the number of zooming times is reached. In this case, the image with a proportion ≤ the qualified threshold of proportion after the number of zooming times is reached is recorded as an unprojectable advertisement.

Images