A live game activity safety judgment point generation method and system

Abstract

The application discloses a kind of live game activity safety determination point generation method and system, method includes: obtaining human body image and part marking line, the part marking line is the marking line passing through human body image;Identify the several marking line segments that part marking line is overlapped with human body image, at least one marking line segment is marked with the part label of at least one corresponding human body part area;According to the part label of the marking line segment, at least one point on at least one marking line segment is obtained as activity safety determination point.The application can find suitable activity safety determination point accurately and quickly according to the actual game scene needs, is not disturbed by complex human body action, when live game is carried out, both avoid the picture that does not meet game plot appears, and reduce the bondage of user too strong.

Description

Technical Field

[0001] This invention belongs to the field of image processing technology, and in particular relates to a method and system for generating security judgment points for live-action game activities. Background Technology

[0002] Motion-sensing games are already quite popular. Current technology can only obtain human motion parameters by filming actual human bodies and then adjust the movement and actions of the virtual characters displayed in the game screen accordingly. If filmed real human bodies are directly displayed in the game screen instead of virtual characters, users can see their own real images and real movements blend well with the game plot, which will greatly enhance the fun and sense of accomplishment of the game.

[0003] In traditional games, the range of movement and actions of virtual characters are controlled by the game software to conform to the needs of the game's plot. However, when real human images are placed into the game screen, since the movements and actions of real people are not controlled by the game software, this can lead to visual effects that are inconsistent with the game's plot or feel uncomfortable. For example, if a user is driving a convertible in the game, and the user moves to the left relative to the camera in real life, the user's image may appear outside the convertible in the game screen, creating an uncomfortable visual effect that does not fit the game's plot. Therefore, it is necessary to limit the user's range of movement to avoid situations that do not conform to the game's plot. Limiting a user's movement to a single area within a human body image can create a strong sense of constraint and negatively impact the user experience. For example, the upper limbs, which are the most frequently used part of the body, have a wide range of motion. In the convertible driving example, including the entire upper limb area within the movement area would require either an excessively large width (i.e., a very wide lateral safety range), affecting the overall game visuals, or severely restricting the user's movements. Furthermore, in actual gameplay, limbs, such as the upper limbs, may need to perform actions like hitting objects; overly restrictive controls would prevent completion. In reality, because it's a convertible, the upper limbs extending beyond the convertible's obstruction wouldn't cause visual discomfort. The key is to limit the torso and other body parts from extending beyond the obstruction. Therefore, specific points within specific body parts need to be identified as safety markers to limit the movement area. This avoids visually incongruous scenes and reduces excessive restriction on the user. However, due to the complexity of human movement, there is currently a lack of mature technology for obtaining such specific points. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and system for generating safety judgment points for live-action game activities, which addresses the shortcomings of the prior art. The generated safety judgment points can solve the problem of excessive restraint for users caused by the setting of safety zones in live-action games. The present invention can also be used in fields such as fun selfies and fitness.

[0005] The first aspect of this invention discloses a method for generating safety judgment points in live-action game activities, comprising the following steps:

[0006] Acquire a human body image and part marking lines, wherein the part marking lines are marking lines that pass through the human body image;

[0007] Identify a number of marker line segments that overlap with a human body image on the body part marking line, wherein at least one of the marker line segments is marked with at least one body part label corresponding to a human body part region;

[0008] Based on the location label of the marked line segment, at least one point on at least one of the marked line segments is identified as an activity safety determination point.

[0009] The above-mentioned method for generating safety judgment points for live-action game activities uses the two endpoints of the obtained marked line segment as activity safety judgment points.

[0010] The aforementioned method for generating safety judgment points for live-action game activities includes, in the case of, one or more of the following human body parts: the human torso, the left arm, the right arm, the left leg, and the right leg.

[0011] The above-mentioned method for generating safety judgment points for live-action game activities includes the following steps for labeling the marked line segments:

[0012] Step a: Generate lines for body parts from the human image based on human feature points:

[0013] Generate the outline of the left arm of the human body based on at least two feature points of the left arm.

[0014] And / or generate the part line of the right arm of the human body based on at least two feature points of the right arm;

[0015] And / or generate the part line of the left leg of the human body based on at least two feature points of the left leg;

[0016] And / or generate the part line of the right leg of the human body based on at least two feature points of the right leg;

[0017] And / or generate part lines of the human torso based on at least two feature points of the human torso, or generate part lines of the human torso based on the vertical midline of the human body.

[0018] Step b: Label the body parts according to the intersection of the marked line segments and the lines of the body parts:

[0019] When the marked line segment intersects with the line representing the left arm of the human body, label the marked line segment with a left arm tag;

[0020] When the marked line segment intersects with the line segment of the right arm of the human body, mark the marked line segment with the right arm label;

[0021] When the marked line segment intersects with the line segment representing the left leg of the human body, label the marked line segment with the left leg tag;

[0022] When the marked line segment intersects with the line segment representing the right leg of the human body, label the marked line segment with the right leg label;

[0023] When a marked line segment intersects with a line segment of the human torso, mark that line segment with a torso label.

[0024] The above-mentioned method for generating safety judgment points for live-action game activities includes the following steps for labeling the marked line segments:

[0025] According to the length and / or position of the marked line segment, mark the body part label; the body part label is a left arm label, right arm label, left leg label, right leg label, or torso label.

[0026] The above-mentioned method for generating safety judgment points for live-action game activities obtains the two endpoints of the marked line segment with the torso label as the activity safety judgment points.

[0027] The above-mentioned method for generating safety judgment points for live-action game activities uses the left endpoint of the marked line segment with the left leg label and the right endpoint of the marked line segment with the right leg label as the activity safety judgment points.

[0028] The above-mentioned method for generating safety judgment points for live-action game activities combines the human body image and the game screen, obtains a human body cut-off line based on the human body occlusion props in the game screen, so that the human body image below the human body cut-off line is not displayed in the game screen; and obtains part marking lines based on the human body cut-off line.

[0029] A second aspect of the present invention also discloses a system for generating safety judgment points for live-action game activities, comprising a first acquisition module, an identification module, and a second acquisition module.

[0030] The first acquisition module is used to acquire a human body image and part marking lines; the part marking lines are marking lines that pass through the human body image;

[0031] The recognition module is used to identify several marking line segments that overlap with the human body image on the part marking line, and at least one of the marking line segments is marked with at least one part label corresponding to a human body part area;

[0032] The second acquisition module is used to acquire at least one point on at least one of the marked line segments as an activity safety determination point based on the location label of the marked line segment.

[0033] Furthermore, when the first acquisition module acquires the part marking line, it synthesizes the human body image and the game screen, acquires the human body cut-off line based on the human body occlusion prop in the game screen, so that the human body image below the human body cut-off line is not displayed in the game screen; and acquires the part marking line based on the human body cut-off line.

[0034] The third aspect of this invention also discloses a method for generating an activity safety zone in a live-action game, which executes the method for generating activity safety judgment points in a live-action game as described in the first aspect, generates activity safety judgment points, and generates an activity safety zone based on the activity safety judgment points.

[0035] The above-mentioned method for generating the safe zone in live-action games, based on the safe zone determination point, includes the following steps: obtaining the marked line segment where the safe zone determination point is located, and generating the safe zone based on the marked line segment.

[0036] A fourth aspect of this invention also discloses a method for generating a safe zone in a live-action game video, comprising the following steps:

[0037] Using the method for generating active safe zones in live-action games as described in the third aspect, the marked line segments and active safe zones on the human body image of the reference frame are obtained;

[0038] Obtain the marked line segment where the activity safety judgment point is located on the current frame human image. Based on the ratio of the marked line segment where the activity safety judgment point is located on the current frame human image to the marked line segment where the activity safety judgment point is located on the reference frame human image, adjust the size of the activity safety area on the reference frame human image, and generate the activity safety area on the current frame human image.

[0039] Compared with the prior art, the present invention has the following advantages: The present invention can accurately and quickly find the appropriate activity safety judgment point according to the actual game scenario needs, without being interfered with by complex human body movements. When playing real-life games, it avoids the appearance of scenes that do not conform to the game plot and reduces excessive restraint on users.

[0040] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0041] Figure 1 This is a flowchart of the method in Example 1.

[0042] Figure 2 This is a schematic diagram showing the composite of a human body image and a game screen.

[0043] Figure 3 This is a schematic diagram showing the marked line segments corresponding to different parts and regions of the human body.

[0044] Figure 4 This is a diagram showing the marked line segments where the torso and left arm intersect.

[0045] Figure 5 This is a comparative illustration showing whether a human body is within the safe zone in a transparent convertible car game.

[0046] Figure 6 This is a comparison illustration of whether a person in a convertible car game is within the safe zone. Figure 1 .

[0047] Figure 7 This is a comparison illustration of whether a person in a convertible car game is within the safe zone. Figure 2 .

[0048] Figure 8 This is a schematic diagram of the lines of parts of the human body.

[0049] Figure 9 This is a system module diagram of Example 2. Detailed Implementation

[0050] Example 1

[0051] like Figure 1 As shown, a method for generating safety judgment points in live-action game activities includes the following steps:

[0052] Step 1: Obtain the human body image and body part marking lines, wherein the body part marking lines are marking lines that pass through the human body image.

[0053] like Figure 2 As shown, in practice, you can use the camera above the TV to take a picture of yourself to obtain a human image. The human image is then combined with the game screen, and the user can observe the combined effect of their own human image and the game screen on the TV screen. It should be noted that at the start of the game, the human image on the user's screen can be scaled, moved, etc., to combine it with the game screen and human body occlusion props in the screen to meet the needs of the game plot, so as to ensure a good visual effect.

[0054] The process involves obtaining part marker lines that pass through a human body image. For example, based on a human body occlusion prop in a game screen, such as a convertible, part marker lines can be obtained by horizontally passing through the occlusion prop and corresponding to the torso. This ensures that the torso obscured by the convertible does not extend beyond the convertible's left and right boundaries, avoiding unpleasant visual effects. Simultaneously, it does not restrict the movement of the upper limbs, reducing the user's sense of constraint. It should be noted that the purpose of these part marker lines is to obtain appropriate activity safety judgment points by acquiring mutually separated marker line segments on these lines and marking them with corresponding part labels, based on the characteristics of human anatomy. These lines can be horizontal, diagonal, vertical, or curved, etc., depending on the specific situation. In terms of game scene selection, for example, when the human body is in an upright position in the game, the activity safety judgment points corresponding to the left and right safe movement zones of the human body can generally be obtained through horizontal part marking lines. When the human body is in an inclined position in the game, such as when the props and the human body rotate together in the game screen, the part marking lines can also be diagonal lines. For another example, when the human body is in an upright position in the game, the activity safety judgment points corresponding to the up and down safe movement zones of the human body can also be obtained through vertical part marking lines. In addition, lines perpendicular to the joint points of the human body parts (such as the line connecting the elbow joint point and the wrist joint point) can be taken to make it more convenient to obtain the marking line segments corresponding to the flexible human body parts (such as the upper limbs). The marking line segments refer to the continuous areas on the part marking lines that overlap with the human body image.

[0055] In this embodiment, it is assumed that the part marking line is a horizontal line, which is used to obtain the activity safety judgment point corresponding to the left and right movement safety zone of the human body.

[0056] It should be noted that the location marker lines can be one or more. For example, when the top of the convertible is a horizontal line, location marker lines can be obtained based on the horizontal line. Furthermore, the human image below the top of the convertible can be cropped out of the game screen, thus eliminating the need to restrict the range of motion of the cropped portion of the human image. This allows for the acquisition of activity safety judgment points based on marker segments on a single location marker line (such as the two endpoints of the torso marker segment), ensuring the reliability of the activity safety judgment points and further improving the accuracy of human movement judgments during gameplay. In terms of freedom, in this case, the convertible is usually an opaque obscuring prop; for example, when the convertible is a transparent obscuring prop according to another plot requirement of the game, it is not possible to crop the human image between the top and bottom of the convertible. In this case, multiple horizontal lines between the top and bottom of the convertible can be obtained. These multiple horizontal lines can be scanned line by line to generate multiple corresponding part marking lines. Then, based on the marking line segments on the multiple part marking lines (such as the two endpoints of the torso marking line segment), activity safety judgment points can be obtained to perform activity safety judgment on a larger range of human images.

[0057] Step 2: Identify several marker line segments that overlap with the human body image on the body part marking line, wherein at least one of the marker line segments is marked with at least one body part label corresponding to a human body part region, specifically as follows: Figure 3 As shown.

[0058] Due to the complexity of human movements and structure (e.g., different parts of the human body may intersect each other, and the edges of different parts of the human body are irregular curves), as well as the diversity of game plot design (e.g., the shapes and sizes of human body covering props are varied), it is extremely difficult to determine the safe zone by relying on existing technology to achieve a good connection between the edges of specific parts of the human body and human body covering props.

[0059] Based on the characteristics of human anatomy, the overlapping part of a part marking line and a single human part (such as the torso, upper limbs, lower limbs, etc.) must be a marking line segment. The overlapping part of a part marking line and multiple non-intersecting human parts is usually multiple discontinuous marking line segments. By using the marking line segments corresponding to specific part labels, the activity safety judgment points on specific human part areas that meet the needs of the game plot can be obtained simply and accurately.

[0060] For example, when a body part marking line crosses the upper body and the arms are spread out in a V-shape, there will be marking line segments overlapping with the torso, the left arm, and the right arm. One of these segments is used to mark the body part. In this embodiment, it is assumed that the segment overlapping the torso is used, and the body part label is the torso label. Of course, the segments overlapping the left and right arms can also be labeled, depending on the actual situation. It should be noted that when a body part marking line crosses the upper body but the arms are close to the torso, there is only one marking line segment. This segment can be labeled only with the torso label, or it can be labeled with the left arm, torso, and right arm labels. Because the arms are close to the torso, the activity safety judgment point will still be located at or near the torso, and the activity safety zone judgment will not cause excessive restraint on the user. Similarly, for example... Figure 4 As shown, when the upper arm and torso of a human body cross, and the crossing part is exactly on the location marking line, the activity safety judgment point will still be located on or near the torso because the horizontal length of the crossing part is usually relatively short.

[0061] It should be noted that for several marker line segments that overlap with the human body image, only the marker line segment corresponding to a certain human body part area can be labeled with a part label, such as only labeling the part label corresponding to the upper limb of the human body, and then obtaining activity safety judgment points on the other unlabeled marker line segments. Alternatively, the marker line segments corresponding to multiple human body parts areas can be labeled with part labels, such as labeling the part labels corresponding to the human torso, the left leg of the human body, and the right leg of the human body, and obtaining activity safety judgment points on the aforementioned marker line segments.

[0062] It should be noted that identifying several marker segments on the part marking line that overlap with the human body image is a common technique in this field. For example, identification based on the transparency features of pixels on the part marking line, or identification based on the intersection of the part marking line and the human body contour line, etc.

[0063] The human body parts typically include one or more of the following: the human torso, the left arm, the right arm, the left leg, and the right leg. The specific number of human body parts used depends on the situation. In this embodiment, the human body parts used are the human torso, the left leg, and the right leg.

[0064] Step 3: Based on the location label of the marked line segment, obtain at least one point on at least one of the marked line segments as the activity safety determination point.

[0065] Here, assuming there are 3 marked line segments, with the body part labels for the 3 marked line segments being the left arm label, torso label, and right arm label, and the human torso or part of the human torso is obscured by an obscuring prop, in actual gameplay, the left and right positions of the human torso need not change too much to avoid being exposed outside the obscuring prop, while the position of the arms is not restricted. In this case, at least one point on the marked line segment of the torso label is taken as the activity safety judgment point. Of course, in this game scenario, in step 2, only the action of marking the body part label can be performed based on the position of the human torso, thus directly obtaining the marked line segment required in step 3. However, in reality, when the user moves away from the camera, causing the left and right legs to move upwards in the game screen and also be obscured by the obscuring prop, then left leg labels and right leg labels are needed. This can be referred to as... Figure 5 , Figure 6 and Figure 7 .

[0066] In a preferred embodiment of the present invention, the two endpoints of the marked line segment are used as activity safety determination points.

[0067] Considering the need to limit the left and right movement of the human body in practice, the safe zone in the game will inevitably have a left safe zone boundary and a right safe zone boundary. In addition, the two endpoints of the marked line segment are basically the points on the two side boundary lines of the corresponding parts of the human body. Using the two endpoints of the marked line segment as the activity safety judgment points makes the determination of whether the activity safety judgment points are within the safe zone more accurate. It should be noted that the two endpoints of the marked line segment can be the two endpoints of a single marked line segment, such as the left and right endpoints of the marked line segment corresponding to the torso, or the two endpoints of two marked line segments, such as the left endpoint of the left leg and the right endpoint of the right leg, etc.

[0068] It should be noted that there are many methods for labeling the marked line segments; please refer to [the relevant documentation]. Figure 8 This section presents a simple, effective, and highly accurate labeling method, which includes the following steps:

[0069] Step a: Generate lines for body parts from the human image based on human feature points:

[0070] The left arm line is generated based on at least two of the following feature points: the armpit or shoulder feature point, the elbow feature point, the wrist feature point, and the hand feature point of the left half of the human body; for example, the left arm line is generated by sequentially connecting the armpit or shoulder feature point, the elbow feature point, the wrist feature point, and the hand feature point of the left half of the human body.

[0071] And / or generate the part line of the right arm of the human body based on at least two of the feature points of the armpit or shoulder, elbow, wrist and hand of the right half of the human body; for example, generate the right arm line by sequentially connecting the feature points of the armpit or shoulder, elbow, wrist and hand of the right half of the human body.

[0072] And / or generate the feature line of the left leg of the human body based on at least two feature points from the hip feature point, knee feature point, ankle feature point and foot feature point of the left half of the human body; for example, generate the left leg line by sequentially connecting the hip feature point, knee feature point, ankle feature point and foot feature point of the left half of the human body.

[0073] And / or generate the feature line of the right leg of the human body based on at least two feature points from the hip feature point, knee feature point, ankle feature point and foot feature point of the right half of the human body; for example, generate the right leg line of the human body by sequentially connecting the hip feature point, knee feature point, ankle feature point and foot feature point of the right half of the human body.

[0074] And / or generate the body part lines of the human torso based on the Adam's apple feature point and hip feature point of the human body. For example, connect the two hip feature points of the human body to take the midpoint, and then connect the Adam's apple feature point and the midpoint to form the body torso line, or take the vertical midline of the human body that is higher than the hip feature point and lower than the head feature point as the body torso line.

[0075] Step b: Label the body parts according to the intersection of the marked line segments and the lines of the body parts:

[0076] When the marked line segment intersects with the line of the left arm of the human body, label the marked line segment with the left arm tag;

[0077] When the marked line segment intersects with the line of the right arm of the human body, label the marked line segment with the right arm tag;

[0078] When the marked line segment intersects with the line of the left leg of the human body, label the marked line segment with the left leg label;

[0079] When the marked line segment intersects with the line of the right leg of the human body, label the marked line segment with the right leg label;

[0080] When a marked line segment intersects with a human torso line, label the marked line segment with a torso tag.

[0081] It should be noted that in step a, the lines of the left arm, right arm, left leg, right leg, and torso can all be generated, or only one or a few can be generated depending on the situation. Furthermore, based on the characteristics of human anatomy, the lines of the described body parts completely pass through the corresponding body part areas. If a marked line segment intersects with it, it can be determined that the marked line segment intersects with the body part area. It should also be noted that the human body feature points are human body part corresponding feature points obtainable through image recognition, including human body part corresponding feature points obtainable through image recognition or feature points calculated based on the relative positional relationship between the human body part corresponding feature points and "other human body part corresponding feature points" / "points or lines on the human body image" / "points or lines on the game screen," etc.

[0082] Obtaining corresponding feature points of human body parts through image recognition is a common technique in this field. For example, human body recognition software based on deep learning can be used to obtain corresponding feature points of human body parts. Alternatively, human body recognition software based on deep learning can be used to obtain human body images and then image analysis can be used to obtain corresponding feature points of human body parts. Another method is to set marker lines or markers to obtain corresponding feature points of human body parts. For example, marker lines can be set on the display screen, prompting the user to make the corresponding feature points of human body parts, such as the top of the head / the vertical midline of the human body, coincide with the corresponding marker lines based on their own perception. Or, specific markers can be marked on specific parts of the human body being photographed and then image analysis can be used to obtain corresponding feature points of human body parts.

[0083] It should be noted that there are many methods for marking the marked line segments with location labels. Here, another marking method with less computation is given, which specifically includes the following steps: marking location labels according to the length and / or position of the marked line segments; the location labels are left arm labels, right arm labels, left leg labels, right leg labels, or torso labels.

[0084] When a human body faces the camera, the width and position of the torso, arms, and legs in the captured planar image of the human body will differ. Based on this, the corresponding body part labels can be marked by combining the length and / or position of the marker line segments. For example, since there are significant differences in the thickness of the torso, arms, and legs, the width of the torso, arms, and legs will also differ in the captured planar image of the human body. Body part labels can be marked by comparing the lengths of the marker line segments. For example, the longest marker line segment can be marked as the torso label. This method has low computational load and fast processing speed. In addition, since human postures are diverse, such as when the arms are extended diagonally, there may be errors in the length of the marker line segments and the thickness of their corresponding parts, especially in the arms and legs where the thickness difference is relatively small. Therefore, the corresponding body part labels can be marked by combining the positional relationship of the marker line segments, such as the torso in the middle, the arms on top, and the legs at the bottom. In addition, the hand area of ​​the upper limbs can also be obtained based on the skin color.

[0085] In a preferred embodiment of the present invention, the two endpoints of the marked line segment with the torso label are used as activity safety determination points.

[0086] In another preferred embodiment of the present invention, the left endpoint of the marker line segment with the left leg label and the right endpoint of the marker line segment with the right leg label are used as activity safety determination points.

[0087] For example, in actual gameplay, it's necessary to control the left-right position of the human body parts covered by the obscuring props to prevent excessive changes, so as not to reveal the body parts outside the obscuring props, while not restricting the position of the arms.

[0088] At the start of the game, the human torso or part of the human torso is obscured by an obscuring prop. At this time, it is necessary to obtain the two endpoints of the marked line segment with the torso tag as the activity safety judgment point. However, in reality, the user's movement is uncontrollable. When the user moves away from the camera, the left and right legs of the human body move upward in the game screen and are also obscured by the obscuring prop. At this time, it is also necessary to obtain the left endpoint of the marked line segment with the left leg tag and the right endpoint of the marked line segment with the right leg tag as the activity safety judgment point.

[0089] In a preferred embodiment of the present invention, the human body image and the game screen are synthesized, and a human body cut-off line is obtained based on the human body occlusion prop in the game screen so that the human body image below the human body cut-off line is not displayed in the game screen; and a part marking line is obtained based on the human body cut-off line.

[0090] for example Figure 6 In the game screen shown, the human body cut-off line is obtained based on the top line of the convertible car, so that the human body image below the cut-off line is not displayed in the game screen. Therefore, there is no need to restrict the range of movement of the cut-off part of the human body image, further improving the freedom of human movement during the game. At this time, the left and right endpoints of the top line of the convertible car are usually taken as the boundary points of the left and right safe zones. Based on the human body cut-off line, the part marking line is obtained. If the boundary point of the torso part marking line segment on the part marking line exceeds the boundary point of the left and right safe zones, an uncomfortable visual effect of the human body being "cut off" will occur. The activity safety judgment point is obtained through the marking line segment on a part marking line (such as the two endpoints of the torso marking line segment). The reliability of the activity safety judgment point is ensured by simple calculation, which further improves the freedom of human movement during the game. It should be noted that when the human body is upright in the game screen and the top line of the convertible car is a diagonal line or a curve, the horizontal line passing through the left and right endpoints of the top line of the convertible car can also be used as the part marking line to meet the need for more accurate left and right safe zone judgment in some game scenarios. The specific choice can be made according to the actual situation.

[0091] Finally, it should be noted that in actual operation, the human body image and game screen are composited. For example, the human body image is matched with human body occlusion props (such as cars, horses, airplanes, ships, etc.) in the game screen. At the start of the game, the human body image on the user's screen can be scaled, moved, and cropped to be composited with the game screen and human body occlusion props in the screen in accordance with the needs of the game plot, so as to ensure good visual effects. Since the actual human body movement is uncontrollable during the game, it is necessary to set a safe zone for human body movement to ensure the visual effect of the subsequent game screen. The safe zone and the part marking line have the following relationship: the part marking line is generally connected to the two opposite boundary lines of the safe zone. In most cases, the part marking line can be perpendicular to the two opposite boundary lines, so that the calculation efficiency when making safety judgment is higher. For example, in the game screen where the human body is upright, the left and right safe zones for human body movement are generally the area between the left and right safe lines generated based on the human body occlusion props. The left and right safe lines are generally the left and right boundary lines of the occlusion props. It should be further explained that, in addition to matching human images with human occlusion props in the game screen and setting safe zones, human images can also be matched with other props (such as human-carrying props like scooters and fire wheels) or areas (such as roads and rivers) in the game screen and set safe zones.

[0092] Example 2

[0093] like Figure 9 As shown, a system for generating security judgment points for live-action game activities includes a first acquisition module, an identification module, and a second acquisition module.

[0094] The first acquisition module is used to acquire a human body image and part marking lines, wherein the part marking lines are marking lines that pass through the human body image;

[0095] like Figure 2 As shown, in practice, you can use the camera above the TV to take a picture of yourself to obtain a human image. The human image is then combined with the game screen, and the user can observe the combined effect of their own human image and the game screen on the TV screen. It should be noted that at the start of the game, the human image on the user's screen can be scaled, moved, etc., to combine it with the game screen and human body occlusion props in the screen to meet the needs of the game plot, so as to ensure a good visual effect.

[0096] The process involves obtaining part marker lines that pass through a human body image. For example, based on a human body occlusion prop in a game screen, such as a convertible, part marker lines can be obtained by horizontally passing through the occlusion prop and corresponding to the torso. This allows for the acquisition of activity safety marker points on both sides of the human torso, ensuring that the torso obscured by the convertible does not extend beyond the convertible's left and right boundaries, thus avoiding unpleasant visual effects. Simultaneously, it does not restrict the movement of the upper limbs, reducing the user's sense of constraint. It should be noted that the purpose of these part marker lines is to obtain appropriate activity safety judgment points by acquiring mutually separated marker line segments on these lines and marking them with corresponding part labels, based on the characteristics of human anatomy. These lines can be horizontal, diagonal, or vertical. The marking lines or curves can be selected based on the actual game scenario. For example, when the human body is in an upright position in the game, horizontal part marking lines can generally be used to obtain the activity safety judgment points corresponding to the left and right safe movement zones of the human body. When the human body is in a tilted position in the game, such as when props and the human body rotate together in the game screen, the part marking lines can also be diagonal lines. Furthermore, when the human body is in an upright position in the game, vertical part marking lines can also be used to obtain the activity safety judgment points corresponding to the up and down safe movement zones of the human body. In addition, lines perpendicular to the lines connecting the joints of the human body parts (such as the line connecting the elbow joint and the wrist joint) can be used to make it easier to obtain the marking line segments corresponding to the flexible human body parts (such as the upper limbs). The marking line segment refers to the continuous area on the part marking line that overlaps with the human body image.

[0097] In this embodiment, it is assumed that the part marking line is a horizontal line, which is used to obtain the activity safety judgment point corresponding to the left and right movement safety zone of the human body.

[0098] It should be noted that the location marker lines can be one or more. For example, when the top of the convertible is a horizontal line, location marker lines can be obtained based on the horizontal line. Furthermore, the human image below the top of the convertible can be cropped out of the game screen, thus eliminating the need to restrict the range of motion of the cropped portion of the human image. This allows for the acquisition of activity safety judgment points based on marker segments on a single location marker line (such as the two endpoints of the torso marker segment), ensuring the reliability of the activity safety judgment points and further improving the accuracy of human movement judgments during gameplay. In terms of freedom, in this case, the convertible is usually an opaque obscuring prop; for example, when the convertible is a transparent obscuring prop according to another plot requirement of the game, it is not possible to crop the human image between the top and bottom of the convertible. In this case, multiple horizontal lines between the top and bottom of the convertible can be obtained. These multiple horizontal lines can be scanned line by line to generate multiple corresponding part marking lines. Then, based on the marking line segments on the multiple part marking lines (such as the two endpoints of the torso marking line segment), activity safety judgment points can be obtained to perform activity safety judgment on a larger range of human images.

[0099] The recognition module is used to identify several marker line segments that overlap with the human body image on the body part marking line, wherein at least one of the marker line segments is marked with at least one body part label corresponding to a human body part region, specifically as follows: Figure 3 As shown;

[0100] It should be noted that due to the complexity of human movements and structure (for example, different parts of the human body may intersect each other, and the edges of different parts of the human body are irregular curves), as well as the diversity of game plot design (for example, the shapes and sizes of human body covering props are varied), it is extremely difficult to determine the safe zone by relying on existing technology to achieve a good connection between the edges of specific parts of the human body and the human body covering props.

[0101] Based on the characteristics of human anatomy, the overlapping part of a part marking line and a single human part (such as the torso, upper limbs, lower limbs, etc.) must be a marking line segment. The overlapping part of a part marking line and multiple non-intersecting human parts is usually multiple discontinuous marking line segments. By using the marking line segments corresponding to specific part labels, the activity safety judgment points on specific human part areas that meet the needs of the game plot can be obtained simply and accurately.

[0102] For example, when a body part marking line crosses the upper body and the arms are spread out in a V-shape, there will be marking line segments overlapping with the torso, the left arm, and the right arm. One of these segments is used to mark the body part. In this embodiment, it is assumed that the segment overlapping the torso is used, and the body part label is the torso label. Of course, the segments overlapping the left and right arms can also be labeled, depending on the actual situation. It should be noted that when a body part marking line crosses the upper body but the arms are close to the torso, there is only one marking line segment. This segment can be labeled only with the torso label, or it can be labeled with the left arm, torso, and right arm labels. Because the arms are close to the torso, the activity safety judgment point will still be located at or near the torso, and the activity safety zone judgment will not cause excessive restraint on the user. Similarly, for example... Figure 4 As shown, when the upper arm and torso of a human body cross, and the crossing part is exactly on the location marking line, the activity safety judgment point will still be located on or near the torso because the horizontal length of the crossing part is usually relatively short.

[0103] It should be noted that for several marker line segments that overlap with the human body image, only the marker line segment corresponding to a certain human body part area can be labeled with a part label, such as only labeling the part label corresponding to the upper limb of the human body, and then obtaining activity safety judgment points on the other unlabeled marker line segments. Alternatively, the marker line segments corresponding to multiple human body parts areas can be labeled with part labels, such as labeling the part labels corresponding to the human torso, the left leg of the human body, and the right leg of the human body, and obtaining activity safety judgment points on the aforementioned marker line segments.

[0104] It should be noted that identifying several marker segments on the part marking line that overlap with the human body image is a common technique in this field. For example, identification based on the transparency features of pixels on the part marking line, or identification based on the intersection of the part marking line and the human body contour line, etc.

[0105] The human body parts typically include one or more of the following: the human torso, the left arm, the right arm, the left leg, and the right leg. The specific number of human body parts used depends on the situation. In this embodiment, the human body parts used are the human torso, the left leg, and the right leg.

[0106] The second acquisition module is used to acquire at least one point on at least one of the marked line segments as an activity safety determination point based on the location label of the marked line segment.

[0107] Here, assuming there are three marked line segments, labeled as left arm, torso, and right arm, and the torso or part of it is obscured by an occlusion prop, the game requires that the left and right positions of the torso remain relatively stable to avoid being visible outside the occlusion prop. However, arm position changes are not restricted. In this case, at least one point on the torso label's marked line segment is taken as the safe movement judgment point. Of course, in this game scenario, the action of marking the body part can be performed based solely on the torso position, directly obtaining the required marked line segments. However, in reality, when the user moves away from the camera, causing the left and right legs to move upwards in the game screen and also be obscured by the occlusion prop, separate left and right leg labels are needed. (See reference here.) Figure 5 , Figure 6 and Figure 7 .

[0108] In a preferred embodiment of the present invention, when the first acquisition module acquires the part marking line, it synthesizes the human body image and the game screen, acquires the human body cut-off line based on the human body occlusion prop in the game screen, so that the human body image below the human body cut-off line is not displayed in the game screen; and acquires the part marking line based on the human body cut-off line.

[0109] for example Figure 6 In the game screen shown, the human body cut-off line is obtained based on the top line of the convertible car, so that the human body image below the cut-off line is not displayed in the game screen. Therefore, there is no need to restrict the range of movement of the cut-off part of the human body image, further improving the freedom of human movement during the game. At this time, the left and right endpoints of the top line of the convertible car are usually taken as the boundary points of the left and right safe zones. Based on the human body cut-off line, the part marking line is obtained. If the boundary point of the torso part marking line segment on the part marking line exceeds the boundary point of the left and right safe zones, an uncomfortable visual effect of the human body being "cut off" will occur. The activity safety judgment point is obtained through the marking line segment on a part marking line (such as the two endpoints of the torso marking line segment). The reliability of the activity safety judgment point is ensured by simple calculation, which further improves the freedom of human movement during the game. It should be noted that when the human body is upright in the game screen and the top line of the convertible car is a diagonal line or a curve, the horizontal line passing through the left and right endpoints of the top line of the convertible car can also be used as the part marking line to meet the need for more accurate left and right safe zone judgment in some game scenarios. The specific choice can be made according to the actual situation.

[0110] In a preferred embodiment of the present invention, the second acquisition module acquires the two endpoints of the marked line segment as active safety determination points.

[0111] Considering the need to limit the left and right movement of the human body in practice, the safe zone in the game will inevitably have a left safe zone boundary and a right safe zone boundary. In addition, the two endpoints of the marked line segment are basically the points on the two side boundary lines of the corresponding parts of the human body. Using the two endpoints of the marked line segment as the activity safety judgment points makes the determination of whether the activity safety judgment points are within the safe zone more accurate. It should be noted that the two endpoints of the marked line segment can be the two endpoints of a single marked line segment, such as the left and right endpoints of the marked line segment corresponding to the torso, or the two endpoints of two marked line segments, such as the left endpoint of the left leg and the right endpoint of the right leg, etc.

[0112] It should be noted that there are many methods for labeling the marked line segments; please refer to [the relevant documentation]. Figure 8 This section presents a simple, effective, and highly accurate labeling method, which includes the following steps:

[0113] Step a: Generate lines for body parts from the human image based on human feature points:

[0114] The left arm line is generated based on at least two of the following feature points: the armpit or shoulder feature point, the elbow feature point, the wrist feature point, and the hand feature point of the left half of the human body; for example, the left arm line is generated by sequentially connecting the armpit or shoulder feature point, the elbow feature point, the wrist feature point, and the hand feature point of the left half of the human body.

[0115] And / or generate the part line of the right arm of the human body based on at least two of the feature points of the armpit or shoulder, elbow, wrist and hand of the right half of the human body; for example, generate the right arm line by sequentially connecting the feature points of the armpit or shoulder, elbow, wrist and hand of the right half of the human body.

[0116] And / or generate the feature line of the left leg of the human body based on at least two feature points from the hip feature point, knee feature point, ankle feature point and foot feature point of the left half of the human body; for example, generate the left leg line by sequentially connecting the hip feature point, knee feature point, ankle feature point and foot feature point of the left half of the human body.

[0117] And / or generate the feature line of the right leg of the human body based on at least two feature points from the hip feature point, knee feature point, ankle feature point and foot feature point of the right half of the human body; for example, generate the right leg line of the human body by sequentially connecting the hip feature point, knee feature point, ankle feature point and foot feature point of the right half of the human body.

[0118] And / or generate the body part lines of the human torso based on the Adam's apple feature point and hip feature point of the human body. For example, connect the two hip feature points of the human body to take the midpoint, and then connect the Adam's apple feature point and the midpoint to form the body torso line, or take the vertical midline of the human body that is higher than the hip feature point and lower than the head feature point as the body torso line.

[0119] Step b: Label the body parts according to the intersection of the marked line segments and the lines of the body parts:

[0120] When the marked line segment intersects with the line of the left arm of the human body, label the marked line segment with the left arm tag;

[0121] When the marked line segment intersects with the line of the right arm of the human body, label the marked line segment with the right arm tag;

[0122] When the marked line segment intersects with the line of the left leg of the human body, label the marked line segment with the left leg label;

[0123] When the marked line segment intersects with the line of the right leg of the human body, label the marked line segment with the right leg label;

[0124] When a marked line segment intersects with a human torso line, label the marked line segment with a torso tag.

[0125] It should be noted that in step a, the lines of the human left arm, human right arm, human left leg, human right leg, and human torso can all be generated, or only one or a few can be generated depending on the situation; in addition, according to the characteristics of human anatomy, the lines of the parts completely pass through the corresponding human body parts area. If the marked line segment intersects with it, it can be determined that the marked line segment intersects with the human body parts area.

[0126] It should be noted that the human body feature points are feature points corresponding to human body parts that can be obtained through image recognition, including feature points corresponding to human body parts that can be obtained through image recognition or feature points calculated based on the relative positional relationship between the human body part corresponding feature points and "feature points corresponding to other human body parts" / "points or lines on human body images" / "points or lines on game screens", etc.

[0127] Obtaining corresponding feature points of human body parts through image recognition is a common technique in this field. For example, human body recognition software based on deep learning can be used to obtain corresponding feature points of human body parts. Alternatively, human body recognition software based on deep learning can be used to obtain human body images and then image analysis can be used to obtain corresponding feature points of human body parts. Another method is to set marker lines or markers to obtain corresponding feature points of human body parts. For example, marker lines can be set on the display screen, prompting the user to make the corresponding feature points of human body parts, such as the top of the head / the vertical midline of the human body, coincide with the corresponding marker lines based on their own perception. Or, specific markers can be marked on specific parts of the human body being photographed and then image analysis can be used to obtain corresponding feature points of human body parts.

[0128] It should be noted that there are many methods for marking the marked line segments with location labels. Here, another marking method with less computation is given, which specifically includes the following steps: marking location labels according to the length and / or position of the marked line segments; the location labels are left arm labels, right arm labels, left leg labels, right leg labels, or torso labels.

[0129] When a human body faces the camera, the width and position of the torso, arms, and legs in the captured planar image of the human body will differ. Based on this, the corresponding body part labels can be marked by combining the length and / or position of the marker line segments. For example, since there are significant differences in the thickness of the torso, arms, and legs, the width of the torso, arms, and legs will also differ in the captured planar image of the human body. Body part labels can be marked by comparing the lengths of the marker line segments. For example, the longest marker line segment can be marked as the torso label. This method has low computational load and fast processing speed. In addition, since human postures are diverse, such as when the arms are extended diagonally, there may be errors in the length of the marker line segments and the thickness of their corresponding parts, especially in the arms and legs where the thickness difference is relatively small. Therefore, the corresponding body part labels can be marked by combining the positional relationship of the marker line segments, such as the torso in the middle, the arms on top, and the legs at the bottom. In addition, the hand area of ​​the upper limbs can also be obtained based on the skin color.

[0130] In a preferred embodiment of the present invention, the second acquisition module acquires the two endpoints of the marked line segment with the torso label as activity safety determination points.

[0131] In another preferred embodiment of the present invention, the second acquisition module acquires the left endpoint of the marker line segment with the left leg label and the right endpoint of the marker line segment with the right leg label as activity safety determination points.

[0132] For example, in actual gameplay, it's necessary to control the left-right position of the human body parts covered by the obscuring props to prevent excessive changes, so as not to reveal the body parts outside the obscuring props, while not restricting the position of the arms.

[0133] At the start of the game, the human torso or part of the human torso is obscured by an obscuring prop. At this time, it is necessary to obtain the two endpoints of the marked line segment with the torso tag as the activity safety judgment point. However, in reality, the user's movement is uncontrollable. When the user moves away from the camera, the left and right legs of the human body move upward in the game screen and are also obscured by the obscuring prop. At this time, it is also necessary to obtain the left endpoint of the marked line segment with the left leg tag and the right endpoint of the marked line segment with the right leg tag as the activity safety judgment point.

[0134] Finally, it should be noted that in actual operation, the human body image and game screen are composited. For example, the human body image is matched with human body occlusion props (such as cars, horses, airplanes, ships, etc.) in the game screen. At the start of the game, the human body image on the user's screen can be scaled, moved, and cropped to be composited with the game screen and human body occlusion props in the screen in accordance with the needs of the game plot, so as to ensure good visual effects. Since the actual human body movement is uncontrollable during the game, it is necessary to set a safe zone for human body movement to ensure the visual effect of the subsequent game screen. The safe zone and the part marking line have the following relationship: the part marking line is generally connected to the two opposite boundary lines of the safe zone. In most cases, the part marking line can be perpendicular to the two opposite boundary lines, so that the calculation efficiency when making safety judgment is higher. For example, in the game screen where the human body is upright, the left and right safe zones for human body movement are generally the area between the left and right safe lines generated based on the human body occlusion props. The left and right safe lines are generally the left and right boundary lines of the occlusion props. It should be further explained that, in addition to matching human images with human occlusion props in the game screen and setting safe zones, human images can also be matched with other props (such as human-carrying props like scooters and fire wheels) or areas (such as roads and rivers) in the game screen and set safe zones.

[0135] Example 3

[0136] A method for generating safe zones in live-action games involves executing the live-action game activity safety determination point generation method described in Example 1 or the live-action game activity safety determination point generation system described in Example 2 to generate activity safety determination points and generate safe zones based on these points.

[0137] Generating an active safety zone based on active safety decision points includes the following steps:

[0138] Obtain the marked line segment where the activity safety decision point is located, and generate the activity safety zone based on the marked line segment.

[0139] Specifically, two boundary lines for obtaining the safe zone are established in the two extension directions of the marked line segment. The marked line segment is perpendicular to the two boundary lines, and the distance between the endpoint of the marked line segment and the boundary line in the extension direction can be freely set according to the actual game scene.

[0140] The advantage of doing this is that it makes the two boundaries of the safe zone perpendicular to the direction of movement of the human body parts, allowing for a more reasonable setting of the boundaries of the safe zone.

[0141] Example 4

[0142] A method for generating a safe zone in a live-action game video includes the following steps:

[0143] Using the method for generating the safe zone in a live-action game as described in Example 3, the marked line segment and the safe zone where the safe zone determination point is located on the human body image of the reference frame are obtained;

[0144] Obtain the marked line segment where the activity safety judgment point is located on the current frame human image. Based on the ratio of the marked line segment where the activity safety judgment point is located on the current frame human image to the marked line segment where the activity safety judgment point is located on the reference frame human image, adjust the size of the activity safety area on the reference frame human image, and generate the activity safety area on the current frame human image.

[0145] For example: if the ratio of the marked line segment where the activity safety determination point is located on the current frame human body image to the marked line segment where the activity safety determination point is located on the reference frame human body image is 1:2, and the activity safety area is a rectangular area, then the ratio of the activity safety area on the current frame human body image to the activity safety area on the reference frame human body image is 1:2 or generated after adjustment based on the ratio.

[0146] It should be noted that in practice, the distance between the human body and the camera may change due to the movement of the human body. If the safe area is a fixed size, it is easy to exceed the safe area. Embodiment 4 can make the safe area change dynamically to solve this problem (for example, the size of the human body blocking prop can be dynamically adjusted based on the dynamically changing safe area).

[0147] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the present invention. Any simple modifications, alterations, or equivalent structural changes made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention. The protection scope of the present invention is determined by the scope of the claims and their equivalents.

Claims

1. A method for generating a safety decision point in a live action game event, characterized by, The method comprises the following steps: Obtaining a human body image and a part marking line, the part marking line being a marking line passing through the human body image; Identifying a plurality of marking line segments on the part marking line that overlap with the human body image, at least one of the marking line segments being marked with a part label of at least one corresponding human body part region; Obtaining at least one point on at least one of the marking line segments as an activity safety judgment point according to the part label of the marking line segment; The activity safety judgment point is used to judge whether the corresponding human body part region is within an activity safety zone.

2. The method of claim 1, wherein the method further comprises: Two end points of the obtained marking line segment are activity safety judgment points.

3. The method of claim 1, wherein the method further comprises: The human body part region comprises one or more of a human body trunk part, a human body left arm part, a human body right arm part, a human body left leg part and a human body right leg part.

4. The method of claim 3, wherein the method further comprises: Marking the part label on the marking line segment comprises the following steps: Step a, generating human body part lines for the human body image according to human body feature points: Generating a part line for the human body left arm part according to at least two feature points of the human body left arm; And / or generating a part line for the human body right arm part according to at least two feature points of the human body right arm; And / or generating a part line for the human body left leg part according to at least two feature points of the human body left leg; And / or generating a part line for the human body right leg part according to at least two feature points of the human body right leg; And / or generating a part line for the human body trunk part according to at least two feature points of the human body trunk, or generating a part line for the human body trunk part according to a vertical center line of the human body; Step b, marking a part label according to the intersection of the marking line segment and the human body part line: When the marking line segment intersects with the part line of the human body left arm part, marking a left arm label on the marking line segment; When the marking line segment intersects with the part line of the human body right arm part, marking a right arm label on the marking line segment; When the marking line segment intersects with the part line of the human body left leg part, marking a left leg label on the marking line segment; When the marking line segment intersects with the part line of the human body right leg part, marking a right leg label on the marking line segment; When the marking line segment intersects with the part line of the human body trunk part, marking a trunk label on the marking line segment.

5. The method of claim 3, wherein the method further comprises: Marking the part label on the marking line segment comprises the following steps: Marking a part label according to the length and / or position of the marking line segment, the part label being a left arm label, a right arm label, a left leg label, a right leg label or a trunk label.

6. A method of generating a safety decision point for a live game activity according to claim 4 or 5, characterized in that, Obtaining two end points of the marking line segment with the trunk label as activity safety judgment points.

7. A method of generating a safety judgment point for a live action game according to claim 4 or 5, wherein Obtaining a left side end point of the marking line segment with the left leg label and a right side end point of the marking line segment with the right leg label as activity safety judgment points.

8. The method of claim 1 or 2, wherein the method further comprises: Synthesizing the human body image and a game picture, obtaining a human body intercepting line based on a human body occlusion prop in the game picture, so that the part of the human body image below the human body intercepting line is not displayed in the game picture, and obtaining the part marking line based on the human body intercepting line.

9. A live action game event safety adjudication point generation system, characterized by, The method comprises a first obtaining module, an identifying module and a second obtaining module, The first obtaining module is configured to obtain a human body image and a part marking line, the part marking line being a marking line passing through the human body image; The identification module is configured to identify a plurality of mark line segments on the part mark line that overlap the human body image, and at least one of the mark line segments is marked with a part label of at least one corresponding human body part region; The second acquisition module is configured to acquire at least one point on at least one of the mark line segments as an activity safety judgment point according to the part label of the mark line segment. The activity safety judgment point is configured to judge whether the corresponding human body part region is in the activity safety zone.

10. A live action game play safety adjudication point generation system according to claim 9, wherein, When the first acquisition module acquires the part mark line, the human body image and the game screen are synthesized, a human body intercepting line is acquired based on a human body shielding prop in the game screen, and a part below the human body intercepting line is not displayed in the game screen; and the part mark line is acquired based on the human body intercepting line.

11. A method for generating a safe zone in a live-action game, characterized in that: The method for generating an activity safety judgment point in a live game according to any one of claims 1-8 is executed to generate an activity safety judgment point, and an activity safety zone is generated based on the activity safety judgment point.

12. The method of claim 11, wherein the method further comprises: The activity safety zone is generated based on the activity safety judgment point, and includes the following steps: A mark line segment where the activity safety judgment point is located is acquired, and an activity safety zone is generated based on the mark line segment.

13. A method for generating a safe zone in a live-action game video, characterized in that: The method for generating an activity safety zone in a live game according to claim 11 or 12 includes the following steps: A mark line segment where the activity safety judgment point is located on a reference frame human body image and an activity safety zone are acquired; A mark line segment where the activity safety judgment point is located on a current frame human body image is acquired, the size of the activity safety zone on the reference frame human body image is adjusted based on the proportion of the mark line segment where the activity safety judgment point is located on the current frame human body image and the mark line segment where the activity safety judgment point is located on the reference frame human body image, and an activity safety zone on the current frame human body image is generated.

Images