WO2020037491A1 - 一种投影系统自动适配投影目标的方法 - Google Patents

一种投影系统自动适配投影目标的方法 Download PDF

Info

Publication number
WO2020037491A1
WO2020037491A1 PCT/CN2018/101505 CN2018101505W WO2020037491A1 WO 2020037491 A1 WO2020037491 A1 WO 2020037491A1 CN 2018101505 W CN2018101505 W CN 2018101505W WO 2020037491 A1 WO2020037491 A1 WO 2020037491A1
Authority
WO
WIPO (PCT)
Prior art keywords
projection
projection target
target
picture
screen
Prior art date
Application number
PCT/CN2018/101505
Other languages
English (en)
French (fr)
Inventor
那庆林
麦浩晃
蒋海滨
黄彦
Original Assignee
神画科技（深圳）有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 神画科技（深圳）有限公司 filed Critical 神画科技（深圳）有限公司
Priority to PCT/CN2018/101505 priority Critical patent/WO2020037491A1/zh
Publication of WO2020037491A1 publication Critical patent/WO2020037491A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G03—PHOTOGRAPHY; CINEMATOGRAPHY; ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ELECTROGRAPHY; HOLOGRAPHY
    • G03B—APPARATUS OR ARRANGEMENTS FOR TAKING PHOTOGRAPHS OR FOR PROJECTING OR VIEWING THEM; APPARATUS OR ARRANGEMENTS EMPLOYING ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ACCESSORIES THEREFOR
    • G03B21/00—Projectors or projection-type viewers; Accessories therefor
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N5/00—Details of television systems
    • H04N5/74—Projection arrangements for image reproduction, e.g. using eidophor

Definitions

  • the present invention relates to a projector, and more particularly, to a method for a projection system to automatically adapt a projection target.
  • the projector in a home scene, if the projector is not fixedly installed after presetting, but is placed on the desktop when in use, then even if the projector is facing the projection screen, it is necessary to repeatedly adjust the position of the projector multiple times. To ensure that the projected picture fits exactly with the projected screen, for example, it fits within the border of the projected screen.
  • the technical problem to be solved by the present invention is the problem that the projector cannot automatically adapt to the projection screen in the prior art.
  • the present invention provides a method for automatically adapting a projection system to a projection system.
  • the projection system includes a projection unit with a projection imaging chip, a monitoring unit with a monitoring imaging chip, and a method for controlling the projection unit.
  • Image control unit wherein, when the projection system is working normally, automatic adaptation processing is performed according to the following steps:
  • Step S1 the monitoring unit recognizes at least two projection target identification points provided on the projection target that are different from the projection target subject, and simultaneously recognizes a projection screen projected on the projection target by the projection unit, and Generating corresponding projection target identification point position information and projection screen position information by the monitoring imaging chip;
  • Step S2 according to the projection target identification point position information and the projection screen position information, the image control unit sends a control instruction to the projection imaging chip according to a predetermined rule to correct the boundary of the projection screen so that the The boundary of the projection picture and the projection target identification point reach a predetermined close level.
  • the steps S1 and S2 may be performed only once, and the boundary of the projection screen and the projection target identification point reach a predetermined close level by one-time correction.
  • the steps S1 and S2 may be repeatedly performed, and correction of cyclic approximation is performed multiple times until the boundary of the projection screen and the projection target identification point reach a predetermined close level.
  • an inside-out method when performing the multiple-cycle approximation correction, an inside-out method may be adopted, and a minimum projection picture is first generated with the center point of the projection picture as a center, and then the minimum projection is used. The picture is gradually increased as a starting point until the boundary of the projection picture and the projection target identification point reach a predetermined close level.
  • an outward-to-inward method may also be adopted, and the current projection picture is used as a starting point to gradually decrease until the boundary of the projection picture and the projection are reduced.
  • the target identification point reaches a predetermined close level.
  • the projection target identification point has a clear color difference from the projection target subject;
  • the projection target may be a rectangular projection screen, and the projection target identification point may be similar to the rectangular projection screen. Points near corners; or the projection target may be a wall, and the projection target identification points may be points arranged horizontally and / or vertically on the wall.
  • step S2 when the monitoring unit recognizes only two projection target identification points on the projection target, in step S2, the connection of the two projection target identification points is performed.
  • the line serves as a boundary of the projection picture, and a complete projection picture is generated according to a predetermined projection picture ratio.
  • the predetermined projection picture ratio may be 16: 9 or 4: 3.
  • a step of identifying a projection distance and an angle may be further included; in step S2, a compensation calculation is also performed in combination with the identified distance and angle, and a predetermined projection screen ratio is generated. Rectangular projection screen. A line having a significant color difference from the projection target subject may also be set between the two projection target identification points.
  • step S2 when the monitoring unit recognizes only three of the projection target identification points on the projection target, in step S2, the approaching of the three projection target identification points is performed.
  • the connecting lines in the vertical and horizontal directions are used as the two boundaries of the projection picture, and then a complete projection picture is generated based on the two boundaries.
  • step S2 a step of identifying a projection distance and an angle may be further included.
  • step S2 a compensation calculation is also performed in combination with the identified distance and angle to generate a rectangular projection screen. It is also possible to set two lines in the vertical and horizontal directions with significant color differences from the projection target subject between the three projection target identification points, and the two lines are perpendicular to each other.
  • step S2 when the monitoring unit recognizes four projection target identification points on the projection target, in step S2, the four projection target identification points are close to vertical.
  • the horizontal lines are corrected as the four boundaries of the projection picture, so that the boundaries of the projection picture and the four boundaries reach a predetermined close level.
  • Four lines in the vertical and horizontal directions with significant color differences from the projection target subject may also be set between the four projection target identification points, and the four lines form a rectangle.
  • Conventional projection screens with black borders belong to this state.
  • the four corners are the aforementioned four projection target identification points, and the black borders are the four lines.
  • the projection picture is the initial projection picture that can be generated by the projection unit monitor, or the maximum projection picture; after the first time, When the step S1 is continuously performed, the projection picture is the current projection picture after the last correction.
  • step S1 when the step S1 is performed for the first time, if at least two of the projection target identification points are not identified, or at least two of the identification points are identified. If it is not located in the projection screen, an adjustment prompt is issued, prompting the user to adjust the positional relationship between the projection system and the projection target to ensure that at least two identification points of the projection target fall within the monitoring range of the monitoring unit. And within the projection picture.
  • step S1 of the method according to the present invention when the monitoring unit recognizes a projection picture projected by the projection unit on the projection target, the monitoring unit obtains the picture by taking a picture and then using an image recognition method.
  • the boundary of the projection picture, or the boundary of the projection picture is obtained by displaying an easily recognizable pattern or marked point in the projection picture.
  • the projection system of the present invention can automatically adapt to a projection target (including a projection screen, a wall, or another plane that can be used to display a projection screen), and ensure that at least two projection target identification points fall on the monitoring unit.
  • a projection target including a projection screen, a wall, or another plane that can be used to display a projection screen
  • different methods two, three, and four
  • the projection screen will fill the quadrilateral space defined by the four identification points, and for the projection screen, it will be a rectangular space defined by the projection screen frame.
  • FIG. 1A is a schematic diagram of the positional relationship between the monitoring range, the projection screen, and the projection screen and the corner point projection in Embodiment 1 of the present invention
  • FIG. 1B is a schematic diagram of a correspondence relationship between coordinates in a monitoring imaging chip and coordinates in a projection imaging chip according to Embodiment 1 of the present invention
  • Embodiment 2A is a schematic diagram of depth calculation in Embodiment 2 of the present invention.
  • FIG. 2B is a schematic diagram of adjustment by distance and angle in Embodiment 2 of the present invention.
  • Embodiment 3A is a schematic diagram of adjustment by an iterative method in Embodiment 3 of the present invention.
  • 3B is a schematic diagram of the partition of the lower left corner of the screen at the relative position of the lower left corner of the projection point in Embodiment 3 of the present invention.
  • FIG. 4A is a schematic diagram of a projection screen having two boundary points outside the projection screen in Embodiment 4 of the present invention.
  • FIG. 4B is a schematic diagram of a projection screen having three boundary points outside the projection screen in Embodiment 4 of the present invention.
  • FIG. 5A is a schematic diagram of a viewing angle of a monitoring lens having two boundary points outside a monitoring range of a projection screen in Embodiment 5 of the present invention
  • FIG. 5B is a schematic perspective view of the projection screen in the vertical direction of the projection screen with two boundary points outside the monitoring range in Embodiment 5 of the present invention
  • 5C is a schematic diagram of a viewing angle of a monitoring lens when only two adjacent edges or two adjacent edges of a projection screen are identified in Embodiment 5 of the present invention.
  • FIG. 6A is a schematic diagram of a viewing angle of a monitoring lens when only two opposite sides of a projection screen are identified in Embodiment 6 of the present invention
  • FIG. 6B is a schematic diagram of a vertical viewing angle of the projection screen when only two opposite sides of the projection screen are recognized in Embodiment 6 of the present invention.
  • FIG. 7A is a schematic diagram of iterating from inside to outside in Embodiment 3 of the present invention.
  • FIG. 7B is a schematic diagram of iterating from outside to inside in Embodiment 3 of the present invention.
  • FIG. 7C is a schematic diagram of mapping first, and then iterating from the inside to the outside in Embodiment 3 of the present invention.
  • FIG. 7B is a schematic diagram of mapping first and iterating from outside to inside in Embodiment 3 of the present invention.
  • FIG. 8 is a schematic diagram of the position of an estimated boundary vertex of an internal marker in Embodiment 8 of the present invention.
  • the corner vertices of the projection screen that are not specifically described in the following implementation examples refer to the corner vertices of the undeformed projection screen.
  • the image collected by the monitoring unit is the same direction as the image projected by the projector, that is, The upper left boundary of the projection screen is in the upper left corner of the image obtained by the monitoring unit, the upper right boundary of the projection screen is in the upper right corner of the image obtained by the monitoring unit, the lower left boundary of the projection screen is in the lower left corner of the image obtained by the monitoring unit, and so on.
  • Embodiment 1 direct mapping method
  • the projection system includes a projection unit having a projection imaging chip, a monitoring unit having a monitoring imaging chip, and an image control unit for controlling the projection unit.
  • the projection system and the projection target can be placed arbitrarily. This embodiment is placed at a slight tilt; the projection target can be a projection screen, a wall, or another plane that can be used to display the projection screen.
  • the monitoring unit can capture a quadrangular shape (WC1, WC2, WC3, WC4), the four corner vertices of the projection screen are (WP1, WP2, WP3, WP4), the four corner vertices of the projection screen Yes (WS1, WS2, WS3, WS4). It can be seen from the figure that the area of these three becomes smaller in order, and the monitoring unit can capture the largest range, the second largest projection screen, and the smallest projection screen. At this time, the four corner vertices of the projection screen and the four corner vertices of the projection screen are all within the range of the quadrangle (WC1, WC2, WC3, WC4), that is, they can be recognized by the monitoring unit.
  • the projector will project the undeformed projection image boundary points (P11, P22, P33, P44) on the projection imaging chip to the projection screen, and the four corner vertices (WP1, WP2, WP3, WP4).
  • the four corner vertices (WP1, WP2, WP3, WP4) of the projection screen and the four corner vertices (WS1, WS2, WS3, WS4) of the projection screen are imaged as the projection screen monitoring boundary on the monitoring imaging chip of the monitoring unit. Points (P1, P2, P3, P4) and projection screen monitoring boundary points (S1, S2, S3, S4).
  • the projection control unit needs to do the projection screen Deformation processing.
  • This deformation processing is to map the projection image boundary points (P11, P22, P33, P44) to (S11, S22, S33, S44) on the projection imaging chip, and then the projector will (S11, S22, S33, S44) when projecting onto the projection screen, so that the corner vertices of the obtained projection screen are exactly on the four corner vertices (WS1, WS2, WS3, WS4) of the projection screen, so that the projection screen completely fills the projection screen, To achieve the best visual effects.
  • This deformation is generally implemented using the perspective transformation formula (1).
  • (x, y) coordinates before perspective transformation and (x ', y') are coordinates after perspective transformation.
  • the optical center of the monitoring unit and the optical center of the projection unit are very close to each other. It can be approximated that the optical center of the projection unit and the optical center of the monitoring unit are at the same point, which is the perspective center. Therefore, monitoring in the same fixed scene There is a one-to-one correspondence between the points on the monitoring imaging chip in the unit and the points on the projection imaging chip in the projection unit.
  • the transformation T can be constructed to map the points (P1, P2, P3, P4) on the surveillance imaging chip to the points (P11, P22, P33, P44) on the projection imaging chip, and then calculate the projection on the surveillance imaging chip The coordinates of the screen monitoring boundary points (S1, S2, S3, S4) on the projection imaging chip (S11, S22, S33, S44).
  • the four corner vertices (WP1, WP2, WP3, WP4) of the projection picture can be obtained by identifying the boundary line or the intersection of the boundary lines of the projection picture, or a specific identifier can be printed by the projector to identify this special And then calculate the positions of the four corner vertices (WP1, WP2, WP3, WP4) of the projection screen according to the coordinates of the marker on the projection imaging chip.
  • the above embodiment corresponds to the situation when the monitoring unit described in the summary of the invention can recognize four projection target identification points on the projection target.
  • the monitoring unit When the monitoring unit only recognizes the two identification points of the projection target or the two endpoints of an edge, it can combine the aspect ratio, left and right rotation angle, up and down elevation angle, internal parameters of the projection unit, internal parameters of the monitoring unit, and projection screen identification. Point to the depth of the projector and other information to calculate the coordinates of the corners and vertices of other projection screens on the projection imaging chip, and then calculate the distortion K of the projection screen.
  • the depths Z1 and Z2 of the two identification points S1 and S2 from the projector need to be obtained. This can be obtained through external equipment or based on other information such as estimates.
  • the point on the projection imaging chip is projected onto the point S through the projection unit P and seen by the monitoring unit C.
  • the (S1) method in Embodiment 1 is used to calculate the coordinates S11 of the point S1 on the projection imaging chip, so the projection direction vector can be calculated according to the internal parameters of the projection unit.
  • the direction vector of the received light can be obtained Simultaneously two equations can be obtained Equation 1, the depth of this point is:
  • f p is the focal length of the projection unit
  • (C xp , C yp ) is the position of the projection unit optical center on the projection imaging chip
  • (x S11 , y S11 ) is the coordinate of S11 on the projection imaging chip
  • f c is the monitoring
  • the focal length of the unit, (C xc , C yc ) is the position of the optical center of the monitoring unit on the monitoring imaging chip
  • (x S1 , y S1 ) is the coordinate of S1 on the monitoring imaging chip. So we can get the spatial position of WS1 relative to the projector as
  • the spatial coordinates of WS2 relative to the projector can be obtained as
  • the normal direction vector of the plane of the projection screen in the projector is calculated according to the elevation angle ⁇ and the left-right rotation angle ⁇ of the projector. So can get The direction vector is Is the vector product of ⁇ vectors. According to the aspect ratio, we can get The length is r *
  • the method (S2) in Embodiment 1 can be used to obtain the distortion K of the projection screen.
  • the initial deformation K0 and the positions of the corner vertices of the projection screen P10P20P30P40, which are projected after the deformation, and the corner vertices (S1, S2, S3, The relative position of S4) is used to get the adjustment screen deformation K1, the point seen in the monitoring unit is P11P21P31P41 and then readjusted to K2 according to the relative position relationship between P11P21P31P41 and (S1, S2, S3, S4); repeat until the deformation K
  • the corner vertices P1nP2nP3nP4n of the n rear projection screen and the corner vertices (S1, S2, S3, S4) of the identified projection screen coincide or the error is less than a certain distance.
  • corner points of a projected image cannot be identified, the boundary points of the projected image can be considered to be processed outside the projection screen.
  • Iterative method that is, the correction of multiple rounds of approximation, generally adopts the reverse principle, taking the point P11 of the projection imaging chip as an example.
  • P11 performs deformation Ki it is P11i
  • P1i the point on the monitoring imaging chip
  • the corresponding corner vertex of the projection screen is S1.
  • S1i the point on the monitoring imaging chip
  • S1i The corresponding corner vertex of the projection screen
  • S1i the it is defined as the center of P1i
  • the iteration step size on the projection imaging chip is defined as (dxi, dyi); when S1 falls on different regions, the adjustment method is as follows:
  • step size (dxi, dyi) needs to select an appropriate step size according to the distance between P1i and S1.
  • the basic principle is that the smaller the distance, the smaller the step size, and the larger the distance, the larger the step size.
  • Embodiment 1 Various methods can be used to iterate the initial deformation. For example, when all four recognition points are identified, the calculated transformation K in Embodiment 1 is used as the initial transformation. When the error of the adaptive projection screen is larger than a preset threshold, Continue to iterate according to this method until the error is less than a preset threshold.
  • a small proportion (such as a quarter) of the projection screen received in the initial transformation is generally used to make the projection screen all projected as much as possible.
  • the projected picture can be deformed to the center of the projected picture in extreme cases.
  • it cannot be transformed into a projection screen that is too small, because the projection screen is too small, which is not conducive to the identification of the projection screen boundary.
  • the iterative process When the compensation setting of each step of the iteration is relatively small, the iterative process will become longer, but in effect, the projection screen gradually shrinks to the projection screen and gradually expands from the interior of the projection screen to the full projection screen.
  • the iteration step size When the iteration step size is set too large, it will often be adjusted too much during the iteration. From the morphological point of view, the entire process will occur. The projection image will alternately oscillate between the projection screen and the outside of the projection screen. The entire process also takes a long time. time. Therefore, to adjust the duration to be shorter, it is important to set reasonable fall compensation.
  • FIG. 7A it is an iterative method from the inside to the outside.
  • the projection picture is reduced to a smaller shape, and then iteratively is gradually performed until the boundary point of the projection picture and the projection target identification point reach a predetermined close. Level before completing the entire iterative process.
  • FIG. 7B it is an iterative method from the outside to the first, starting from the initial position of the projection screen, and then iterating inward gradually until the boundary point of the projection screen and the projection target identification point reach a predetermined close level To complete the entire iteration process.
  • mapping is performed first. If the boundary point of the projection screen and the identification point of the projection target do not reach a predetermined close level, then iterate so that the boundary point of the projection screen and the identification point of the projection target reach a predetermined close level. As shown in Figures 7C and 7D.
  • a part of the corner points of the projection screen is beyond the range of the projection area, and the projection screen has two identification points outside the quadrangle of the projection screen and recognized within the range of the monitoring unit. In this position, the projection screen cannot cover the entire projection screen, and only a part of the projection screen can be selected as the deformed projection range to achieve the display effect on the projection screen.
  • the specific method is as follows. First, calculate the maximum ratio that the projection picture can reach on the projection screen. From this ratio, calculate the position of the vertex of the largest similar quadrilateral box in the monitoring unit. Calculate the projection distortion.
  • step 2 when the four corner vertices of the projection screen and the four vertices of the projection screen are identified, if there are other edges outside the projection range, the other projection screen areas and intersection vertices (such as D1D2D3D4D5) shown in FIG. 4B is added to step 1, and the largest inscribed square (VM1, VM2, VM3, VM4) in step 2 is reconstructed.
  • the other projection screen areas and intersection vertices such as D1D2D3D4D5
  • the monitoring unit cannot monitor all the points of the projection screen, which makes it impossible to find the maximum ratio of the projection image to the projection screen, so it is impossible to calculate the subsequent N1N2N3N4 monitoring imaging chip. Location. In this case, it is necessary to calculate the spatial coordinate information of the corner vertices D1D2D3D4D5 at the intersection of the projection screen region and the projection screen region according to the method in Embodiment 2.
  • D4D5 and D4D3 are two Edge and calculate the unit direction vector of the two edges D1D2D3D4D5 are all on the projection screen, so there are
  • (ai, bi) is the coordinates in the coordinate system on the projection screen.
  • the internal parameters of the space are converted according to the internal parameters of the projection unit.
  • the points are mapped onto the projection imaging chip as N11N22N33N44, so as to obtain the distortion K of the projection picture.
  • the monitoring unit recognizes two opposite sides of the projection screen.
  • WN1, WN2, WN3, WN4 is selected as the most suitable projection screen area within the region, and then calculated as the spatial coordinates WN1, WN2, WN3, WN4, the position N11N22N33N44 in the projection imaging chip of WN1, WN2, WN3, WN4 in the projection imaging chip is calculated according to the internal parameters of the projection unit, and the deformation K of the projection screen.
  • the methods in the above implementation cases are all implementation methods that are established when the position of the projector does not move. This embodiment explains that if there is an automatic projection position control unit, some projection screens can be processed by controlling the position of the projector. Identify the situation where the point is outside the projection screen.
  • the projector can be controlled to rotate to the left or to move backward, so that the projection screen S1S2S3S4 seen by the monitoring unit moves to the right, thereby allowing S1S2S3S4 to fall into the quadrangular (P1, P2, P3, P4) range.
  • the method in Embodiment 1 can be further used to make the projection screen completely fit the projection screen, instead of the deformed projection screen in Embodiment 4 being smaller than the size of the projection screen.
  • the projector can be controlled to rotate to the left, or to move backwards and downwards, so that the projection screen S1S2S3S4 seen by the monitoring unit moves to the right and moves downwards, thereby allowing S1S2S3S4 to fall into the quadrilateral (P1, P2, P3, P4)
  • the method in Embodiment 1 can be further used to make the projection screen completely fit the projection screen, instead of the deformed projection screen in Embodiment 4 being smaller than the size of the projection screen.
  • the projection screen S1S2S3S4 seen by the monitoring unit moves to the right, and the distance ratio of S1S2 is smaller than the distance ratio of S3S4 when rotated.
  • the projection screen S1S2S3S4 seen by the monitoring unit moves to the left, and the distance ratio of S1S2 is larger than the distance ratio of S3S4 when rotated.
  • the projection screen S1S2S3S4 seen by the monitoring unit moves upward, and the distance ratio of S2S3 becomes larger than the distance ratio of S4S1 when rotated.
  • the projection screen S1S2S3S4 seen by the monitoring unit moves downward, and the distance ratio of S2S3 is smaller than the distance ratio of S4S1 when rotated.
  • prompts can be given on the projector according to the above rules to allow users, installers and other installation tips to adjust the position of the projector to achieve better projection results.
  • the boundary vertices of the projection screen may be projected outside the projection screen. Due to the complexity of the actual environment outside the projection screen, the boundary vertices of the projection screen cannot be recognized by the monitoring unit. In order to obtain the corner vertices of the projection screen more easily and more stably, a special mark or pattern that can be easily identified can be projected on the fixed point position inside the projection screen, and then according to the relative position relationship between the position of the fixed point and the boundary apex of the projection screen The proportional relationship and the coordinate position of the identification point in the monitoring imaging chip are used to calculate the position of the projection picture boundary point on the monitoring imaging chip.
  • (M1, M2, M3, M4) are internal identification points recognized by the monitoring unit, and (M11, M22, M33, M44) are coordinates of the identification point on the projection imaging chip. Therefore, we can construct a mapping that transforms Q to (M11, M22, M33, M44) to (M1, M2, M3, M4). Therefore, by substituting (P11, P22, P33, P44) into the transformation Q, the coordinates (P1, P2, P3, P4) of the projection screen boundary point on the monitoring imaging chip can be obtained. Thereby, the subsequent deformation can be further obtained according to the method of the foregoing implementation example 1-66.
  • the transformation T can be used to directly convert (S1, S2, S3) from (M1, M2, M3, M4) to (M11, M22, M33, M44).
  • S4 is substituted into the transformation to obtain the position of the corner points of the projection screen on the projection imaging chip (S11, S22, S33, S44).
  • the deformation K of the projection pictures from (P11, P22, P33, P44) to (S11, S22, S33, S44) is obtained.

Abstract

本发明公开一种投影系统自动适配投影目标的方法，其中，投影系统在工作时按以下步骤进行自动适配处理：由监控单元对投影目标上设置的至少两个标识点进行识别、同时对投影画面进行识别，分别生成相应的投影目标标识点位置信息及投影画面位置信息；然后由图像控制单元根据前述位置信息，按预定规则向投影成像芯片发出控制指令，以对投影画面的边界进行校正，使所述投影画面的边界与所述投影目标标识点达到预定的接近水平。

Description

一种投影系统自动适配投影目标的方法 技术领域
本发明涉及投影机，更具体地说，涉及一种投影系统自动适配投影目标的方法。
背景技术
在投影机的日常使用中，经常碰到摆放问题，调整调校问题。随着人们对投影画面质量需求的提高，光栅屏这种能吸收环境光并增加投影画面对比度的投影屏幕越来越受欢迎。目前很多光栅屏采用的是固定安装模式，如何使投影画面很好地适配这种固定安装、固定比例的投影屏幕，一直没有良好的解决方案。
例如在家用场景中，如果投影机不是预调之后固定安装，而是使用时才拿出来摆放于桌面上，则此时即使投影机正对投影屏幕，也需要反复多次调整投影机的位置，才能确保投影画面正好与投影屏幕适配，例如正好显示在投影屏幕的边框范围之内。
另一种情况是投影机不在投影屏幕的正前方，例如摆放于卧室的床头柜上，投影机的光轴与投影屏幕之间呈斜角关系，此时投影画面与投影屏幕之间的适配问题会更为复杂。
发明内容
针对现有技术的上述缺陷，本发明要解决的技术问题在于现有技术上投影机不能自动适配投影屏幕的问题。
为解决上述技术问题，本发明提供一种投影系统自动适配投影目标的方法，所述投影系统包括具有投影成像芯片的投影单元、具有监控成像芯片的监控单元、以及用于控制所述投影单元的图像控制单元；其中，所述投影系统正常工作时，按以下步骤进行自动适配处理：
步骤S1、所述监控单元对投影目标上设置的至少两个与投影目标主体有差异的投影目标标识点进行识别、同时对所述投影单元投射到所述投影目标上的投影画面进行识别，并由所述监控成像芯片分别生成相应的投影目标标识点位置信息及投影画面位置信息；
步骤S2、所述图像控制单元根据所述投影目标标识点位置信息及投影画面位置信息，按预定规则向所述投影成像芯片发出控制指令，以对所述投影画面的边界进行校正，使所述投影画面的边界与所述投影目标标识点达到预定的接近水平。
本发明所述的方法中，可以只执行一遍所述步骤S1、步骤S2，通过一次性校正使所述投影画面的边界与所述投影目标标识点达到预定的接近水平。
本发明所述的方法中，也可以重复执行所述步骤S1、步骤S2，进行多次循环逼近的校正，直到所述投影画面的边界与所述投影目标标识点达到预定的接近水平。
本发明所述的方法中，在进行所述多次循环逼近的校正时，可采取由内向外的方式，先以所述投影画面的中心点为中心生成一个最小投影画面，再以该最小投影画面为起点逐步增大，直到所述投影画面的边界与所述投影目标标识点达到预定的接近水平。
本发明所述的方法中，在进行所述多次循环逼近的校正时，也可采取由外向内的方式，以当前投影画面为起点逐步减小，直到所述投影画面的边界与所述投影目标标识点达到预定的接近水平。
本发明所述的方法中，所述投影目标标识点与所述投影目标主体具有明显的颜色差异；所述投影目标可为矩形投影屏幕，所述投影目标标识点可为所述矩形投影屏幕相邻角部的点；或者，所述投影目标可为墙壁，所述投影目标标识点可为所述墙壁上水平和/或垂直方向排列的点。
本发明所述的方法中，当所述监控单元在所述投影目标上仅识别出两个所述投影目标标识点时，在所述步骤S2中，以所述两个投影目标标识点的连线作为投影画面的一个边界，按照预定的投影画面比例生成完整的投影画面。
所述预定投影画面比例可以是16:9或4:3。此时，在所述步骤S2之前，还可包括对投影距离与角度进行识别的步骤；在所述步骤S2中，还结合所识别出的距离与角度进行补偿计算，按照预定的投影画面比例生成矩形投影画面。还可在所述两个投影目标标识点之间设置与所述投影目标主体具有明显颜色差异的连线。
本发明所述的方法中，当所述监控单元在所述投影目标上仅识别出三个所述投影目标标识点时，在所述步骤S2中，以所述三个投影目标标识点的接近垂直及水平方向的连线作为投影画面的两个边界，再基于所述两个边界生成完整的投影画面。
同样，在所述步骤S2之前，还可包括对投影距离与角度进行识别的步骤；在所述步骤S2中，还结合所识别出的距离与角度进行补偿计算，以生成矩形投影画面。还可在所述三个投影目标标识点之间设置与所述投影目标主体具有明显颜色差异的垂直及水平方向的两条连线，所述两条连线相互垂直。
本发明所述的方法中，当所述监控单元在所述投影目标上识别出四个所述投影目标标识点时，在所述步骤S2中，以所述四个投影目标标识点的接近垂直及水平方向的连线作为投影画面的四个边界进行校正，以使所述投影画面的边界与所述四个边界达到预定的接近水平。
还可在所述四个投影目标标识点之间设置与所述投影目标主体具有明显颜色差异的垂直及水平方向的四条连线，所述四条连线围成一个矩形。常规带有黑边框的投影屏幕就属于这种状态，四个角部是前述四个投影目标标识点，黑边框即为四条连线。
在使用本发明所述的方法进行自动适配处理过程中，首次执行所述步骤S1时，所述投影画面为所述投影单元监的可生成的初始投影画面、或最大投影画面；在首次之后继续执行所述步骤S1时，所述投影画面为上一次校正后的当前投影画面。
在使用本发明所述的方法进行自动适配处理过程中，首次执行所述步骤S1时，如果未能识别出至少两个所述投影目标标识点，或识别出的至少两个所述标识点没有位于所述投影画面内，则发出调整提示，提示用户需调整所述投影系统与投影目标之间的位置关系，确保至少两个所述投影目标标识点落在所述监控单元的监控范围之内，且落在所述投影画面内。
在使用本发明所述的方法所述步骤S1中，所述监控单元对所述投影单元投射到所述投影目标上的投影画面进行识别时，是通过拍摄图片再经图像识别方法得出所述投影画面的边界，或者通过在所述投影画面内显示便于识别的图案或标记点来得出所述投影画面的边界。
由于采取了以上技术方案，本发明的投影系统可以自动适配投影目标(包括投影屏幕、墙壁、或其他可用于显示投影画面的平面)，在确保至少两个投 影目标标识点落在监控单元的监控范围之内且落在投影画面内的前提下，会根据识别出的标识点的不同情况(两个、三个、四个)，分别使用不同的方法进行自动适配调整，以使投影画面处于最佳显示状态；在四个标识点的情况下则投影画面会充满这四个标识点限定的四边形空间，针对投影屏幕则是充满投影屏幕边框所限定的矩形空间。这种自动适配方法可省去人工调整的繁琐，大大增加投影机的使用便利性，尤其适用于家用投影机产品。
附图说明
下面将结合附图及实施例对本发明作进一步说明，附图中：
图1A是本发明实施例1中监控范围、投影画面、投影屏幕之间的位置关系及其边角点投射示意图；
图1B是本发明实施例1的监控成像芯片中的坐标与投影成像芯片中坐标之间的对应关系示意图；
图2A是本发明实施例2中的深度计算示意图；
图2B是本发明实施例2中通过距离和角度进行调整的示意图；
图3A是本发明实施例3中的通过迭代法进行调整的示意图；
图3B是本发明实施例3中的左下角屏幕交点在投影左下角点的相对位置的分区示意图；
图4A是本发明实施例4中投影屏幕有两个边界点在投影画面之外的示意图；
图4B是本发明实施例4中投影屏幕有三个边界点在投影画面之外的示意图；
图5A是本发明实施例5中投影屏幕有两个边界点在监控范围之外的监控 镜头视角示意图；
图5B是本发明实施例5中投影屏幕有两个边界点在监控范围之外的投影屏幕垂直方向视角示意图；
图5C是本发明实施例5中只识别到投影屏幕的两条相邻边或者断裂的两条相邻边时监控镜头视角的示意图；
图6A是本发明实施例6中只识别到投影屏幕的两条对边时监控镜头视角的示意图；
图6B是本发明实施例6中只识别到投影屏幕的两条对边时投影屏幕垂直方向视角示意图；
图7A是本发明实施例3中由内向外迭代的示意图；
图7B是本发明实施例3中由外向内迭代的示意图；
图7C是本发明实施例3中先映射、再由内向外迭代的示意图；
图7B是本发明实施例3中先映射、由外向内迭代的示意图；
图8是本发明实施例8中内部标识点估算边界顶点位置示意图。
具体实施方式
为了描述简单起见，以下实施案例中没特殊说明的投影画面的边角顶点都是指未形变的投影画面的边角顶点，监控单元采集的图像与投影机投出的图像同方向，也就是说，投影画面左上边界在监控单元所得图像的左上角方向，投影画面右上边界在监控单元所得图像的右上角方向，投影画面左下边界在监控单元所得图像的左下角方向，依此类推。
实施例1，直接映射法
如图1A所示，该投影系统包括具有投影成像芯片的投影单元、具有监控成像芯片的监控单元、以及用于控制所述投影单元的图像控制单元。该投影系统与投影目标之间可以任意摆放，本实施例为略微倾斜摆放；其中的投影目标可以是投影屏幕、墙壁、或其他可用于显示投影画面的平面)。
在图1A中，监控单元能拍摄到范围是四边形(WC1,WC2,WC3,WC4)，投影画面的四个边角顶点是(WP1,WP2,WP3,WP4)，投影屏幕的四个边角顶点是(WS1,WS2,WS3,WS4)。从图中可以看出，这三者的面积依次变小，监控单元能拍摄到范围最大、投影画面次之、投影屏幕最小。此时，投影画面的四个边角顶点、投影屏幕的四个边角顶点都在四边形(WC1,WC2,WC3,WC4)的范围内，也就是都能被监控单元识别到。
初始状态时，投影机会将投影成像芯片上未变形的投影图像边界点(P11,P22,P33,P44)投射到投影屏幕，即得到前述投影画面的四个边角顶点(WP1,WP2,WP3,WP4)。
投影画面的四个边角顶点(WP1,WP2,WP3,WP4)和投影屏幕的四个边角顶点(WS1,WS2,WS3,WS4)在监控单元的监控成像芯片上会成像为投影画面监控边界点(P1,P2,P3,P4)和投影屏幕监控边界点(S1,S2,S3,S4)。
如图1B所示，为了将投影图像边界点(P11,P22,P33,P44)投射到投影屏幕的四个边角顶点(WS1,WS2,WS3,WS4)上，投影控制单元需要对投影画面做形变处理，此形变处理的作用是在投影成像芯片上将投影图像边界点(P11,P22,P33,P44)映射到(S11,S22,S33,S44)，然后由投影机将(S11,S22,S33,S44)投射到投影屏幕上时，以使所得投影画面的边角顶点恰好在投影屏幕的四个边角顶点(WS1,WS2,WS3,WS4)上，从而使投影画面完全充满投影屏幕，以达到最好的视觉效果。此变形一般使用透视变换公式(1)来 实现的。
Figure PCTCN2018101505-appb-000001
其中(x,y)透视变换前坐标，(x',y')为透视变换后坐标。
在图1B中，监控单元光心和投影单元光心间距很近，可以近似认为，投影单元光心和监控单元光心是在同一个点，这个点就是透视中心，因此在同一固定场景下监控单元中监控成像芯片上的点和投影单元中投影成像芯片上的点存在一一对应的关系。
(S1)可以构造变换T，将监控成像芯片上的点(P1,P2,P3,P4)映射到投影成像芯片上的点(P11,P22,P33,P44)，然后计算出监控成像芯片上投影屏幕监控边界点(S1,S2,S3,S4)在投影成像芯片上的坐标(S11,S22,S33,S44)。
(S2)可以构造变换K，将原始未变形的投影图像边界点(P11,P22,P33,P44)映射到投影屏幕边界点在投影成像芯片上对应的点(S11,S22,S33,S44)，进而得到原始投影画面中任一点在变形后投影成像芯片上的新坐标。假设投影单元分辨率为(W*H)：则一般有P11＝(H-1,0)，P22＝(0,0)，P33＝(0,W-1)，P44＝(H-1,W-1)。
其中，投影画面的四个边角顶点(WP1,WP2,WP3,WP4)可通过识别投影画面的边界线或边界线的交点得到，也可以通过投影机打出特定的标识物，再去识别这个特殊的标识物，然后根据标识物在投影成像芯片上的坐标去推算出投影画面的四个边角顶点(WP1,WP2,WP3,WP4)所在的位置。
以上实施例对应于发明内容中所述的监控单元可在投影目标上识别出四 个投影目标标识点时的情况。
实施例2，角度距离法
当监控单元只识别到了投影目标的两个标识点、或者一条边线的两个端点时，可以结合宽高比例、左右旋转角、上下俯仰角、投影单元内部参数、监控单元内部参数、投影屏幕标识点到投影机的深度等信息，来推算其他的投影屏幕边角顶点在投影成像芯片上的坐标，进而计算得到投影画面的变形K。
下面以只识别到两个标识点S1、S2为例说明利用距离角度的方法。首先需要得到两个标识点S1、S2距离投影机的深度Z1、Z2。这个可以通过外部设备获取或者根据其他信息如估算得到。如图2A所示，投影成像芯片上的点经过投影单元P投射到点S上，被监控单元C看到。利用实施例1中的(S1)方法计算出点S1在投影成像芯片上的坐标S11，因此根据投影单元的内部参数可以计算得到的投射方向向量
Figure PCTCN2018101505-appb-000002
同时根据监控单元的内部参数，可以得到接收光线的方向向量
Figure PCTCN2018101505-appb-000003
联立两方程可以得到方程1，此点的深度为：
Figure PCTCN2018101505-appb-000004
Figure PCTCN2018101505-appb-000005
(f p*x S11-C xp)*L＝(f c*x S1-C xc)*L+D  (方程1)
Figure PCTCN2018101505-appb-000006
其中f p为投影单元的焦距，(C xp,C yp)为投影单元光心在投影成像芯片上的位置， (x S11,y S11)为S11在投影成像芯片上的坐标，f c为监控单元的焦距，(C xc,C yc)为监控单元光心在监控成像芯片上的位置，(x S1,y S1)为S1在监控成像芯片上的坐标。所以可以得到WS1相对于对于投影机的空间位置为
Figure PCTCN2018101505-appb-000007
同理可以得到WS2相对于投影机的空间坐标为
Figure PCTCN2018101505-appb-000008
简记为(X1,Y1,Z1)(X2,Y2,Z2)
其次，以上计算出此边线的长度
Figure PCTCN2018101505-appb-000009
和方向向量
Figure PCTCN2018101505-appb-000010
并根据投影机的俯仰角α和左右旋转角β，计算得到投影屏幕的平面在投影机中的法线方向向量
Figure PCTCN2018101505-appb-000011
因此可以得到的
Figure PCTCN2018101505-appb-000012
方向向量为
Figure PCTCN2018101505-appb-000013
为×向量的向量积。根据宽高比例关系可以得到
Figure PCTCN2018101505-appb-000014
的长度为r*|WS1WS2|，继而计算得到WS3，WS4在投影机种的空间坐标为。
Figure PCTCN2018101505-appb-000015
最后得到WS3在成像芯片上的坐标SS3＝(X3/(f p*Z3)+C xp,Y3/(f p*Z3)+C yp)，和WS4在成像芯片上的坐标为SS4＝(X4/(f p*Z4)+C xp,Y4/(f p*Z4)+C yp)。进一步利用实施例1中的(S2)方法可以得到投影画面的形变K。
以上实施例相当于发明内容中所述的监控单元仅在投影目标上识别出两个投影目标标识点时的情况。
实施例3，迭代法
如图3A所示，可以不采用计算方式，通过设置初始的形变K0和形变完后投射出的投影画面边角顶点的位置P10P20P30P40，和识别得到的投影屏幕上边角顶点(S1,S2,S3,S4)的相对位置来得到调整画面形变为K1，在监控单元看到的点为P11P21P31P41然后根据P11P21P31P41和(S1,S2,S3,S4)的相对位置关系重新调整为K2；重复多次直到形变K n后投影画面的边角顶点P1nP2nP3nP4n和识别到的投影屏幕的边角顶点(S1,S2,S3,S4)重合或误差小于一定距离为止。当有投影图像的边角点无法被识别，可以认为该投影图像的边界点在投影屏幕外部来处理。
迭代法，也就是多次循环逼近的校正，一般采取反向原则，以投影成像芯片的点P11为例说明，当P11执行变形Ki后为P11i，投影出来后在监控成像芯片上点为P1i，投影屏幕对应边角顶点为S1。如图3B所示，定义为P1i中心，并且定义投影成像芯片上的迭代步长为(dxi,dyi)；当S1落在不同各区域上时的调整方法按以下规律：
(1)、1区；S1落在1区，则P1i落在投影屏幕区域左内外下侧，因此需要投影画面向左上侧移动，则有下一步变形(x P1(i+1),y P1(i+1))＝(x P1i-dxi,y P1i-dyi)；
(2)、2区；S1落在2区，则P1i落在投影屏幕区域左外下侧，因此需要将投影画面能够往右上侧移动，则有下一步变形(x P1(i+1),y P1(i+1))＝(x P1i-dxi,y P1i+dyi)；
(3)、3区；S1落在3区，则P1i落在投影屏幕区域左内上侧，因此将投影画面能够往左下侧移动移动，则有下一步变形(x P1(i+1),y P1(i+1))＝(x P1i+dxi,y P1i-dyi)；
(4)、4区；S1落在3区，则P1i落在投影屏幕区域左外上侧，因此将投影画面能够往右下侧移动移动，则有下一步变形(x P1(i+1),y P1(i+1))＝(x P1i+dxi,y P1i+dyi)；
(5)、1区2区的交线上，则P1i落在投影屏幕区域下侧，因此将投影画面 能够往上侧移动，则有下一步变形(x P1(i+1),y P1(i+1))＝(x P1i-dxi,y P1i)；
(6)、3区4区的交线上，则P1i落在投影屏幕区域上侧，因此将投影画面能够往下侧移动，则有下一步变形(x P1(i+1),y P1(i+1))＝(x P1i+dxi,y P1i)；
(7)、1区3区的交线上，则P1i落在投影屏幕区域左内侧，因此将投影画面能够往下侧移动，则有下一步变形(x P1(i+1),y P1(i+1))＝(x P1i,y P1i-dyi)；
(8)、2区4区的交线上，则P1i落在投影屏幕区域左外侧，因此将投影画面能够往下侧移动，则有下一步变形(x P1(i+1),y P1(i+1))＝(x P1i,y P1i+dyi)；
(9)、步长的选取(dxi,dyi)需要根据P1i和S1的距离来选取合适的步长，基本原则是，距离越小步长越小，距离越大步长越大。
类似的，根据不同顶点及其不同的相对位置关系给出不同的投影成像芯片的点迭代公式，获得这次迭代变形的投影成像芯片上的位置，进一步来获取投影成像芯片上任意点的形变公式。
迭代初始变形可以采用多种方法，例如在4个识别点都被识别到时，利用实施例1中的计算的变换K当作初始变换，当适配投影屏幕的误差大于预设阈值时，可以继续按照此方法继续迭代，直至误差小于预设阈值。
为了防止初始的投影画面的边角顶点在投影画面的背景中无法被识别，一般先将初始的变换内收到投影画面的一小比例(例如四分之一)来尽量使得投影画面全部在投影屏幕内，极端情况下可以让投影画面变形到投影画面的中心附近。但是不能形变成到过小的投影画面，因为投影画面过小将不利于投影画面边界的识别。
当迭代每步的补偿设置比较小时，迭代过程将变长，但效果上看，投影画面呈逐步收缩到投影屏幕和从投影屏幕内部慢慢扩充到满投影屏幕的形态。迭代步长设置很大时，会导致迭代时会经常调整过头，从形态上看，整个过程会 出现，投影画面在投影屏幕内部和投影画面外部来回交替变化的震荡形态，整个过程也需要很长的时间。因此要调整时长变短，设置合理的跌打补偿很重要。
如图7A所示，是一种由内向外的迭代方法，首先将投影画面缩小到一个较小的形状，然后逐步向外进行迭代，直到投影画面的边界点与投影目标标识点达到预定的接近水平，才完成整个迭代过程。同样，如图7B所示，是一种由外向内的迭代方法，首先从投影画面的初始位置开始，然后逐步向内进行迭代，直到投影画面的边界点与投影目标标识点达到预定的接近水平，才完成整个迭代过程。
本实施例还有一种情况是先进行映射，如果投影画面的边界点与投影目标标识点没有达到预定的接近水平，再进行迭代，使投影画面的边界点与投影目标标识点达到预定的接近水平，如图7C、图7D所示。
此实施例中可以复合其他的实施例中描述的方法来获取其初始变换或其中的某些步骤的变换。
实施例4
在投影屏幕标识点中部分点无法被识别或者超出投影区域范围时，需要根据实际情况推算或者设置合理的边角位置。
如图4A所示，投影屏幕有一部分的边角点超出了投射区域的范围，投影屏幕有两个标识点在投影画面的四边形外并且在监控单元范围内被识别。在这种摆放位置下投影画面无法覆盖整个投影屏幕，只能选取投影屏幕的部分来做变形后的投射范围，用以达到在投影屏幕中的显示效果。具体方法如下，首先计算投影画面在投影屏幕中能达到的最大比例，通过这个比例计算出最大相似四边形框的顶点在监控单元的位置，再根据实施例1中(S1)和(S2)的方法 计算出投影形变。
步骤1，计算投影画面在投影屏幕中能达到最大比例的方法为：获取到投影屏幕两个边S2S3，S1S4与投影画面边界线P1P2的交点D2D1，建立映射变形F将(S1,S2,S3,S4)映射到虚拟单位正方形的对应四个顶点(V1,V2,V3,V4)上，其中V1＝(1,0)，V2＝(0,0)，V3＝(0,1)，V4＝(1,1)。计算出在虚拟单位正方形中的坐标VD1＝(x1,y1),VD2＝(x2,y2)。因此可以计算出最大比例为k＝1.0-max(x1,y1,y2)。
步骤2，计算投影屏幕最大相似框的顶点在监控单元中的坐标的方法为：构造映射变形F -1，将(V1,V2,V3,V4)，映射到(S1,S2,S3,S4)，并构造最大内接正方形(VM1,VM2,VM3,VM4)在虚拟单位正方形中的坐标VM1＝(1,k)，VM2＝(k,k)，VM3＝(k,1)，VM4＝(1,1)，映射到监控成像芯片上的坐标N1N2N3N4，同时利用N1N2N3N4替换(S1,S2,S3,S4)根据实施例1的方法，得到最终得到投影画面的形变。
更一般的在识别到投影画面的四个边角顶点和投影屏幕的四个顶点时，如果有其他的边也在投影范围之外，需要把其他的投影画面区域和交集的边角顶点(如图4B中所示的D1D2D3D4D5)加入到步骤1中，并且重新构造步骤2中的最大的内接正方形(VM1,VM2,VM3,VM4)。
实施例5
如图5A所示，与实施例中的4B有点类似，监控单元无法监控到投影屏幕的全部点，导致无法求出投影图像占投影屏幕的最大比例，所以无法计算出后续的N1N2N3N4在监控成像芯片中的位置。在此情况下需要按实施例2中的方法，计算出投影画面区域和投影屏幕区域交集的边角顶点D1D2D3D4D5的空间坐标信息。
按照实施例2中的方法计算出D1D2D3D4D5的空间坐标为(Xi,Yi,Zi),i＝1:5，从其中选取相邻的投影屏幕的边，图5A中为D4D5和D4D3，这两条边并且计算两条边的单位方向向量
Figure PCTCN2018101505-appb-000016
D1D2D3D4D5都在投影屏幕上，所以有
Figure PCTCN2018101505-appb-000017
其中(ai,bi)，为投影屏幕上的坐标系下的坐标。选取其中最合适的按照设置的固定长宽比例的矩形的顶点并换算到摄像机空间中的坐标WN1,WN2,WN3,WN4，按照实施例2中的方法在通过投影单元的内部参数，将空间的点映射到投影成像芯片上为N11N22N33N44，从而得到投影画面的形变K。
同理这种也可以处理如图5C所示的情况，有些投影屏幕的边角点无法被识别，仅被识别到两条相邻边，或者两段不相交的边(延长两条不相交的边可以相交到投影屏幕的边角顶点)，只是选取最合适的按照固定比例的位置方式不同。
实施例6
如图6A所示，监控单元识别到投影屏幕两条对边。采用与实施例5相似的方法，先在监控画面中找到投影屏幕的边角顶点四边形S1S2S3S4与投影画面四边形(P1,P2,P3,P4)的交集的多边形的边角顶点D1D2D3D4D5。
按照实施例2中的方法计算出交集多边形的顶点的空间坐标(Xi,Yi,Zi),i＝1:5，取投影屏幕边界的一个顶点为基准点以为D1例，计算出的单位方向向量
Figure PCTCN2018101505-appb-000018
因为识别的或上仰，斜投等原因，可能导致这两个向量不一定垂直，需要构造两个相互垂直的单位方向向量
Figure PCTCN2018101505-appb-000019
其中
Figure PCTCN2018101505-appb-000020
Figure PCTCN2018101505-appb-000021
其它都可以分解为
Figure PCTCN2018101505-appb-000022
则有(ai,bi)，是各个点在投影屏幕空间的坐标，在其中选取WN1,WN2,WN3,WN4，为区域内部最合适的投影屏幕区域，进而计算为空间坐标WN1,WN2,WN3,WN4，在根据投影单元的内部参数计算出WN1,WN2,WN3,WN4在投影成像芯片中的位置N11N22N33N44，投影画面的形变K。
实施例7
以上实施案例的方法都是建立在投影机位置不动的情况下的实施方法，本实施例说明的是如果有自动投影位置控制单元的情况下，可以借助控制投影机的位置来处理一些投影屏幕识别点在投影画面外的情况。
针对图4A的情况，可以控制投影机向左旋转，或者向后移动，使得监控单元看到的投影屏幕S1S2S3S4向右移动，进而让S1S2S3S4落入到的四边形(P1,P2,P3,P4)范围内，然后进一步可以使用实施例1中的方法，使投影画面完全贴合投影屏幕，而不是实施例4中的变形后的投影画面小于投影屏幕的大小。
针对图4B的情况，可以控制投影机向左旋转，或者向后移动和向下旋转，使得监控单元看到的投影屏幕S1S2S3S4向右移动向下移动，进而让S1S2S3S4落入到的四边形(P1,P2,P3,P4)范围内，进一步可以使用实施例1中的方法，使投影画面完全贴合投影屏幕，而不是实施例4中的变形后的投影画面小于投影屏幕的大小。
下面总结下摄像机移动时，监控单元看到的投影目标的移动规律：
当投影机向左旋转或移动时，监控单元看到的投影屏幕S1S2S3S4向右移动，并且旋转时S1S2的距离比S3S4的距离比例会变小。
当投影机向有旋转或移动时，监控单元看到的投影屏幕S1S2S3S4向左移动，并且旋转时S1S2的距离比S3S4的距离比例会变大。
当投影机向下旋转或移动时，监控单元看到的投影屏幕S1S2S3S4向上移动，并且旋转时S2S3的距离比S4S1的距离比例会变大。
当投影机向上旋转或移动时，监控单元看到的投影屏幕S1S2S3S4向下移动，并且旋转时S2S3的距离比S4S1的距离比例会变小。
当投影机向后移动时，监控单元看到的投影屏幕S1S2S3S4向中心收缩。
当投影机向前移动时，监控单元看到的投影屏幕S1S2S3S4向外扩散。
其他不在示意图中的情况，都可以参照以上移动规律，尽量将投影屏幕画面调整到未变形的投影画面内，让变形后的投影能很好的贴合被识别到的标记点。
在没有投影位置自动控制单元的情况下，可以根据以上规律在投影机上给出提示，来让用户、安装者等安装提示调整投影机的位置，来实现更好的投影效果。
实施例8
在实际使用过程中，投影画面的边界顶点有可能投射到投影屏幕的外面，由于投影屏幕外面实际环境的复杂性，会导致投影画面的边界顶点无法被监控单元正常识别到。为了更容易和更稳定的得到投影画面的边角顶点，可以在投影画面内部的固定点位置上投射出易于被识别特殊标识或图案，然后根据固定点位置和投影画面边界顶点的相对位置关系、比例关系，以及标识点在监控成像芯片中的坐标位置，来计算出投影画面边界点在监控成像芯片上的位置。
与实施例1相似，如图8所示，(M1,M2,M3,M4)为监控单元识别的内部标 识点，(M11,M22,M33,M44)为标识点在投影成像芯片上的坐标，因此可以构建变换Q为(M11,M22,M33,M44)到(M1,M2,M3,M4)的映射。因此将(P11,P22,P33,P44)代入变换Q就可以得到投影屏幕边界点在监控成像芯片上的坐标(P1,P2,P3,P4)。从而可以进一步按照以上实施示例1-66的方法得到后续的变形。
特别的是，如果在实施例1中采用以上内部标识点时，可以利用变换T，为(M1,M2,M3,M4)到(M11,M22,M33,M44)直接将(S1,S2,S3,S4)代入到变换得到投影屏幕边角点在投影成像芯片上的位置(S11,S22,S33,S44)。进而得到(P11,P22,P33,P44)到(S11,S22,S33,S44)的投影画面的形变K。
应当理解的是，以上实施例仅用以说明本发明的技术方案，而非对其限制，对本领域技术人员来说，可以对上述实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而所有这些修改和替换，都应属于本发明所附权利要求的保护范围。

Claims (18)

  1. 一种投影系统自动适配投影目标的方法，所述投影系统包括具有投影成像芯片的投影单元、具有监控成像芯片的监控单元、以及用于控制所述投影单元的图像控制单元；
    其特征在于，所述投影系统正常工作时，按以下步骤进行自动适配处理：
    步骤S1、所述监控单元对投影目标上设置的至少两个与投影目标主体有差异的投影目标标识点进行识别、同时对所述投影单元投射到所述投影目标上的投影画面进行识别，并由所述监控成像芯片分别生成相应的投影目标标识点位置信息及投影画面位置信息；
    步骤S2、所述图像控制单元根据所述投影目标标识点位置信息及投影画面位置信息，按预定规则向所述投影成像芯片发出控制指令，以对所述投影画面的边界进行校正，使所述投影画面的边界与所述投影目标标识点达到预定的接近水平。
  2. 根据权利要求1所述的投影系统自动适配投影目标的方法，其特征在于，其中只执行一遍所述步骤S1、步骤S2，通过一次性校正使所述投影画面的边界与所述投影目标标识点达到预定的接近水平。
  3. 根据权利要求1所述的投影系统自动适配投影目标的方法，其特征在于，其中重复执行所述步骤S1、步骤S2，进行多次循环逼近的校正，直到所述投影画面的边界与所述投影目标标识点达到预定的接近水平。
  4. 根据权利要求3所述的投影系统自动适配投影目标的方法，其特征在于，在进行所述多次循环逼近的校正时，先以所述投影画面的中心点为中心生成一个最小投影画面，再以该最小投影画面为起点逐步增大，直到所述投影画 面的边界与所述投影目标标识点达到预定的接近水平。
  5. 根据权利要求3所述的投影系统自动适配投影目标的方法，其特征在于，在进行所述多次循环逼近的校正处理时，以当前投影画面为起点逐步减小，直到所述投影画面的边界与所述投影目标标识点达到预定的接近水平。
  6. 根据权利要求1-5中任一项所述的投影系统自动适配投影目标的方法，其特征在于，所述投影目标标识点与所述投影目标主体具有明显的颜色差异；
    所述投影目标为矩形投影屏幕，所述投影目标标识点为所述矩形投影屏幕相邻角部的点；或者，所述投影目标为墙壁，所述投影目标标识点为所述墙壁上水平和/或垂直方向排列的点。
  7. 根据权利要求6所述的投影系统自动适配投影目标的方法，其特征在于，当所述监控单元在所述投影目标上仅识别出两个所述投影目标标识点时，在所述步骤S2中，以所述两个投影目标标识点的连线作为投影画面的一个边界，按照预定的投影画面比例生成完整的投影画面。
  8. 根据权利要求7所述的投影系统自动适配投影目标的方法，其特征在于，所述预定投影画面比例为16:9或4:3。
  9. 根据权利要求7所述的投影系统自动适配投影目标的方法，其特征在于，
    在所述步骤S2之前，还包括对投影距离与角度进行识别的步骤；
    在所述步骤S2中，还结合所识别出的距离与角度进行补偿计算，按照预定的投影画面比例生成矩形投影画面。
  10. 根据权利要求7所述的投影系统自动适配投影目标的方法，其特征在于，所述两个投影目标标识点之间设有与所述投影目标主体具有明显颜色差异的连线。
  11. 根据权利要求6所述的投影系统自动适配投影目标的方法，其特征在于，当所述监控单元在所述投影目标上仅识别出三个所述投影目标标识点时，在所述步骤S2中，以所述三个投影目标标识点的接近垂直及水平方向的连线作为投影画面的两个边界，再基于所述两个边界生成完整的投影画面。
  12. 根据权利要求11所述的投影系统自动适配投影目标的方法，其特征在于，
    在所述步骤S2之前，还包括对投影距离与角度进行识别的步骤；
    在所述步骤S2中，还结合所识别出的距离与角度进行补偿计算，以生成矩形投影画面。
  13. 根据权利要求11所述的投影系统自动适配投影目标的方法，其特征在于，所述三个投影目标标识点之间设有与所述投影目标主体具有明显颜色差异的垂直及水平方向的两条连线，所述两条连线相互垂直。
  14. 根据权利要求6所述的投影系统自动适配投影目标的方法，其特征在于，当所述监控单元在所述投影目标上识别出四个所述投影目标标识点时，在所述步骤S2中，以所述四个投影目标标识点的接近垂直及水平方向的连线作为投影画面的四个边界进行校正，以使所述投影画面的边界与所述四个边界达到预定的接近水平。
  15. 根据权利要求14所述的投影系统自动适配投影目标的方法，其特征在于，所述四个投影目标标识点之间设有与所述投影目标主体具有明显颜色差异的垂直及水平方向的四条连线，所述四条连线围成一个矩形。
  16. 根据权利要求1-5中任一项所述的投影系统自动适配投影目标的方法，其特征在于，在所述进行自动适配处理过程中，首次执行所述步骤S1时， 所述投影画面为所述投影单元监的可生成的初始投影画面、或最大投影画面；在首次之后继续执行所述步骤S1时，所述投影画面为上一次校正后的当前投影画面。
  17. 根据权利要求16所述的投影系统自动适配投影目标的方法，其特征在于，在所述进行自动适配处理过程中，首次执行所述步骤S1时，如果未能识别出至少两个所述投影目标标识点，或识别出的至少两个所述标识点没有位于所述投影画面内，则发出调整提示，提示用户需调整所述投影系统与投影目标之间的位置关系，确保至少两个所述投影目标标识点落在所述监控单元的监控范围之内，且落在所述投影画面内。
  18. 根据权利要求1所述的投影系统自动适配投影目标的方法，其特征在于，所述步骤S1中，所述监控单元对所述投影单元投射到所述投影目标上的投影画面进行识别时，是通过拍摄图片再经图像识别方法得出所述投影画面的边界，或者通过在所述投影画面内显示便于识别的图案或标记点来得出所述投影画面的边界。
PCT/CN2018/101505 2018-08-21 2018-08-21 一种投影系统自动适配投影目标的方法 WO2020037491A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2018/101505 WO2020037491A1 (zh) 2018-08-21 2018-08-21 一种投影系统自动适配投影目标的方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2018/101505 WO2020037491A1 (zh) 2018-08-21 2018-08-21 一种投影系统自动适配投影目标的方法

Publications (1)

Publication Number Publication Date
WO2020037491A1 true WO2020037491A1 (zh) 2020-02-27

Family

ID=69592391

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/101505 WO2020037491A1 (zh) 2018-08-21 2018-08-21 一种投影系统自动适配投影目标的方法

Country Status (1)

Country Link
WO (1) WO2020037491A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111953950A (zh) * 2020-08-14 2020-11-17 青岛海信移动通信技术股份有限公司 投影设备及其投影镜头的姿态调整方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1489382A (zh) * 2002-07-23 2004-04-14 日本电气视象技术株式会社 投影仪
CN1584729A (zh) * 2003-08-22 2005-02-23 日本电气株式会社 图像投影方法和设备
CN101697056A (zh) * 2009-10-27 2010-04-21 苏州巨像科技有限公司 具有画面自适应功能的智能摄像投影系统及其投影方法
CN102162979A (zh) * 2011-04-08 2011-08-24 广东威创视讯科技股份有限公司 投影仪投影图像自动调整方法及其装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1489382A (zh) * 2002-07-23 2004-04-14 日本电气视象技术株式会社 投影仪
CN1584729A (zh) * 2003-08-22 2005-02-23 日本电气株式会社 图像投影方法和设备
CN101697056A (zh) * 2009-10-27 2010-04-21 苏州巨像科技有限公司 具有画面自适应功能的智能摄像投影系统及其投影方法
CN102162979A (zh) * 2011-04-08 2011-08-24 广东威创视讯科技股份有限公司 投影仪投影图像自动调整方法及其装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111953950A (zh) * 2020-08-14 2020-11-17 青岛海信移动通信技术股份有限公司 投影设备及其投影镜头的姿态调整方法

Similar Documents

Publication Publication Date Title
CN109151415B (zh) 一种投影系统自动适配投影目标的方法
CN106952311B (zh) 基于全景拼接数据映射表的辅助泊车系统及方法
US20180218485A1 (en) Method and apparatus for fusing plurality of depth images
CN110099266B (zh) 投影机画面校正方法、装置及投影机
JP2022528659A (ja) プロジェクタの台形補正方法、装置、システム及び読み取り可能な記憶媒体
CN114727081B (zh) 投影仪投影校正方法、装置及投影仪
JP6299234B2 (ja) 表示制御方法、情報処理装置、および表示制御プログラム
CN107689033B (zh) 一种基于椭圆分割的鱼眼图像畸变校正方法
CN105308503A (zh) 利用短程相机校准显示系统的系统和方法
CN106604003B (zh) 一种短焦投影实现曲面幕布投影的方法及系统
CN112734860B (zh) 一种基于弧形幕先验信息的逐像素映射投影几何校正方法
WO2018112898A1 (zh) 一种投影方法、装置及机器人
CN108737799A (zh) 一种投影方法、装置及系统
KR20140090775A (ko) 어안 렌즈를 사용하여 얻은 왜곡영상에 대한 보정방법 및 이를 구현하기 위한 영상 디스플레이 시스템
CN106886976B (zh) 一种基于内参数修正鱼眼像机的图像生成方法
WO2020037491A1 (zh) 一种投影系统自动适配投影目标的方法
JP6022423B2 (ja) 監視装置及び監視装置の制御プログラム
CN107346530A (zh) 一种修正鱼眼图像的投影方法及系统
CN104363421A (zh) 实现多角度摄像头监控效果的方法和装置
CN111353945B (zh) 鱼眼图像校正方法、装置及存储介质
WO2022062604A1 (zh) 投影画面调节方法、装置、投影仪和存储介质
WO2023015868A1 (zh) 图像背景生成方法及装置、计算机可读存储介质
KR20170020864A (ko) 이미지 처리 방법 및 카메라
JP2016142991A (ja) 画像処理システム、情報処理装置、画像処理方法、及びプログラム
CN107945104A (zh) 一种基于太空虚拟现实相机的全景成像方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18930914

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18930914

Country of ref document: EP

Kind code of ref document: A1