TWI616100B - Curved surface projection correction control method and its augmented reality application system - Google Patents

Abstract

The present invention provides a projection correction method for a curved surface, the method comprising: a projection unit projecting a correction image on a projection surface; an image capture unit capturing the correction image and defining a grayscale image range of the correction image Binarizing into a binarized image; dividing the binarized image into a background region and a projection region; defining a parallelogram range in the projection region as a conversion region, and defining a correction parameter by the conversion region Corresponding to correcting the image projected by the projection unit; thereby, the invention can correct the image projected by the projection unit, prevent the image from being distorted, and can conform to various projection surfaces, such as wall corners, ceiling seams or curved surfaces. The projection surface can correctly project the image, and the invention can be hung on the chest of the user, which is convenient for carrying, the convenience of the cookware and the universal applicability.

Description

Projection correction control method for curved surface and its augmented reality application system

The invention provides a projection correction control method for a curved surface and an augmented reality application system thereof, in particular, an image capable of correcting projection according to a projection surface of various shapes.

According to the technology of manipulating and interacting with objects or information in the digital world, there have been many achievements in the field of virtual reality (VR), in which the presentation of information is mainly through head-mounted displays. To meet the immediate needs of observers, but the real need for computer technology applications is not the digital world simulated by computers, or the objects or environments in the virtual world. The core need is how to provide digital information in the real world. Let people still feel that they are in the real world, not to adapt to the virtual world, which makes Augmented Reality (AR) come into being.

Although most of the technologies in the virtual reality field can be directly applied to augmented reality applications, how to correctly combine the real world and the digital world is the most important technical difference, such as: The posture of the virtual object is determined by the observer's posture; the corresponding digital information is projected on the real world object, etc., all of which involve the above technical topics. The source of human information is mainly through visual, auditory, olfactory, taste and touch. Body sensory receptors are received, in which the visual capsule contains 70% of the total body sensory receptors. Therefore, obtaining information through images is the most important and convenient way for human beings. Moreover, human beings are most sensitive to visual quality requirements. ,because This ideal expansion application system requires high-quality display capabilities to conquer the visual satisfaction of users.

In virtual reality applications, head-mounted displays (helmet-type or glasses-type) are used to obtain information, making information display simpler without technical thresholds, but it also allows for virtual reality applications. It is isolated from the real world and limits its relevance to daily life. Therefore, the main uses are mainly simulation training. However, technology will ultimately come from life and human needs. It is necessary to apply virtual reality technology to daily life, which will obviously make information The display becomes the threshold of technology. Because it needs to meet the needs of people's activities, it is impossible to limit people to the traditional computer screens, and to display information in the environment. Therefore, it is necessary to research and develop a kind of display that allows users to display directly in the living environment. The required information needs have become the primary problem to be solved in the development of augmented reality application platforms.

Augmented Reality uses the technology to calculate the position and angle of camera images in the field, and through the technology of image presentation, it has a wide range of applications, and can be applied to satellite, mobile devices, surgical medical, industrial and entertainment technologies, but it is used. Augmented reality platform, presented on your own portable screen (such as: smart phone, tablet, or computer screen), or projected to the desired location using the projector.

However, the portable screen cannot produce enough abundance of real-world presentation, so it is difficult to combine with the actual image. With the projector projector, thanks to the miniaturization technology, the micro-projector has been developed to make it light and easy to carry. It can meet the display range of personal use, but still needs to reflect the image through the projection surface, so it has the problem of geometric distortion of the projection surface, and the projection range can be taken by high-order photographic equipment, and the projection range can be any irregular wall and corner. , double curtains, brick walls, etc., and use the high-end photographic equipment to shoot the projection range, to correct the change of each pixel, in theory, although it can correct any projection surface, but it requires high-order photographic equipment to complete Correction, which requires a lot of cost, and if it can only produce a uniform size or standard rectangular display range like a conventional projector, it will cause the augmented reality information to lose its realism and reduce its applicability.

In view of this, our inventors are concentrating on further research on projection correction methods and amplification. The application system, and the development and improvement of the system, is designed to solve the above problems, and the invention has been continuously tested and modified.

The reason is that the purpose of the present invention is to solve the problem that the portable portable screen cannot generate sufficient abundance of reality, so it is difficult to combine with the actual image; if the projector is projected through the projector, the geometric distortion of the projection surface often needs to be utilized. High-end photographic equipment shoots a range of projections, which requires a large amount of cost, and can only produce a uniform size or a standard rectangular display range like a conventional projector, which will result in the loss of realism and reduce the applicability of the augmented reality information.

In order to achieve the above object, the inventors provide a projection correction method for a curved surface, the method comprising: (a) a projection unit projecting a first curved surface correction image on a projection surface, and the image capturing unit captures the image a first surface correction picture; (b) the projection unit projects a pair of second surface correction pictures that should be formed on the first surface correction picture forming grid, the image capturing unit captures the second surface correction picture, and defines the first a gray scale image value field of a curved surface correction picture and a second curved surface correction picture is binarized into a binary image; (c) thinning the binary image into a thin line image, and is defined in the In the thinned image, if the first curved surface correction image and the second curved surface corrected image have intersection points to form the mesh, the projection surface is a curved surface; and the binarized image is divided into a background region and a projection. And (d) defining a range of parallelograms within the projection area as a transition region, and defining a correction parameter by the transition region to correspond to correcting an image of the projection of the projection unit.

According to the projection correction method described above, the steps further include: (e) defining a reference line within the thinned image to determine a plurality of coordinate points of the grid; and (f) defining an area formed by the coordinate points as the projected area.

The present invention further provides a projection control method comprising the projection correction method as described above.

According to the projection control method described above, the method further comprises: (g) capturing, by the image capturing unit, a hand image; (h) binarizing the hand image into a hand binarized image, and (i) establishing a training database that stores and defines a probability value for each color according to the skin color region and defines a skin color threshold value; (j) determining The color whose probability value is greater than the color threshold value is the skin color value, and forms a hand region according to the skin color value; and (k) after the hand region area is greater than a certain area value and continues to be fixed for a time, the control is opened and closed. The projection unit.

According to the projection control method described above, the method further comprises: (1) the image capturing unit extracts ambient luminosity; and (m) defines a luminosity threshold value, and starts when the ambient luminosity is less than the photometric threshold value; The projection unit.

The present invention further provides an augmented reality application system, comprising: a projection unit corresponding to projecting a first curved surface correction picture and a pair of second curved surface correction pictures that should be formed on the first curved surface correction picture forming mesh An image capturing unit corresponding to the first curved surface corrected image and the second curved surface corrected image; and a processing unit coupled to the projection unit and the image capturing unit The processing unit identifies the projection surface imaged by the correction picture, and divides Parsing the grayscale image value range of the first curved surface correction picture and the second curved surface correction picture, binarizing into a binary image, and thinning the binary image into a thinned image, and defining In the thinned image, if the first curved surface correction image and the second curved surface corrected image have intersection points to form the mesh, the projection surface is a curved surface, and the area formed by the mesh is defined as a projection area. And searching for the parallelogram range in the projection area as a conversion area, and correcting the projection unit corresponding to the image projected on the projection surface through the conversion area.

According to the augmented reality application system described above, a reflection unit is further included, which is correspondingly disposed on the projection unit, thereby adjusting the angle of projection of the projection unit.

According to the augmented reality application system described above, the projection unit is a projection image of the augmented reality.

According to the augmented reality application system described above, the image capturing unit further extracts a hand image, the processing unit is identified as a hand region according to a skin color value, and the processing unit is provided with an area setting value and a The time is set, and when the area of the hand area is greater than a certain value and continues to the time value, the projection unit is turned on and off.

According to the augmented reality application system described above, wherein the image capturing unit further extracts ambient luminosity, the processing unit is provided with a luminosity threshold value, and the projection is started when the ambient luminosity is less than the luminosity threshold value. unit.

With the above arrangement, the present invention can display digital information in the real world, make the information acquisition method closer to life, and solve the problem of using the augmented reality to project the projection information into the real world, which is costly and unfavorable for carrying and The invention is designed to allow the user to hang on the chest without burden, and the reflective unit can be manually operated to adjust the height of the projection unit and adjust the projection height or the projection surface of various shapes. The invention is characterized in that the image is distorted and deformed. Therefore, the present invention projects the correction image through the projection unit, and the image capturing unit captures the projection surface of the correction image on various shapes, such as a single plane, a double plane, and a curved surface, and corrects the projection unit. Projected image, obvious The invention allows the user to search for the projection surface anytime and anywhere, such as: wall corners, ceiling seams, ball arches, etc., or projection deformation due to different depths of the plane, such as: curtains, walls, columns, etc., the present invention The invention can overcome the problem and project the correct image; furthermore, the invention considers the convenience of use, and can define the area of the hand region after detecting the skin color, and when the value of the area is determined and the time is fixed, By closing the projection unit, the projection unit can be automatically turned on by the image capturing unit to reduce the power consumption, and the convenience and wide applicability of the present invention can be improved.

1‧‧‧projection unit

11‧‧‧Correction screen

2‧‧‧Projection surface

3‧‧‧Image capture unit

4‧‧‧ binarized image

41‧‧‧Background area

42‧‧‧Projection area

Edge of 421‧‧

422‧‧‧ Straight line

423‧‧‧ edge line

424‧‧‧ Faces

43‧‧‧Conversion area

44, 44a, 44b‧‧‧ regional corner points

45a, 45b‧‧‧ conversion point

46‧‧‧ Adjacent lines

5‧‧‧Hand image

51‧‧‧Hand binarized image

52‧‧‧ skin color area

53‧‧‧Non-skin area

6‧‧‧Reflective unit

7‧‧‧First surface correction screen

71‧‧‧Longitudinal line

8‧‧‧Second surface correction screen

81‧‧‧ horizontal line

9‧‧‧ Thin line image

91‧‧‧ baseline

92‧‧‧Punctuation

Figure 1 is a schematic view of the structure of the present invention.

Fig. 2 is a flow chart showing the first embodiment of the present invention.

Figure 3 is a schematic diagram of a binarized image of the first embodiment of the present invention.

Figure 4 is a schematic diagram of edge detection in the first embodiment of the present invention.

Fig. 5 is a schematic diagram showing the Hough conversion of the first embodiment of the present invention.

Fig. 6 is a schematic view showing the linear filtration of the first embodiment of the present invention.

Figure 7 is a schematic view showing the definition of the conversion area of the first embodiment of the present invention.

Figure 8 is a schematic view showing a projection image of a conversion correcting projection unit of the first embodiment of the present invention.

Figure 9 is a schematic diagram of the image capture unit of the first embodiment of the present invention.

Figure 10 is a schematic view showing a hand binarized image of the first embodiment of the present invention.

Figure 11 is a flow chart showing the second embodiment of the present invention.

Figure 12 is a schematic diagram of a binarized image of a second embodiment of the present invention.

Figure 13 is a schematic diagram of edge detection in accordance with a second embodiment of the present invention.

Fig. 14 is a schematic diagram showing the Hough conversion of the second embodiment of the present invention.

Fig. 15 is a schematic view showing the linear filtration of the second embodiment of the present invention.

Figure 16 is a schematic view showing the definition of the conversion area of the second embodiment of the present invention.

Figure 17 is a schematic view showing the adjacent line of the second embodiment of the present invention.

Figure 18 is a schematic view showing a second embodiment of the present invention adjacent to a line drawing sub-area.

Figure 19 is a schematic view showing a projection image of a projection conversion unit of the left side of the second embodiment of the present invention.

Figure 20 is a schematic diagram of a projection image of a field conversion correction projection unit on the right side of the second embodiment of the present invention.

Figure 21 is a flow chart showing the third embodiment of the present invention.

Figure 22 is a schematic diagram of a first curved surface correction screen of a third embodiment of the present invention.

Figure 23 is a schematic view showing a second curved surface correction screen of the third embodiment of the present invention.

Fig. 24 is a schematic view showing the thinning of the binarized image of the third embodiment of the present invention into a thinned image.

Figure 25 is a schematic view showing the definition of coordinate points of the reference line of the third embodiment of the present invention.

Figure 26 is a schematic view showing the formation of a projection area by the coordinate points of the third embodiment of the present invention.

Figure 27 is a schematic view showing a conversion area of a third embodiment of the present invention.

The invention will be described in detail below with reference to the drawings.

Please refer to FIG. 1 to FIG. 10 , which is a first embodiment of the present invention. The present invention is an augmented reality application system, comprising: a projection unit 1 corresponding to projecting a correction picture 11 a projection surface 2, the projection unit 1 is a projection image of the augmented reality, and the projection surface 2 is a single plane, a double plane or a curved surface; an image capturing unit 3 corresponding to the image captured by the projection surface 2 A processing unit is coupled to the projection unit 1 and the image capturing unit 3, the processing unit identifies the projection surface 2 imaged by the correction screen 11, and analyzes the grayscale image value of the correction screen 11. The domain is binarized into a binarized image 4, and the binarized image 4 is imaged into a background region 41 and a projection region 42, and the region of the parallelogram is searched for the transition region 43 in the projection region 42. Correcting the image of the projection unit 1 corresponding to the projection surface 2 through the conversion area 43 , the image capturing unit 3 captures the hand image 5 , the processing unit is identified as the hand area according to the skin color value, and the processing unit An area setting value and a time setting value are provided, and when the area of the hand area is greater than a certain value and continues to the time value, the projection unit 1 is controlled to be turned on and off; the image capturing unit 3 takes an ambient luminosity The processing unit is provided with a illuminance threshold value, and the projection unit 1 is activated when the ambient luminosity is less than the luminosity threshold value; and a reflection unit 6 correspondingly disposed on the projection unit 1 to adjust the projection unit 1 The angle of the projection.

Continuing to refer to FIG. 1 to FIG. 10, the present invention provides a projection correction method for a curved surface, the method comprising: a projection unit 1 projecting a correction image 11 on a projection surface 2; an image capture unit 3 captures The correction screen 11 assumes that the color space of the image capturing unit 3 is the three primary colors (RGB), and the corrected picture 11 captured by the image capturing unit 3 is defined as IC(x _c , y _c ). {R, G, B}, and define the grayscale image range of the correction screen 11 as [0, 255], and set the grayscale correction screen 11 to ICg (x _cg , y _cg ) to obtain ICg (x _cg , y _cg ) [0, 255], as shown in FIGS. 3 and 4, and then binarized into a binarized image 4, thereby dividing the binarized image 4 into a background region 41 and a projection region 42, and defining the Binarized image 4 is ICb (x _cb , y _cb ) {0,1}; detecting the edge 421 of the projection area 42 in the binarized image 4, and obtaining an edge image of the binarized image 4 as ICe(x _ce , y _ce ) {0, 1}, and as shown in Figures 5 and 6, the complex line 422 of the edge 421 is formed by the Hough transform definition, which will cause a misjudgment of the edge 421 detection, so the line 422 is Redundant linear filtering, defining an edge line 423 forming the edge 421 of the projection area 42, and defining the edge line 423 to form four regional corner points 44, the projection surface 2 is a single plane; defined by the edge The line corners 44 intersect at the intersection of the line 423. As shown in FIG. 4, the coordinates of the corner point 44 of the area can be obtained, and the area surrounded by the corner point 44 of the area is determined, which is the coordinate equation of the area of the projection area 42. The range of the parallelogram defining the coordinate equation of the projection area is the conversion area 43. For example, as shown in FIG. 7, the projection area 42 is a quadrilateral, and the corner point 44 of the area is the apex of the quadrilateral The edge line 423 is a quadrilateral edge, and a corner point 44a of the upper end is projected toward the y-axis direction to the opposite side of the edge line 423a to form a transition point 45a, and the corner point 44b of the other upper end is formed. Corresponding to the slope of the edge line 423a, corresponding to the transition point 45a The change point 45b is used to define the parallelogram formed by the area corner points 44a, 44b and the conversion points 45a, 45b as the conversion area 43; the second is shown in the upper left diagram of Fig. 8, which is the preview before the correction in the projection unit 1. The projected picture assumes that its coordinate points are I ₀ (0,0), I ₁ (iX-1,0), I ₂ (iX-1,iY-1), and I ₃ (0,iY-1), The result of the projection on the projection surface 2 is as shown in the upper right diagram of FIG. 8, which is the image captured by the image capturing unit 3, so that the coordinate points of the correction screen 11 are respectively P ₀ (x ₀ , y ₀ ), P ₁ (x ₁ , y ₁ ), P ₂ (x ₂ , y ₂ ), and P ₃ (x ₃ , y ₃ ), defining points I ₀ , I ₁ , I ₂ , I _{3 ,} and point P _{0, P 1, P 2,} P converter ₃ the relationship _{{P 0, P 1, P} 2, P 3} = T 1 {I 0, I 1, I 2, I 3}, so as to search for the projection unit 1 For the projection surface 2, the correction picture 11 is projected as a function of distortion, and as shown in the lower right diagram of FIG. 8, the coordinate points of the conversion area 43 are PI ₀ (xpi ₀ , ypi ₀ ), P ₁ ( Xpi ₁ , ypi ₁ ), P ₂ (xpi ₂ , ypi ₂ ) and P ₃ (xpi ₃ , ypi ₃ ), and the conversion relationship between the correction picture 11 and the conversion area 43 is {PI ₀ , PI ₁ , PI ₂ , PI ₃ }=T ₂ {P ₀ , P ₁ , P ₂ , P ₃ }, and by T ₁ and T ₂ inversely convert the projection unit 1 to correct the preview projection image, thus obtaining a correction parameter The coordinate points of the screen for previewing the projected projection unit 1 are PT ₀ (PTx ₀ , PTy ₀ ), P ₁ (PTx ₁ , PTy ₁ ), P ₂ (PTx ₂ , PTy ₂ ), and P ₃ ( PTx ₃ , PTy ₃ ), as shown in the lower left diagram of FIG. 8 , corresponds to the image of the projection of the projection unit 1 .

Continuing to refer to FIG. 9 and FIG. 10, the present embodiment provides a projection control method, the method comprising: the image capturing unit 3 captures a hand image 5; the present invention utilizes a YCbCr color space as a color. The representation is used to match the changes of the application scene and the ambient lighting, and the RGB color space has no separation of color and brightness. Therefore, the hand image captured by the image capturing unit 3 needs to first convert the RGB color space. For the YCbCr color space, the conversion is as follows:

In order to determine the correct skin color, it is necessary to establish a training database in the processing unit for learning the skin color determination. Therefore, the hand image 5 is first binarized into the hand binarized image 51, and is used. The user is customized to be divided into a skin color region 52 and a non-skin color region 53; stored in the training database and defining a probability value of each color according to the skin color region 52, and defining a skin color threshold value, determining that the probability value is greater than the skin color threshold value The color is a skin color value, and a hand region is formed according to the skin color value; the area of the hand region of the image capturing unit 3 is larger than a certain area value and continues to be fixed for a time, for example, the user puts the hand About two seconds before the image capturing unit 3 is placed, the processing unit controls the opening and closing of the projection unit 1; the image capturing unit 3 further extracts ambient luminosity; and the processing unit defines a luminosity threshold And the projection unit 1 is activated when the ambient luminosity is less than the luminosity threshold.

Referring again to FIGS. 11 to 20, which is a second embodiment of the present invention, which differs from the first embodiment in that six of said edge lines 423 are defined and said edge line 423 is defined. If three pairs of angle approximations are formed in the polar coordinates, the projection surface is a double plane, such as a corner of a wall or a seam of a ceiling, and the edge line 423 intersects and covers the projection area 42. The corner point 44 of the area; As shown in FIG. 16, a horizontal line L1 which is tangent to the corner 44 of the region having the largest inner angle is defined in the edge line 423, and extends perpendicularly to the edge line 423 from the point P0 of the horizontal line L1. Line L2, and the intersection of the horizontal line L1 and the vertical line L2, respectively, is defined by the distance from the edge line 423 to form three reference points P1, P2, P3, and a reference rectangle is determined as the reference points P1, P2 on the horizontal line L1. The distance between the horizontal line L1 and the vertical line L2, and the distance between the reference points P3 on the vertical line L2 is high; the reference rectangle is defined to extend the intersection point, and is proportionally enlarged to any of the reference rectangles When the vertex intersects the edge line 423, the transition region 43 is formed; as shown in FIGS. 17 to 20, a straight line defined by the corner point 44 of the region having the largest inner angle and the corner point 44 of the region opposite thereto is defined as Adjacent to the line 46, and defining the projection area 42 by the abutment line 46 to form a dihedral area 424, and the parallelograms in the area surrounded by the area corners 44 of the area 424 are The conversion area 43 is converted in the same manner as the first embodiment, so This will not be repeated here.

Continuing to refer to FIG. 21 to FIG. 27, which is a third embodiment of the present invention, which differs from the first embodiment in that the projection unit 1 projects a first curved surface correction picture 7 and is composed of the image. The first surface correction screen 7 is the longitudinal line 71 arranged in the interval, but is not limited thereto; as shown in FIG. 21 and FIG. 23 As shown in the figure, the projection unit 1 projects a pair of second curved surface correction pictures 8 which should be formed on the first curved surface correction picture 7. In the embodiment, the second curved surface correction picture 8 is a horizontal line 81 arranged at intervals. The image capturing unit 3 captures the second curved surface correction image 8 and thins the binary image 4 into a thinned image 9, as shown in FIG. 24, and is defined in the thinned image 9 The first curved surface correction screen 7 and the second curved surface corrected image 8 have intersections to form the mesh, and the projection surface 2 is a curved surface; see FIG. 25 and FIG. 26 for the second time. A reference line 91 is defined in the thinned image 9 to search for intersection coordinates of the longitudinal line 71 and the transverse line 81, respectively. To determine the grid coordinates of the complex Point 92; as shown in FIG. 27, the area formed by the coordinate points 92 is defined as the projection area 42. In this embodiment, the parallelogram range is determined in the projection area 42 as the conversion mode of the conversion area 43. It is the same as the first embodiment, so it will not be described here.

It is obvious from the above description and setting that the present invention has the following several advantages and effects, which are detailed as follows:

1. The present invention captures a correction picture through an image capturing unit, and determines that the edge line system forms four corner points, the projection surface is a single plane; defining six of the edge lines, and defining the Wherein the edge line forms three pairs of angle approximations in the polar coordinates, the projection surface is a double plane; defined in the thinned image, the first curved surface correction picture and the second curved surface corrected picture have intersections to form the net In the case of the grid, the projection surface is a curved surface; the invention is used to correct the projection of the projection unit, and the invention can be applied to various projection surfaces, such as wall corners, ceiling seams or spherical arches, without causing The image projected by the projection unit is distorted, so the invention has wide applicability.

2. The projection unit of the present invention can be a pico projector, and can project an augmented reality image, and the image capturing unit is a network camera, and the volume thereof is relatively small, so that the user can be hung without burden. The chest is convenient to carry and can be projected on various projection surfaces at any time, and can project the correct image without distorting the distortion, so the invention has the convenience and applicability, and if the projection is augmented reality Will make the augmented reality image more authentic.

3. The invention learns the skin color determination through the training database, and the user only needs to place the hand in the image capturing unit for a period of time, so that the area of the hand region on the image capturing unit screen is larger than an area. After the value is continued for a period of time, the projection unit can be opened and closed, and if the object is not a skin color, the projection unit cannot be opened and closed, and the image capturing unit is prevented from being misjudged to close the projection unit, thereby improving the present invention. The convenience of the switch.

4. The image capturing unit of the present invention captures ambient luminosity and has low ambient luminosity When the photometric threshold is used, the projection unit is automatically activated to perform projection, so as to achieve the convenience of use.

In summary, the technical means disclosed by the present invention can effectively solve the problems of the prior knowledge, achieve the intended purpose and efficacy, and are not found in the publication before publication, have not been publicly used, and have long-term progress, The invention referred to in the Patent Law is correct, and the application is filed according to law, and the company is invited to give a detailed examination and grant a patent for invention.

The above is only the preferred embodiments of the present invention, and the scope of the present invention is not limited thereto, that is, the equivalent changes and modifications made by the scope of the invention and the contents of the invention are all It should remain within the scope of this invention.

1‧‧‧projection unit

11‧‧‧Correction screen

2‧‧‧Projection surface

3‧‧‧Image capture unit

6‧‧‧Reflective unit

Claims (9)

A projection correction control method for a curved surface, the method comprising: (a) a projection unit projecting a first curved surface correction image on a projection surface, and capturing, by the image capturing unit, the first curved surface correction image; (b) the The projection unit projects a pair of second surface correction pictures that should be formed on the first surface correction picture forming grid, the image capturing unit captures the second surface correction picture, and defines the first surface correction picture and the second surface correction picture The grayscale image range is binarized into a binarized image; (c) the binarized image is thinned into a thinned image and defined in the thinned image, the first curved surface The correction picture and the second curved surface correction picture have intersection points to form the mesh, and the projection surface is a curved surface; and the binarized image is divided into a background area and a projection area; (d) defining the projection area The parallelogram in the range is a conversion area, and the correction parameter is defined by the conversion area to correspondingly correct the image projected by the projection unit; (e) the image capturing unit captures a hand image to form a hand area; After and (f) area is greater than a given value to the area of the hand and a duration time setting, the control of the opening and closing of the projection unit. The method for controlling a projection correction of a curved surface according to claim 1, wherein the step further comprises: (g) defining a reference line in the thinned image to determine a plurality of coordinate points of the grid; (h) defining the area formed by the coordinate points as the projected area. The projection correction control method for a curved surface according to claim 1, wherein the step further comprises: (i) binarizing the hand image into a hand binarized image, and dividing the user by a user. a skin color region and a non-skin color region; (j) establishing a training database that stores and defines a probability value for each color according to the skin color region, and defines a skin color threshold; (k) determining that the probability value is greater than the skin color threshold The color of the value is the skin color value, and a hand region is formed according to the skin color value. The projection correction control method for a curved surface according to claim 1, wherein the method further comprises: (1) the image capturing unit extracts ambient luminosity; and (m) defines a luminosity threshold value and is in an environment The projection unit is activated when the luminosity is less than the photometric threshold. An augmented reality application system, comprising: a projection unit corresponding to projecting a first curved surface correction picture and a pair of second curved surface correction pictures that should be formed on the first curved surface correction picture on a projection surface; an image a capture unit that captures the first curved surface corrected image and the second curved surface corrected image and captures a hand image; and a processing unit coupled to the projection unit and the image a capture unit that recognizes a projection surface imaged by the correction image, and analyzes a grayscale image value field of the first curved surface correction image and the second curved surface correction image, and binarizes the image into a binary image. And thinning the binarized image into a thinned image, and defining the thinned image, the first curved surface corrected image and If the second curved surface correction image has an intersection to form the mesh, the projection surface is a curved surface, and the area formed by the mesh is defined as a projection area, and the parallelogram range is searched for in the projection area. And correcting, by the conversion area, the image projected by the projection unit on the projection surface; the processing unit is configured to identify the hand image as a hand area, and the processing unit is provided with an area setting value and a time setting value. And when the area of the hand area is greater than a certain value and continues to the time value, the projection unit is controlled to be opened and closed. The augmented reality application system of claim 5, further comprising a reflection unit corresponding to the projection unit, thereby adjusting an angle of projection of the projection unit. The augmented reality application system of claim 5, wherein the projection unit projects an image of the augmented reality. The augmented reality application system of claim 5, wherein the processing unit is identified as the hand region according to a skin color value. The augmented reality application system of claim 5, wherein the image capturing unit further extracts ambient luminosity, the processing unit is provided with a photometric threshold value, and when the ambient luminosity is less than the photometric threshold value, The projection unit is activated.