WO2018107776A1

WO2018107776A1 - Video data rotation method and system

Info

Publication number: WO2018107776A1
Application number: PCT/CN2017/096479
Authority: WO
Inventors: 林文富
Original assignee: 威创集团股份有限公司
Priority date: 2016-12-13
Filing date: 2017-08-08
Publication date: 2018-06-21
Also published as: CN106791481A

Abstract

The present invention relates to a video data rotation method and system. The method for rotating video data comprises the following steps: obtaining an original display range corresponding to video data to be rotated which is on a display interface; calculating a target coordinate after a display coordinate corresponding to each pixel in the original display range is rotated according to a target rotation angle; determining a rotation correspondence relationship between a pixel and the target coordinate according to an original correspondence relationship between a pixel and a display coordinate in video data; filling each pixel in the video data to a target coordinate position corresponding to a pixel respectively according to the rotation correspondence relationship so as to rotate each pixel in said video data in the display interface; performance parameters of each pixel in the rotated video data remain unchanged, image quality corresponding thereto is well maintained, and a rotation process of each pixel in the video data is relatively simple to implement, while having high rotation efficiency.

Description

Video data rotation method and system

Technical field

The present invention relates to the field of video signal processing technologies, and in particular, to a video data rotation method and system.

Background technique

In recent years, the application frequency of display devices such as video splicing walls is getting higher and higher, and the application range is wider and wider. In the process of displaying the video data by the above display device, the data acquisition angle of the video source acquisition device usually changes (such as the scanning lens rotation of the scanner or the rotation of the camera lens, etc.), if the video data is to be During the acquisition process, the video data acquired under the acquisition of the angle change is sent to the display device for display, and the video data acquired in the above-mentioned acquisition angle change needs to be rotated during the display process of the display device to ensure the above video. The effect of the data on the display device.

In the traditional video data rotation scheme, there is a technique to realize the rotation of the corresponding image data by a two-step method, but in the first transformation, the high frequency part in the image is lost, and the image quality is degraded obviously, and then Paeth and Tanaka propose three on this basis. In the step implementation method, the image rotation processing can be completed by three times of translation, but the image quality is obviously different from the original image. There is also a method for realizing image rotation by using fast Fourier transform, which combines coordinate transformation and image interpolation into one, but the calculation amount is large and the implementation process is complicated. It can be seen that the traditional video data rotation scheme has a complicated implementation process, and the image quality in the video signal deteriorates after the rotation.

Summary of the invention

Based on this, it is necessary to provide a video data rotation method and system for the conventional video data rotation scheme, which has a complicated implementation process and a poor image quality in the rotated video signal.

A video data rotation method includes the following steps:

Obtaining an original display range corresponding to the video data to be rotated on the display interface;

Calculating target coordinates after the display coordinates corresponding to the respective pixels in the original display range are rotated according to the target rotation angle;

Determining a rotation correspondence relationship between the pixel and the target coordinate according to an original correspondence relationship between the pixel and the display coordinate in the video data;

Each pixel in the video data is respectively filled to a target coordinate position corresponding to the pixel according to the rotation correspondence.

A video data rotation system comprising:

The obtaining module is configured to obtain an original display range corresponding to the video data to be rotated on the display interface;

a calculation module, configured to calculate target coordinates after the display coordinates corresponding to the respective pixels in the original display range are rotated according to the target rotation angle;

a determining module, configured to determine a rotation correspondence relationship between the pixel and the target coordinate according to an original correspondence relationship between the pixel and the display coordinate in the video data;

And a filling module, configured to fill each pixel in the video data to a target coordinate position corresponding to the pixel according to the rotation correspondence.

The above video data rotation method and system can obtain the original display range corresponding to the video data to be rotated on the display interface and the corresponding display coordinates in the original display range in the video data, and calculate the target after the display coordinate is rotated according to the target rotation angle. Coordinates, determining a rotation correspondence relationship between the pixel and the target coordinate according to an original correspondence relationship between the pixel and the display coordinate in the video data, so that each pixel in the video data is respectively filled to a target coordinate position corresponding to the pixel according to the rotation correspondence relationship, In order to realize the rotation of each pixel in the video data on the display interface, the performance parameters of each pixel in the rotated video data remain unchanged, the corresponding image quality is better maintained, and each pixel in the above video data is rotated. The process is relatively simple and has a high rotational efficiency.

DRAWINGS

1 is a flow chart of a video data rotation method of an embodiment;

2 is a schematic diagram of an original display range and a target display range of an embodiment;

Figure 3 is a schematic diagram showing the original display range of an embodiment;

Figure 4 is a schematic diagram showing the original display range of an embodiment;

FIG. 5 is a schematic structural diagram of a video data rotation system of an embodiment.

detailed description

The specific embodiments of the video data rotation method and system of the present invention are described in detail below with reference to the accompanying drawings.

Referring to FIG. 1, FIG. 1 is a flowchart of a video data rotation method according to an embodiment, which includes the following steps:

S10. Acquire an original display range corresponding to the video data to be rotated on the display interface.

Normally, the data acquisition angle of the video source acquisition device changes (such as the scan lens rotation of the scanner, the projection angle of the projector, or the rotation of the camera lens, etc.), and the video data acquired in this state is required. After the rotation process is performed, the display is performed on the corresponding display interface to ensure the display effect of the above video data on the display interface. The video data includes a plurality of pixel points, and each pixel point has a corresponding display coordinate on the display interface, and each of the display coordinates constitutes an original display range corresponding to the video data on the display interface.

S20: Calculate target coordinates after the display coordinates corresponding to the respective pixels in the original display range are rotated according to the target rotation angle;

The target rotation angle may be determined according to the fluctuation range of the data acquisition angle of the video source acquiring device. Generally, the target rotation angle is a variation angle of the data acquisition angle of the video source acquiring device with respect to the horizontal direction or the vertical direction. Calculating a target coordinate of the display coordinate corresponding to each pixel in the original display range according to the target rotation angle, that is, converting each display coordinate in the original display range according to the corresponding target rotation angle to determine each display coordinate transformation Post coordinates (target coordinates).

S30. Determine a rotation correspondence relationship between the pixel and the target coordinate according to an original correspondence relationship between the pixel and the display coordinate in the video data.

The above steps may display the correspondence between the coordinates and the target coordinates by the correspondence between the pixels and the display coordinates, and determine the correspondence between the pixels and the target coordinates to obtain a rotation correspondence.

S40. Fill each pixel in the video data to a target coordinate position corresponding to the pixel according to the rotation correspondence.

The above steps may sample the pixels according to the arrangement features of the respective pixels in the video data to acquire respective pixels in the video data and perform corresponding filling.

The video data rotation method provided in this embodiment can obtain the original display range corresponding to the video data to be rotated on the display interface and the display coordinates corresponding to each display in the original display range in the video data, and calculate the display coordinate to rotate according to the target rotation angle. a target coordinate, determining a rotation correspondence relationship between the pixel and the target coordinate according to an original correspondence relationship between the pixel and the display coordinate in the video data, thereby filling each pixel in the video data to a target coordinate corresponding to the pixel according to the rotation correspondence relationship Position, in order to realize the rotation of each pixel in the above video data on the display interface, the performance parameter of each pixel in the rotated video data remains unchanged, the corresponding image quality is better maintained, and each pixel in the above video data is realized. The process of rotation is relatively simple and has a high rotational efficiency.

In one embodiment, the process of calculating the target coordinates of the display coordinates corresponding to each pixel in the original display range according to the target rotation angle may include:

Converting the original coordinate system corresponding to the display interface to the target coordinate system according to the target rotation angle;

Calculating a target display range corresponding to the original display range in the target coordinate system;

The target coordinates corresponding to the display coordinates of the respective pixels in the target coordinate system are respectively calculated.

In this embodiment, the conversion from the original coordinate system to the target coordinate system is performed according to the target rotation angle, thereby realizing the conversion from the display coordinate to the target coordinate, and has high conversion accuracy.

As an embodiment, the original display range ABCD shown in FIG. 2 is rotated to the target display range A'B'C'D', and the corresponding target rotation angle is Φ, and the original coordinate system in which the original display range ABCD is located may be converted first. For the target coordinate system with D as the coordinate origin (0, 0), the transformation from the original display range ABCD to the target display range A'B'C'D' is realized. The transformation process of the above-described original display range ABCD to the target display range A'B'C'D' may include:

If the original display range is a rectangular area, the width of the rectangular area is w, the height is l, and the origin of the coordinates is selected as (x ₀ , y ₀ ), where x ₀ =0, y ₀ =0, four vertices of the rectangular area The coordinates are A(x ₀ , y _k1 ), B(x _j1 , y _k ), C(x _j , y _k2 ), D(x _j2 , y ₀ ), respectively.

As shown in FIG. 3, the lengths of the respective horizontal segments include: the abscissa of the point D is j1, the abscissa of the point B is j2, and the abscissa of the point C is j, and the length of each segment is obtained by mathematical geometric calculation as follows:

J1=wcosΦ (1)

J2=lsinΦ (2)

j=wcosΦ+lsinΦ=j1+j2 (3)

As shown in FIG. 4, the lengths of the respective vertical segments include: the ordinate of the point A is k1, the ordinate of the point C is k2, and the ordinate of the point B is k, and the length of each segment is obtained by mathematical geometric calculation. :

K1=lcosΦ (4)

K2=wsinΦ (5)

k=lcosΦ+wsinΦ=k1+k2 (6)

Through Equations 1-6, the coordinates of the four points A, B, C, and D can be calculated, and the rectangular area circled by these four points is the area to be rotated (original display range). Assuming that the pixel coordinates of the rotation area are (x, y), when the image is taken, there is a limitation that 0 < x < x _j , 0 < y < y _k .

After the specific coordinates of the four points A, B, C, and D in the original coordinate system are defined by the above calculation formula 1-6, a new coordinate system (target coordinate system) is newly established. After the image is rotated, the new Cartesian coordinate system is established with the lower left point of the rectangle as the origin. As shown in Fig. 2, the D point is used as the origin (D' after rotation). After the new coordinates are defined, the original video pixels are filled into the new coordinate system, and the original pixel data is not changed. Here, the difference between the old rotation method and the complex calculation of the Fourier transform is omitted.

In an embodiment, before the process of calculating the target coordinates of the display coordinates of the respective pixels in the original display range according to the target rotation angle, the method may further include:

In the process of acquiring video data, detecting a rotation angle of the video capturing unit; wherein the video capturing unit is a camera of the imaging device or a scanning lens of the scanner;

A target rotation angle of the video data is determined according to the rotation angle.

In this embodiment, the target rotation angle of the video data is determined according to the rotation angle of the video collection unit, and the accuracy of the determined target rotation angle is ensured.

In one embodiment, the original correspondence between the pixels and the display coordinates in the video data is as described above. The process of determining the rotation correspondence between the pixel and the target coordinates may include:

Obtaining a correspondence between each pixel in the video data and each display coordinate in the original display range, to obtain an original correspondence relationship;

Obtaining a correspondence between each display coordinate in the original display range and each target coordinate in the target coordinate range, and obtaining a coordinate correspondence relationship;

Corresponding relationship between each pixel and each target coordinate is determined according to the original correspondence relationship and the coordinate correspondence relationship, and a rotation correspondence relationship is obtained.

In this embodiment, the rotation corresponding relationship is determined according to the original correspondence relationship and the coordinate correspondence relationship, and the accuracy is high, thereby ensuring the accuracy of the filling position of each pixel after the rotation, and improving the display effect of the corresponding video data after the rotation.

In an embodiment, the process of filling each pixel in the video data to the target coordinate position corresponding to the pixel according to the rotation correspondence may include:

Detecting an arrangement order of each pixel in the video data;

Each pixel is sequentially sampled from the video data according to the arrangement order, and the acquired pixels are respectively filled to a target coordinate position corresponding to the pixel.

In this embodiment, the sampling of the pixels is performed according to the order of arrangement of the pixels in the video data, so that the sampling process is more orderly.

A double rate synchronous dynamic random buffer is set, and the video data to be rotated is buffered to the double rate synchronous dynamic random buffer.

As an embodiment, the process of filling each pixel in the video data to the target coordinate position corresponding to the pixel according to the rotation correspondence may include:

And each pixel in the video data buffered by the double rate synchronous dynamic random buffer is respectively filled to a target coordinate position corresponding to the pixel according to the rotation correspondence.

The double rate synchronous dynamic random buffer can be established by using DDR (Double Rate Synchronous Dynamic Random Access Memory), and the double rate synchronous dynamic random buffer is a new video frame data storage. The area uses the double rate synchronous dynamic random buffer area to buffer the video data, and then performs corresponding sampling and filling, so that the subsequent sampling and filling process is more stable.

In one embodiment, the original display range is a triangular area; the vertices of the triangular area correspond to a first pixel in the video data;

The process of filling each pixel in the video data to the target coordinate position corresponding to the pixel according to the rotation correspondence relationship includes:

Starting from the first pixel in the video data, sampling down line by line; wherein, in the sampling process of the pixel, each line is downsampled, the ordinate of the sampling coordinate is decremented by one, and the abscissa of the sampling coordinate is decremented by one to the left. , add one to the right;

And sampling the pixel obtained according to the rotation correspondence to a target coordinate position corresponding to the pixel.

The original display range of this embodiment is a triangular area, and pixel sampling of the area can be started from the pixel corresponding to the vertex position, and sampled line by line. For each next line, the y coordinate (ordinate) is reduced by one until yk2 (original display) For the x-axis coordinate (abscissa), for each x-axis coordinate, the coordinates are decremented by one to the left and one to the right.

In one embodiment, the original display range is a parallel four deformation region; the upper left vertex of the parallel four deformation region corresponds to the first pixel in the video data;

Starting from the first pixel in the video data, sampling down row by row; wherein, in the sampling process of the pixel, each row is downsampled, the abscissa coordinate of the sampling coordinate is decremented to the left by one, and subtracted to the right by the sampling. The ordinate of the coordinate is decremented by one;

The original display range of this embodiment is a parallel four-deformation region. When collecting pixels corresponding to the upper left vertex of the parallel four-deformation region, another acquisition rule may be adopted. That is, the x coordinate (abscissa) is decremented by one to the left, and the right is also decremented by one, and the absolute value of the abscissa of the two edges is kept consistent until the adoption of A The point, that is, the x coordinate is 0; the y coordinate (ordinate) can be used for each next line, and the y coordinate is reduced by one until the y coordinate is the lowest point of the original display range.

In one embodiment, the original display range is an inverted triangle, and the sample can be sampled row by row from the upper left vertex point of the inverted triangle region, and the next y coordinate (ordinate) is reduced by one to the inverted triangle region. The lower vertex; the x-axis coordinate (abscissa), every next line of sampling, the coordinate is increased to the left, minus one to the right. When the target rotation angle is 45° or 90°, the sampling process is relatively simple.

Referring to FIG. 5, FIG. 5 is a schematic structural diagram of a video data rotation system according to an embodiment, including:

The obtaining module 10 is configured to obtain an original display range corresponding to the video data to be rotated on the display interface;

The calculation module 20 is configured to calculate target coordinates after the display coordinates corresponding to the respective pixels in the original display range are rotated according to the target rotation angle;

a determining module 30, configured to determine a rotation correspondence relationship between the pixel and the target coordinate according to an original correspondence relationship between the pixel and the display coordinate in the video data;

The filling module 40 is configured to fill each pixel in the video data to a target coordinate position corresponding to the pixel according to the rotation correspondence.

The video data rotation system provided by the present invention has a one-to-one correspondence with the video data rotation method provided by the present invention, and the technical features and advantages thereof in the embodiment of the video data rotation method are applicable to the embodiment of the video data rotation system. , hereby declare.

The technical features of the above-described embodiments may be arbitrarily combined. For the sake of brevity of description, all possible combinations of the technical features in the above embodiments are not described. However, as long as there is no contradiction between the combinations of these technical features, All should be considered as the scope of this manual.

The above-described embodiments are merely illustrative of several embodiments of the present invention, and the description thereof is more specific and detailed, but is not to be construed as limiting the scope of the invention. It should be noted that a number of variations and modifications may be made by those skilled in the art without departing from the spirit and scope of the invention. Therefore, the scope of the invention should be determined by the appended claims.

Claims

A video data rotation method, comprising the following steps:

Obtaining an original display range corresponding to the video data to be rotated on the display interface;

Calculating target coordinates after the display coordinates corresponding to the respective pixels in the original display range are rotated according to the target rotation angle;

Determining a rotation correspondence relationship between the pixel and the target coordinate according to an original correspondence relationship between the pixel and the display coordinate in the video data;

Each pixel in the video data is respectively filled to a target coordinate position corresponding to the pixel according to the rotation correspondence.
The video data rotation method according to claim 1, wherein the calculating the target coordinates of the display coordinates corresponding to each pixel in the original display range according to the target rotation angle comprises:

Converting the original coordinate system corresponding to the display interface to the target coordinate system according to the target rotation angle;

Calculating a target display range corresponding to the original display range in the target coordinate system;

The target coordinates corresponding to the display coordinates of the respective pixels in the target coordinate system are respectively calculated.
The video data rotation method according to claim 1, wherein before the process of calculating the target coordinates of the display coordinates corresponding to each pixel in the original display range according to the target rotation angle, the method further includes:

In the process of acquiring video data, detecting a rotation angle of the video capturing unit;

A target rotation angle of the video data is determined according to the rotation angle.
The video data rotation method according to claim 1, wherein the process of determining a rotation correspondence relationship between the pixel and the target coordinate according to an original correspondence relationship between a pixel and a display coordinate in the video data comprises:

Obtaining a correspondence between each pixel in the video data and each display coordinate in the original display range, to obtain an original correspondence relationship;

Obtaining a correspondence between each display coordinate in the original display range and each target coordinate in the target coordinate range, and obtaining a coordinate correspondence relationship;

Corresponding relationship between each pixel and each target coordinate is determined according to the original correspondence relationship and the coordinate correspondence relationship, and a rotation correspondence relationship is obtained.
The video data rotation method according to claim 1, wherein the process of filling each pixel in the video data to the target coordinate position corresponding to the pixel according to the rotation correspondence relationship comprises:

Detecting an arrangement order of each pixel in the video data;

Each pixel is sequentially sampled from the video data according to the arrangement order, and the acquired pixels are respectively filled to a target coordinate position corresponding to the pixel.
The video data rotation method according to claim 1, wherein before the process of calculating the target coordinates of the display coordinates corresponding to each pixel in the original display range according to the target rotation angle, the method further includes:

A double rate synchronous dynamic random buffer is set, and the video data to be rotated is buffered to the double rate synchronous dynamic random buffer.
The video data rotation method according to claim 6, wherein the process of filling each pixel in the video data to the target coordinate position corresponding to the pixel according to the rotation correspondence relationship comprises:

And each pixel in the video data buffered by the double rate synchronous dynamic random buffer is respectively filled to a target coordinate position corresponding to the pixel according to the rotation correspondence.
The video data rotation method according to any one of claims 1 to 7, wherein the original display range is a triangular area; the vertices of the triangular area correspond to a first pixel in the video data;

The process of filling each pixel in the video data to the target coordinate position corresponding to the pixel according to the rotation correspondence relationship includes:

Starting from the first pixel in the video data, sampling down line by line; wherein, in the sampling process of the pixel, each line is downsampled, the ordinate of the sampling coordinate is decremented by one, and the abscissa of the sampling coordinate is decremented by one to the left. , add one to the right;

Filling the sampled pixels to the target coordinates corresponding to the pixels according to the rotation correspondence position.
The video data rotation method according to any one of claims 1 to 7, wherein the original display range is a parallel four-deformation region; the upper left vertex of the parallel four-deformation region corresponds to the first pixel in the video data. ;

The process of filling each pixel in the video data to the target coordinate position corresponding to the pixel according to the rotation correspondence relationship includes:

Starting from the first pixel in the video data, sampling down row by row; wherein, in the sampling process of the pixel, each row is downsampled, the abscissa coordinate of the sampling coordinate is decremented to the left by one, and subtracted to the right by the sampling. The ordinate of the coordinate is decremented by one;

And sampling the pixel obtained according to the rotation correspondence to a target coordinate position corresponding to the pixel.
A video data rotation system, comprising:

The obtaining module is configured to obtain an original display range corresponding to the video data to be rotated on the display interface;

a calculation module, configured to calculate target coordinates after the display coordinates corresponding to the respective pixels in the original display range are rotated according to the target rotation angle;

a determining module, configured to determine a rotation correspondence relationship between the pixel and the target coordinate according to an original correspondence relationship between the pixel and the display coordinate in the video data;

And a filling module, configured to fill each pixel in the video data to a target coordinate position corresponding to the pixel according to the rotation correspondence.