TWI493502B - Processing method for image rotating and apparatus thereof - Google Patents

Description

Method and device for processing image rotation

The present invention relates to a method and apparatus for processing image rotation, and more particularly to a method and apparatus for processing image rotation during decompression.

In general, a digital camera or a digital camera often needs to rotate the image after taking a picture or recording. FIG. 1 is a flow chart of a method for generating a rotated image during decompression, please refer to FIG. 1. First, as described in step S110, when decompression is started, decompression is sequentially performed on a plurality of bands of each compressed image by using an image compression standard. Next, in step S120, each decompressed image is scaled to an appropriate size. Finally, as described in step S130, the scaled image is calculated by a central processing unit (CPU) to find the relative position of the rotated image in the memory, and then according to the relative position in the memory. The data is moved to complete the action of rotating the image.

In step S120 of the above flow, the image is usually reduced and then calculated by the central processing unit, because the reduced image can reduce the amount of calculation. However, if the user wants to view the full image size instead of the scaled size, the central processing unit rotates the full image size in the current image size (for example, 14 Mb), which requires a considerable amount of The amount of calculation is too time consuming.

In addition, if the decompressed image is a YCbCr 420 or YCbCr 422 format image, the Y component data and the Cb and Cr component data have a special data arrangement manner. Therefore, if the YCbCr format image is directly rotated, the computational complexity is quite high, and the rotation is relatively high. The relative position of the post-image in the memory is too discrete, which makes it impossible to accelerate the movement of data in the memory. Accordingly, the performance of the image rotation processing is reduced.

In view of the above, the present invention provides a method and apparatus for processing image rotation, by separating the image into luminance and chrominance data, and using a look-up table to perform rotation processing to improve image processing efficiency.

From a point of view, the present invention provides a method of processing image rotation suitable for receiving a plurality of compressed images, the method comprising the following steps. First, a plurality of blocks of each compressed image are sequentially decompressed by using an image compression standard to generate a plurality of first images corresponding to the blocks. And separating each of the first images into luminance data, first chrominance data, and second chrominance data. Then, according to the rotation angle correspondence table, the brightness table data, the first chromaticity data and the second chromaticity data are respectively rotated to generate the rotated brightness, the first chromaticity and the second chromaticity data. And combining the rotated brightness, the first chromaticity and the second chromaticity data into a second image, and sequentially arranging the second images to generate a rotated image.

In an embodiment of the invention, the rotation angle correspondence table records a correspondence between a rotation angle and a direct memory access (DMA) position.

In an embodiment of the invention, the method for processing image rotation further includes enlarging or reducing the rotated image to generate an adjusted image.

In an embodiment of the invention, the first image and the second image are YCbCr 420 format images or YCbCr 422 format images.

In an embodiment of the invention, the brightness data is Y component data, and the first chrominance data and the second chrominance data are Cb component data and Cr component data.

In an embodiment of the invention, the image compression standard uses a Joint Photographic Experts Group (JPEG) compression standard.

From another point of view, the present invention provides a device for processing image rotation, which is adapted to receive a plurality of compressed images. The device includes an image compression module, a separation module, a rotation module, and a merge module. The image compression module performs decompression on a plurality of blocks of each compressed image in sequence by using an image compression standard to generate a plurality of first images corresponding to the blocks. The separation module is coupled to the image compression module to separate each of the first images into luminance data, first chrominance data, and second chrominance data. The rotating module is coupled to the separating module, and is inspected according to the rotation angle correspondence table, respectively rotating the brightness data, the first chromaticity data and the second chromaticity data, thereby generating the rotated brightness, the first chromaticity and the second color Information. The merging module is coupled to the rotating module, and combines the rotated brightness, the first chromaticity and the second chromaticity data into a second image, merges the modules, and sequentially arranges the second images to generate a rotated image.

In an embodiment of the invention, the rotation module establishes a rotation angle correspondence table, wherein the rotation angle corresponds to a correspondence between the record rotation angle and the direct memory access position.

In an embodiment of the present invention, the device for processing image rotation further includes a size adjustment module coupled to the module, and the size adjustment module enlarges or reduces the rotated image generated by the merged module to generate Adjusted image.

In an embodiment of the invention, the first image and the second image are YCbCr 420 format image or YCbCr 422 format image.

In an embodiment of the invention, the brightness data is a Y component data, and the first chrominance data and the second chrominance data are Cb component data and Cr component data.

In an embodiment of the present invention, the image compression standard adopted by the image compression module is a joint image expert group compression standard.

Based on the above, the method and apparatus for processing image rotation provided by the present invention divides the image into a plurality of blocks, and sequentially separates the image of each block into luminance and chrominance data, and uses the look-up table to measure the brightness. Rotate separately from the chromaticity data, and finally combine the rotated luminance and chrominance data to generate a rotated image. According to this, the performance of the image rotation processing can be increased.

The above described features and advantages of the present invention will be more apparent from the following description.

2 is a block diagram of an apparatus for processing image rotation in accordance with an embodiment of the present invention. Referring to FIG. 2, the device 200 for processing image rotation in this embodiment is, for example, a digital camera, a monocular camera, a video recorder, or other smart phone, tablet computer or the like having image processing functions, and is not limited to the above. The apparatus 200 for processing image rotation is adapted to receive a plurality of compressed images, including an image compression module 210, a separation module 220, a rotation module 230, and a merge module 240. The function is described as follows: The image compression module 210 can be implemented by software, hardware, or a combination thereof, and is not limited herein. The image compression module 210 compresses and decompresses the image by using an image compression standard, such as the Joint Photographic Experts Group (JPEG) compression standard.

The separation module 220, the rotation module 230, and the merge module 240 are all functional modules implemented by hardware. The hardware may include a central processing unit, a programmable general purpose or special purpose microprocessor (Microprocessor), a digital signal processor (DSP), and the like, or a hardware device having a computing function or the foregoing hardware device. The combination. The separation module 220 is coupled to the image compression module 210 for separating the received images into different brightness and chrominance data. The rotation module 230 is coupled to the separation module 220, and different brightness and chromaticity data are respectively inspected and rotated according to the rotation angle correspondence table. Finally, the merge module 240 coupled to the rotation module 230 combines the rotated luminance and chrominance data to generate a rotated image.

FIG. 3 is a flow chart of a method for processing image rotation according to an embodiment of the invention. Referring to FIG. 3, the method of the present embodiment is applied to the device 200 for processing image rotation of FIG. 2, and the detailed steps of the method for processing image rotation in this embodiment are described below with reference to each module in FIG. 2: first, as shown in step S310. The image compression module 210 sequentially decompresses a plurality of blocks of each compressed image by using an image compression standard to generate a plurality of first images corresponding to the blocks. In this embodiment, the image compression standard is, for example, a joint imaging expert group compression standard. When the image compression module 210 decompresses the compressed image, the partitioning mode is adopted. For example, every 256 columns in the compressed image can be divided into one block. FIG. 4 is a schematic diagram of image partition execution rotation according to an embodiment of the invention. Please refer to FIG. 2, FIG. 3 and FIG. 4 together. A compressed image can be divided into N blocks, and the image compression module 210 decompresses the N blocks to generate N first images 410_1~ 410_N, the N first images 410_1~410_N may be YCbCr 420 or YCbCr 422 format images. The YCbCr 420 format image means that the Cb chrominance component and the Cr chrominance component are only half of the sampling rate in the horizontal and vertical directions; the YCbCr 422 format image means that the Cb chrominance component and the Cr chrominance component have the same vertical sampling rate, but the level The sampling rate is only half.

Next, as described in step S320, the separation module 220 sequentially separates the N first images 410_1~410_N into luminance data, first chrominance data, and second chrominance data. As shown in FIG. 4, the first image 410_1 is a YCbCr 420 or YCbCr 422 format image. Therefore, in this embodiment, the first image 410_1 can be divided into a Y component data 411, a Cb component data 412, and a Cr component data 413.

Next, in step S330, the rotation module 230 establishes a rotation angle correspondence table, wherein the rotation angle correspondence table records the correspondence relationship between the rotation angle and the direct memory access (DMA) position. In this embodiment, according to the rotation angle correspondence table, the rotation module 230 can respectively rotate the Y component data 411, the Cb component data 412, and the Cr component data 413 to generate the rotated Y component data 411' and after the rotation. The Cb component data 412' and the rotated Cr component data 413'.

The merge module 240 merges the rotated Y component data 411', the rotated Cb component data 412' and the rotated Cr component data 413' into a second image 420_1, and the remaining second images 420_2~420_N are The order is arranged to produce a complete rotated image. Since the merged second images 420_1~420_N are ordered partitions, after the second image 420_1 is aligned to the corresponding memory data position, the remaining second images 420_2~420_N are sequentially arranged in order. Yes, no more operations are required, so that the problem that the relative position of the image after rotation in the memory is too discrete as described in the prior art can be avoided.

As described above, in the image rotation, the YCbCr format image is first separated into individual luminance and chrominance component data, and is not directly calculated and rotated in the YCbCr format image as described in the prior art. Therefore, the present invention utilizes individual luminance and chrominance component data for rotation, which greatly reduces computational complexity. In addition, the present invention utilizes a rotating module to pre-establish a rotation angle correspondence table, and according to the manner of looking up the table, finds the relative position of the image in the memory after being rotated, and calculates each pixel in the image compared to completely using the software. The method and apparatus for processing image rotation provided by the present invention are faster in calculating the relative position of the dots in the memory.

The following additional embodiments are provided as examples in which the present invention can be implemented. FIG. 5 is a block diagram of an apparatus for processing image rotation according to another embodiment of the present invention. In this embodiment, the apparatus 500 for processing image rotation includes a size adjustment module 550 in addition to the image compression module 210, the separation module 220, the rotation module 230, and the merge module 240. The size adjustment module 550 is coupled to the merge module 240. The size adjustment module 550 can also be a functional module implemented by a hardware to enlarge or reduce an image. Since the device 500 for processing the image rotation has the same or similar functions as the modules in the device 200 for processing the image rotation shown in FIG. 2, the remaining modules are not described herein.

FIG. 6 is a flow chart of a method for processing image rotation according to another embodiment of the invention. The mode of operation of the apparatus 500 for processing image rotation will be described below with reference to FIG. Please refer to FIG. 5 and FIG. 6 at the same time.

The image compression module 210 sequentially decompresses a plurality of blocks of each compressed image by using an image compression standard to generate a plurality of first images corresponding to the blocks (step S610). Next, the separation module 220 sequentially separates the first images into luminance data, first chrominance data, and second chrominance data (step S620). The rotation module 230 performs a table lookup according to the rotation angle correspondence table, and respectively rotates the brightness data, the first chromaticity data and the second chromaticity data to generate the rotated brightness, the first chromaticity and the second chromaticity data (step S630). ). The merging module 240 combines the rotated brightness, the first chromaticity and the second chromaticity data into a second image, and sequentially arranges the second images to generate a rotated image (step S640). Finally, the resizing module 550 can also zoom in or out on the rotated image generated by the merge module 240 to generate an adjusted image. This adjusted image can be directly output to the screen for playback.

It should be noted that, as described in the prior art, since the amount of calculation for calculating the image rotation by the software is too large, it is necessary to perform the operation of reducing the image before the image is rotated to reduce the amount of calculation. However, since the present invention uses a rotating module to establish a table lookup method, the amount of calculation can be saved, and the image does not need to be scaled before the image is rotated. The size adjustment module 550 is an additional function for adjusting the image size to a suitable image size for playback and viewing before the image is output.

The experimental performance test is performed by the embodiment shown in FIG. 5 and FIG. 6. When the test image size is 4288×3216, the test result is as follows: the time taken by the separation module 220 to perform its work is about 148 milliseconds (ms). The time it takes for the rotation module 230 to perform its work is about 472 ms; the time it takes for the merge module 240 to perform its work is about 191 ms; the time it takes for the size adjustment module 550 to perform its work is about It is 251 ms; the time taken to complete the entire image rotation process is approximately 1067 ms.

In contrast, the experimental performance test is performed by using the method flow as shown in FIG. 1. When the test image size is also 4288×3216, the time taken to complete the entire image rotation process is about 4034 ms. Accordingly, the method and apparatus for processing image rotation provided by the present invention can indeed improve the performance of image processing.

In summary, the method and apparatus for processing image rotation provided by the present invention divide the image into a plurality of blocks, and sequentially separate the image of each block into brightness and chrominance data, and use the hardware to look up the table. The method of rotating the luminance and chrominance data separately, and finally combining the rotated luminance and chrominance data to generate a rotated image. The invention adopts a combination of software and hardware to increase the performance of image rotation processing, in addition to avoiding the problem that the calculation amount is too large and the processing time is too long, and the use of hardware operations can be avoided. The problem of excessive cost.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the invention, and any one of ordinary skill in the art can make some modifications and refinements without departing from the spirit and scope of the invention. The scope of the invention is defined by the scope of the appended claims.

200, 500. . . Device for processing image rotation

210. . . Image compression module

220. . . Separation module

230. . . Rotary module

240. . . Merge module

410_1~410_N. . . First image

411, 411’. . . Y component data

412, 412’. . . Cb component data

413, 413’. . . Cr component data

420_1~420_N. . . Second image

S110~S130. . . Steps of the method of generating a rotated image

S310~S340. . . Each step of the method for processing image rotation according to an embodiment of the present invention

S610~S650. . . Various steps of a method for processing image rotation according to another embodiment of the present invention

1 is a flow chart of a method of generating a rotated image during the execution of a decompression process.

2 is a block diagram of an apparatus for processing image rotation in accordance with an embodiment of the present invention.

FIG. 3 is a flow chart of a method for processing image rotation according to an embodiment of the invention.

FIG. 4 is a schematic diagram of image partition execution rotation according to an embodiment of the invention.

FIG. 5 is a block diagram of an apparatus for processing image rotation according to another embodiment of the present invention.

FIG. 6 is a flow chart of a method for processing image rotation according to another embodiment of the invention.

S310~S340. . . Steps in the method of processing image rotation

Claims (12)

A method for processing image rotation, which is adapted to receive a plurality of compressed images, comprising: performing decompression on a plurality of blocks of each compressed image in sequence by using an image compression standard to generate a majority corresponding to the blocks; Separating each of the first images into a brightness data, a first chromaticity data, and a second chromaticity data; performing a table lookup according to a rotation angle correspondence table, respectively rotating the brightness data, the first color Degree data and the second chromaticity data to generate the rotated brightness, the first chromaticity and the second chromaticity data; and the rotated brightness, the first chromaticity and the second chromaticity The data is merged into a second image, and each of the second images is sequentially arranged to generate a rotated image that is rotated at a spatial position. The method for processing image rotation as described in claim 1, wherein the rotation angle corresponds to a correspondence between a record rotation angle and a direct memory access (DMA) position. The method for processing image rotation as described in claim 1, further comprising: zooming in or out the rotated image to generate an adjusted image. The method for processing image rotation according to the first aspect of the invention, wherein the first image and the second image are YCbCr 420 format image or YCbCr 422 format image. The method for processing image rotation as described in claim 4, wherein the brightness data is Y component data, and the first chromaticity data and the second chromaticity data are Cb component data and Cr component data. The method for processing image rotation as described in claim 1, wherein the image compression standard adopts a Joint Photographic Experts Group (JPEG) compression standard. A device for processing image rotation is adapted to receive a plurality of compressed images, comprising: an image compression module, wherein an image compression standard sequentially performs decompression on a plurality of blocks of each of the compressed images to generate a pair a plurality of first images of the plurality of blocks; a separate module coupled to the image compression module to separate each of the first images into a luminance data, a first chrominance data, and a second chrominance data; The rotation module is coupled to the separation module, and is inspected according to a rotation angle correspondence table, and respectively rotates the brightness data, the first chromaticity data and the second chromaticity data to rotate, to generate the brightness after rotation The first chromaticity and the second chromaticity data; and a merging module coupled to the rotating module, combining the rotated brightness, the first chromaticity and the second chromaticity data into a second And merging the modules and sequentially arranging each of the second images to generate a rotated image that is rotated in a spatial position. The apparatus for processing image rotation according to claim 7, wherein: the rotation module establishes the rotation angle correspondence table, wherein the rotation angle is The correspondence table records the correspondence between the rotation angle and the position of the direct memory access (DMA). The apparatus for processing image rotation as described in claim 7 further includes: a size adjustment module coupled to the merge module, wherein the size adjustment module performs the rotated image generated by the merged module Zoom in or out to produce an adjusted image. The device for processing image rotation according to claim 7, wherein the first image and the second image are YCbCr 420 format image or YCbCr 422 format image. The apparatus for processing image rotation according to claim 10, wherein: the brightness data is Y component data, and the first chromaticity data and the second chromaticity data are Cb component data and Cr component data. The apparatus for processing image rotation according to claim 7, wherein the image compression standard used by the image compression module is a joint image expert group compression standard.