WO2021004126A1

WO2021004126A1 - Image coding method and apparatus

Info

Publication number: WO2021004126A1
Application number: PCT/CN2020/086334
Authority: WO
Inventors: 高鹏; 范志刚; 周毅
Original assignee: 西安万像电子科技有限公司
Priority date: 2019-07-09
Filing date: 2020-04-23
Publication date: 2021-01-14
Also published as: CN110505483A

Abstract

The present disclosure relates to the technical field of electronic information, and provided are an image coding method and apparatus, which may resolve the problem in a frame image processing process in which the coding effect is poor due to a macro block that comprises image information and text information being identified as a text macro block. A specific technical solution is as follows: the image coding method provided in embodiments of the present disclosure: obtaining a first macro block in a target frame image, and then according to color information of the first macro block, dividing the first macro block to then generate a text macro block and an image macro block, wherein an image in the image macro block is an image that comprises text information; once hierarchical processing is performed on the image macro block, generating a text layer and an image layer, using a text coding means for the text layer and the text macro block, and using an image coding means for the image layer. The present disclosure is used for image coding.

Description

Image coding method and device

Technical field

The present disclosure relates to the field of electronic information technology, and in particular to image coding methods and devices.

Background technique

With the rapid development of image processing algorithms, the technology of image encoding and decoding processing algorithms also develops rapidly. In the prior art, when encoding and decoding an image, the image is first recognized, text macroblocks or image macroblocks are determined according to the recognition results, and then corresponding encoding and decoding algorithms are determined according to different macroblocks. Among them, in view of the simple color of text macroblocks, the frame image is divided into fixed-size macroblocks, and the macroblocks are divided into two types: picture and text according to the proportion of basic colors in the macroblock and the number of high-gradient pixels. The basic color refers to the basic color, base color, the basic color refers to the gray value that appears most frequently in the image as the basic color, that is, the main color in an image. The disadvantage of this algorithm is that when processing images based on these two features, macroblocks that are not real text will be recognized as text macroblocks. For example, in the process of processing desktop icon images, because the macroblocks corresponding to the desktop icon images contain many types of colors, if the blocks that are not real text are recognized as text macroblocks, the text encoding algorithm is used to process the non-text Macro blocks, the coding effect cannot meet the preset requirements, and obvious color layering between adjacent macro blocks will occur.

Summary of the invention

The embodiments of the present disclosure provide an image encoding method and device, which can solve the problem of poor encoding effect due to the recognition of macroblocks containing images as text macroblocks during frame image processing. The technical scheme is as follows:

According to a first aspect of the embodiments of the present disclosure, there is provided an image encoding method, the method including:

Acquiring a first macro block and color information corresponding to the first macro block, where the first macro block includes: text information and image information;

Determining the text macro block and the image macro block in the first macro block according to the color information corresponding to the first macro block, and the image in the image macro block is an image containing text information;

After dividing and processing the image macroblock according to a preset algorithm, an image layer and a text layer are generated;

The text macroblock and the text layer are coded according to the first coding algorithm, and the image layer is coded according to the second coding algorithm to generate the coded macroblock.

In an embodiment, before acquiring the first macroblock, the method further includes:

Acquiring a frame image and ratio information of at least one macroblock in the frame image, the ratio information including: ratio information of high gradient pixels and ratio information of basic color pixels;

When the ratio information of at least one macro block in the frame image meets a preset requirement, it is determined that at least one macro block in the frame image is the first macro block.

In an embodiment, the method determining the text macroblock and the image macroblock in the first macroblock includes:

Obtaining the number of color types in at least one sub-macroblock in the first macroblock according to the color information;

When the number of color types in the sub-macroblock is greater than the preset number, determining that the sub-macroblock is the image macroblock;

When the number of color types in the sub-macroblock is less than or equal to the preset number, determining that the sub-macroblock is the text macroblock;

Determine the text macro block and the image macro block in the first macro block according to the image macro block or the text macro block corresponding to the sub macro block.

In an embodiment, generating an image layer and a text layer in the method includes:

Analyze the image macroblock, and obtain high-gradient pixels in the image macroblock according to the analysis result;

Divide the high-gradient pixels into a text layer, and divide the pixels in the image macroblock except the high-gradient pixels into an image layer;

After processing all the pixels in the image macroblock, fill the text layer or the blank pixels in the image layer according to the preset pixel value, and then generate the image layer and the text layer.

In one embodiment, the method generates a text layer based on high gradient pixels, including

Acquiring target position information and an average pixel value corresponding to the target position information, where the target position information refers to position information corresponding to pixels other than high gradient pixels in the text layer;

Generating and determining preset pixels according to the target position information and the average pixel value corresponding to the target position information;

The text layer is generated according to the high gradient pixels and the preset pixels.

The image coding method provided by the embodiment of the present disclosure obtains the first macro block in the target frame image, and the first macro block, that is, the macro block containing text information and image information in the frame image is determined as a text macro in the prior art. After dividing the first macro block according to the color information of the first macro block, the text macro block and the image macro block are generated. The image in the image macro block is an image containing text information; After the layering process, the text layer and the image layer are generated, the text encoding method is adopted for the text macroblock and the text layer map, and the image encoding method is adopted for the image layer layer. By improving the accuracy of recognizing text macroblocks in frame images, it is avoided that macroblocks containing image information and text information are recognized as text macroblocks, and matching coding algorithms are selected according to the different characteristics of text macroblocks and image macroblocks, thereby The coding effect of the frame image is improved, and then the visual effect and compression ratio effect of the frame image after the coding is improved.

According to a second aspect of the embodiments of the present disclosure, there is provided an image encoding device, which includes: a first acquisition module, a first determination module, a layer generation module, and an image generation module;

The first acquisition module is configured to acquire the first macro block and the color information corresponding to the first macro block, and the first macro block includes: text information and image information;

The first determining module is configured to determine the text macroblock and the image macroblock in the first macroblock according to the color information corresponding to the first macroblock, and the image in the image macroblock is an image containing text information;

The layer generation module is used to generate image layers and text layers after dividing image macroblocks according to preset algorithms;

The image generation module is used to encode the text macroblock and the text layer according to the first encoding algorithm, and to generate the encoded macroblock after encoding the image layer according to the second encoding algorithm.

In an embodiment, the device further includes: an image acquisition module and a second determination module,

The image acquisition module is configured to acquire a frame image and the proportion information of at least one macroblock in the frame image, the proportion information including: proportion information of high gradient pixels and proportion information of basic color pixels;

The second determining module is configured to determine that at least one macroblock in the frame image is the first macroblock when the ratio information of at least one macroblock in the frame image meets a preset requirement.

In an embodiment, the first determining module of the device includes: an acquiring unit, a first determining unit, a second determining unit, and a third determining unit,

The acquiring unit is configured to acquire the number of color types in at least one sub-macroblock in the first macroblock according to the color information;

The first determining unit is configured to determine that the sub-macroblock is an image macroblock when the number of color types in the sub-macroblock is greater than the preset number;

The second determining unit is configured to determine that the sub-macroblock is a text macroblock when the number of color types in the sub-macroblock is less than or equal to the preset number;

The third determining unit is configured to determine the text macro block and the image macro block in the first macro block according to the image macro block or the text macro block corresponding to the sub macro block.

In one embodiment, the layer generating module of the device includes a first acquiring unit and a generating unit

The first obtaining unit is configured to analyze the image macroblock, and obtain high gradient pixels in the image macroblock according to the analysis result;

The generating unit is configured to generate a text layer based on the high gradient pixels, and generate an image layer based on pixels other than the high gradient pixels in the image.

In one embodiment, the layer generating module of the device includes a determining unit, a second acquiring unit, a first generating unit, and a second generating unit,

The second acquiring unit is used to acquire the target position information and the average pixel value corresponding to the target position information. The target position information refers to the position information corresponding to pixels other than high gradient pixels in the text layer;

The first generating unit is configured to generate and determine a preset pixel according to the target position information and the average pixel value corresponding to the target position information;

The second generating unit is used to generate a text layer according to the high gradient pixels and the preset pixels.

It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure.

Description of the drawings

The drawings herein are incorporated into the specification and constitute a part of the specification, show embodiments in accordance with the disclosure, and together with the specification are used to explain the principle of the disclosure.

Fig. 1 is a flowchart of an image coding method provided by an embodiment of the present disclosure;

Figure 2 is a structural diagram of an image encoding device provided by an embodiment of the present disclosure;

Fig. 3 is a structural diagram 1 of an image encoding device provided by an embodiment of the present disclosure;

Fig. 4 is a structural diagram 2 of an image encoding device provided by an embodiment of the present disclosure;

Fig. 5 is a structural diagram 3 of an image encoding device provided by an embodiment of the present disclosure;

Fig. 6 is a structural diagram 4 of an image encoding device provided by an embodiment of the present disclosure.

Detailed ways

Here, exemplary embodiments will be described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present disclosure. Rather, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

Example one

The embodiment of the present disclosure provides an image encoding method. As shown in FIG. 1, the image encoding method includes the following steps:

101. Acquire a first macro block and color information corresponding to the first macro block.

Among them, the first macro block includes: text information and image information.

The color information is used to indicate the color information corresponding to the pixels in the first macroblock, and may include information such as the number of color types included in the first macroblock and the color value corresponding to the pixel.

Before acquiring the first macro block, the method of the present disclosure includes determining the first macro block in the frame image, that is, determining the image macro block and the text macro block in the target frame image, including:

Acquiring a frame image and ratio information of at least one macroblock in the frame image, where the ratio information includes: ratio information of high gradient pixels and ratio information of basic color pixels, and the basic color is determined according to the gray value in the frame image;

When the ratio information of at least one macroblock in the frame image meets the preset requirement, the first macroblock of the at least one macroblock in the frame image is determined.

Wherein, obtaining the ratio information of at least one macroblock in the frame image includes:

First determine the basic color and high gradient pixels corresponding to at least one macro block in the frame image, and then according to the ratio of the number of basic color pixels of at least one macro block in the frame image to all pixels in the frame image, that is, the basic color pixel Proportion; according to the ratio of the number of high gradient pixels in at least one macroblock in the frame image to all pixels in at least one macroblock in the frame image, that is, the ratio of high-pixel image pixels.

Further, the basic color of the image frame, that is, the main color in the image frame, may be determined according to the frequency histogram of the frame image, or may be determined according to the number of occurrences of each color value in the frame image.

Further, the number of medium and high gradient pixels obtained in the image frame includes:

Determine whether the difference between the pixel value of each pixel in the frame image and the pixel value of any adjacent pixel is greater than or equal to the third threshold;

If the difference between the pixel value of the first pixel in the frame image and the pixel value of any adjacent pixel is greater than or equal to the third threshold, the first pixel is determined as a high gradient pixel;

Count the number of first pixels. That is, by comparing the difference between adjacent pixels, the statistics of high gradient pixels are realized.

The above-mentioned first macroblock is the text macroblock determined by the prior art. The prior art determines the text macroblock based on the characteristics of simple color of the text macroblock; according to the basic color pixel ratio and high gradient in the frame image The number of pixels, etc., divide the frame image into text macroblocks or picture macroblocks.

However, the strategy of identifying image macroblocks and text macroblocks in an image based on the basic color ratio and high gradient pixels based on the prior art: For example, a certain macroblock of an image contains a large number of high gradient pixels, and the macroblock contains The proportion of basic color pixels is high, indicating that the macroblock contains fewer colors, the difference is large, and the color is relatively single, so it is divided into text macroblocks; but this strategy is to determine whether a pixel is a basic color pixel. Calculated according to the pixel values in a range. If the color in the target image is a gradual color, several pixels with similar pixel values will be divided into basic color pixels, which increases the proportion of basic color pixels in the macro block. As a result, the macro block containing the image is determined to be a text macro block.

Therefore, the present disclosure takes the text macroblock after the target frame image is divided as the first macroblock in the prior art, and further processes the first macroblock to improve the data accuracy and the encoding effect.

In view of the above, the method of the present disclosure aims to solve the problem that when dividing frame images into text macroblocks and image macroblocks, the type of macroblock cannot be accurately determined due to the algorithm based on the prior art, and the macroblock containing image information and text information is divided into It is a text macro block, where the frame image can be an image formed by superimposing multiple layers of images, or an image with rich colors.

102. Determine the text macroblock and the image macroblock in the first macroblock according to the color information.

Wherein, the image in the image macroblock is an image containing text information. For example, the image may be an icon with embedded text or a picture with embedded text.

Based on the above analysis, in view of the defects of dividing text macroblocks in the prior art, frame images with gradient colors or complex colors cannot be accurately divided into text macroblocks or image macroblocks. Therefore, according to the color information of the first macroblock, the first macroblock The blocks are further divided into text macro blocks and image macro blocks, which specifically include:

Here are specific examples to illustrate: Obtain the number of color types in an image macroblock. For example, for an image macroblock of 16*16 size, determine whether the number of color types in the image macroblock exceeds a threshold. If not, It is determined that the image macro block is a text type macro block, and if it exceeds, it is determined that the text macro block is a picture or icon type embedded with text.

103. After performing division processing on the image macroblock according to a preset algorithm, an image layer and a text layer are generated.

In an optional embodiment, the method of the present disclosure, when generating an image layer and a text layer according to the high gradient feature data in the image macroblock, includes:

Divide the high-gradient pixels in the image macroblock into the text layer, and divide the pixels except the high-gradient pixels in the image macroblock into the image layer;

After all the pixels in the image macroblock are processed, the blank pixels in the text layer or the image layer are filled according to the preset pixel values to generate the image layer and the text layer.

For the above method, before obtaining the high gradient pixels in the image macroblock according to the analysis result, it includes determining the high gradient pixels in the image macroblock:

Traverse each pixel contained in the image macroblock, and obtain the gradient value corresponding to each pixel. According to the pixel value corresponding to each pixel, the absolute value of each pixel is calculated by the difference between the pixel values of the four pixels of up, down, left, and right, and it is determined whether the sum of the absolute values exceeds the preset threshold; when the target gradient value exceeds the preset Threshold, the pixel is determined to be a high gradient pixel.

Aiming at the above method, filling the blank pixels in the text layer or the image layer according to the preset pixel value includes

Determine the target position information of the blank pixels in the text layer or the image layer according to the corresponding position information in the at least one pixel divided into the text layer or the image layer;

Obtaining an average pixel value of pixels within a preset range of the target position information;

According to the average pixel value, the blank pixels corresponding to the target position information are filled.

104. Encode the text macroblock and the text layer according to the first encoding algorithm, and after encoding the image layer according to the second encoding algorithm, generate an encoded macroblock.

In an alternative embodiment, the first encoding algorithm may be an encoding algorithm for text types, and the second encoding algorithm may be an encoding algorithm for image types, and the second encoding algorithm may be encoded in a JPEG progressive encoding manner. Among them, the typical JPEG progressive encoding scheme is to arrange the data after DCT transformation (DCT, Discrete Cosine Transform, commonly used image compression algorithm) in a zigzag scanning manner, and first encode the low-frequency components, and then display it on the decoding display side. See the rough outline of a frame, and then transmit the high-frequency components that carry detailed information.

Example two

Based on the image coding method described in the embodiment corresponding to FIG. 1, the following are embodiments of the disclosed device, which can be used to implement the method embodiments of the disclosure.

An embodiment of the present disclosure provides an image encoding device. As shown in FIG. 2, the image encoding device 20 includes: a first acquisition module 201, a first determination module 202, a layer generation module 203, and an image generation module 204;

The first acquisition module 201 is used to acquire the color information corresponding to the first macro block and the first macro block, the first macro block includes: text information and image information.

The first determining module 202 is configured to determine the text macroblock and the image macroblock in the first macroblock according to the color information corresponding to the first macroblock, and the image in the image macroblock is an image containing text information.

As shown in FIG. 3, in an optional embodiment, the first determining module 202 includes: an acquiring unit 2021, a first determining unit 2022, a second determining unit 2023, and a third determining unit 2024,

The acquiring unit 2021 is configured to acquire the number of color types in at least one sub-macroblock in the first macroblock according to the color information;

The first determining unit 2022 is configured to determine that the sub macro block is an image macro block when the number of color types in the sub macro block is greater than the preset number;

The second determining unit 2023 is configured to determine that the sub-macroblock is a text macroblock when the number of color types in the sub-macroblock is less than or equal to the preset number;

The third determining unit 2023 is configured to determine the text macro block and the image macro block in the first macro block according to the image macro block or the text macro block corresponding to the sub macro block.

The layer generating module 203 is configured to generate image layers and text layers after dividing the image macroblocks according to a preset algorithm.

As shown in FIG. 4, in an alternative embodiment, the layer generating module 203 in the device of the present disclosure includes a first obtaining unit 2031 and a generating unit 2032,

The first obtaining unit 2031 is configured to analyze the image macroblock, and obtain the high gradient pixels in the image macroblock according to the analysis result;

The generating unit 2032 is configured to generate a text layer according to the high gradient pixels, and generate an image layer according to pixels other than the high gradient pixels in the image.

As shown in FIG. 5, in an alternative embodiment, the layer generation module 203 of the device of the present disclosure includes a second acquisition unit 2033, a first generation unit 2034, and a second generation unit 2035,

The second acquiring unit 2033 is configured to acquire target position information and an average pixel value corresponding to the target position information, where the target position information refers to the position corresponding to pixels other than high-gradient pixels in the text layer information;

The first generating unit 2034 is configured to generate and determine preset pixels according to the target position information and the average pixel value corresponding to the target position information;

The second generating unit 3035 is configured to generate the text layer according to the high gradient pixels and the preset pixels.

As shown in FIG. 6, in an optional embodiment, the device 20 of the present disclosure further includes: an image acquisition module 205 and a second determination module 206,

The image acquisition module 205 is configured to acquire the frame image and the proportion information of at least one macroblock in the frame image. The proportion information includes: proportion information of high gradient pixels and proportion information of basic color pixels;

The second determining module 206 is configured to determine that at least one macroblock in the frame image is the first macroblock when the ratio information of at least one macroblock in the frame image meets a preset requirement.

The image encoding device provided by the embodiment of the present disclosure obtains the first macro block in the target frame image, and the first macro block, that is, in the prior art, determines the macro block containing text information and image information in the frame image as a text macro After dividing the first macro block according to the color information of the first macro block, the text macro block and the image macro block are generated. The image in the image macro block is an image containing text information; After the layering process, the text layer and the image layer are generated, the text encoding method is adopted for the text macroblock and the text layer map, and the image encoding method is adopted for the image layer layer. By improving the accuracy of recognizing text macroblocks in frame images, it is avoided that macroblocks containing image information and text information are recognized as text macroblocks, and matching coding algorithms are selected according to the different characteristics of text macroblocks and image macroblocks, thereby The coding effect of the frame image is improved, and then the visual effect and compression ratio effect of the frame image after the coding is improved.

Based on the image encoding method described in the embodiment corresponding to FIG. 1, embodiments of the present disclosure also provide a computer-readable storage medium. For example, the non-transitory computer-readable storage medium may be a read-only memory (English: Read Only Memory, ROM), random access memory (English: Random Access Memory, RAM), CD-ROM, magnetic tape, floppy disk and optical data storage device, etc. The storage medium stores computer instructions for executing the image encoding method described in the embodiment corresponding to FIG. 1, which will not be repeated here.

After considering the specification and practicing the disclosure disclosed herein, those skilled in the art will easily think of other embodiments of the present disclosure. This application is intended to cover any variations, uses, or adaptive changes of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or conventional technical means in the technical field not disclosed in the present disclosure . The description and the embodiments are to be regarded as exemplary only, and the true scope and spirit of the present disclosure are pointed out by the following claims.

Claims

An image coding method, characterized in that the method includes:

Acquiring a first macro block and color information corresponding to the first macro block, where the first macro block includes: text information and image information;

Determine, according to the color information corresponding to the first macro block, corresponding text macro blocks and image macro blocks in the first macro block, and the image in the image macro block is an image containing text information;

After dividing and processing the image macroblocks according to a preset algorithm, an image layer and a text layer are generated;

The text macroblock and the text layer are coded according to the first coding algorithm, and the image layer is coded according to the second coding algorithm to generate coded macroblocks.
The method according to claim 1, characterized in that, before said obtaining the first macroblock, the method further comprises:

Acquiring a frame image and ratio information of at least one macroblock in the frame image, where the ratio information includes: ratio information of high gradient pixels and ratio information of basic color pixels;

When the ratio information of at least one macro block in the frame image meets a preset requirement, it is determined that at least one macro block in the frame image is the first macro block.
The method according to claim 1, wherein the determining the text macroblock and the image macroblock in the first macroblock comprises:

Acquiring, according to the color information, the number of color types in at least one sub-macroblock in the first macroblock;

When the number of color types in the sub-macroblock is greater than a preset number, determining that the sub-macroblock is the image macroblock;

When the number of color types in the sub-macroblock is less than or equal to a preset number, determining that the sub-macroblock is the text macroblock;

Determine the text macro block and the image macro block in the first macro block according to the image macro block or the text macro block corresponding to the sub macro block.
The method according to claim 1, wherein said generating an image layer and a text layer comprises:

Parse the image macroblock, and obtain high gradient pixels in the image macroblock according to the analysis result;

A text layer is generated according to the high gradient pixels, and an image layer is generated according to pixels other than the high gradient pixels in the image.
The method according to claim 4, wherein said generating a text layer according to said high gradient pixels comprises:

Acquiring target position information and an average pixel value corresponding to the target position information, where the target position information refers to position information corresponding to pixels other than high gradient pixels in the text layer;

Determine a preset pixel according to the target position information and the average pixel value corresponding to the target position information;

The text layer is generated according to the high gradient pixels and the preset pixels.
An image coding device, characterized by comprising: a first acquisition module, a first determination module, a layer generation module, and an image generation module;

The first obtaining module is configured to obtain a first macro block and color information corresponding to the first macro block, and the first macro block includes: text information and image information;

The first determining module is configured to determine the text macroblock and the image macroblock in the first macroblock according to the color information corresponding to the first macroblock, and the image in the image macroblock contains text information Image;

The layer generating module is configured to generate an image layer and a text layer after the image macroblock is divided and processed according to a preset algorithm;

The image generation module is configured to encode the text macroblock and the text layer according to a first encoding algorithm, and generate an encoded macroblock after encoding the image layer according to a second encoding algorithm.
The device according to claim 6, further comprising: an image acquisition module and a second determination module,

The image acquisition module is configured to acquire a frame image and scale information of at least one macroblock in the frame image, and the scale information includes: scale information of high gradient pixels and scale information of basic color pixels;

The second determining module is configured to determine that at least one macroblock in the frame image is the first macroblock when the ratio information of at least one macroblock in the frame image meets a preset requirement.
The device according to claim 6, wherein the first determining module comprises: an acquiring unit, a first determining unit, a second determining unit, and a third determining unit,

The acquiring unit is configured to acquire the number of color types in at least one sub-macroblock in the first macroblock according to the color information;

The first determining unit is configured to determine that the sub-macroblock is the image macroblock when the number of color types in the sub-macroblock is greater than a preset number;

The second determining unit is configured to determine that the sub-macroblock is the text macroblock when the number of color types in the sub-macroblock is less than or equal to a preset number;

The third determining unit is configured to determine the text macro block and the image macro block in the first macro block according to the image macro block or the text macro block corresponding to the sub macro block.
The device according to claim 6, wherein the layer generating module includes a first acquiring unit and a generating unit,

The first obtaining unit is configured to analyze the image macroblock, and obtain high gradient pixels in the image macroblock according to the analysis result;

The generating unit is configured to generate a text layer according to the high gradient pixels, and generate an image layer according to pixels other than the high gradient pixels in the image.
The device according to claim 6, wherein the layer generating module comprises a second acquiring unit, a first generating unit and a second generating unit,

The second acquiring unit is configured to acquire target position information and an average pixel value corresponding to the target position information, where the target position information refers to position information corresponding to pixels other than high gradient pixels in the text layer ；

The first generating unit is configured to generate and determine preset pixels according to the target position information and the average pixel value corresponding to the target position information;

The second generating unit is configured to generate the text layer according to the high gradient pixels and the preset pixels.