CN1972445A - Video encoder and video encoding method capable of updating quantization table - Google Patents

Abstract

The invention discloses a Video Encoder capable of updating a Quantization Table, which is used for compressing Video information and is characterized by comprising a Random Access Memory (RAM) for storing the Quantization Table (Quantization Table), wherein the Quantization Table can be updated according to the requirements of a user or special environment so as to change the selection standard of Quantization (Quantization) and obtain better compression ratio. The invention also discloses a video coding method, which can update a Quantization Table (Quantization Table) according to the content of the video information in the video coding process, and quantize the video information of the frequency domain according to the Quantization Table so as to more flexibly remove redundant signals in the video information.

Description

Video encoder and video encoding method capable of updating quantization table

technical field

The present invention relates to a video encoder and an encoding method, in particular to a random access memory (RAM, Random Access Memory) that can store a quantization table (Quantization Table), and the quantization table can be configured according to user needs or special circumstances. is updated to change the selection criteria for Quantization.

Background technique

Video compression technology originates from the need to efficiently transmit and store video data in limited bandwidth and storage space. For example, a 640×480, 30frame/s, 30-minute full-color video requires about 47GB of storage space if it is not compressed, which obviously cannot achieve the benefits of video transmission.

Existing video compression technology mainly achieves the purpose of compression by removing redundancy, information that is not easily perceived by the human eye, and entropy coding. As shown in FIG. 1 , it is a schematic diagram of an existing video encoder. In the video encoder 10, the video information in the spatial domain is input into the conversion unit 101, and converted into video information in the frequency domain by means of DCT (Discrete Cosine Transformation). Quantization is then carried out, through the quantization unit 102, the quantization table 1021 (Quantization Table) is used to select or reject the video information in the frequency domain, and remove the high-frequency information in the video information that is not easy for human eyes to detect. Finally, entropy coding 103 is performed, such as Huffman or Arithmetic coding, to store information in an efficient manner.

The above-mentioned quantization process, in short, is to divide the video information in the frequency domain, and the denominator of the division is different according to the frequency in the quantization table, that is, the selection of the quantization scale (Quantization Scale). In principle, high-frequency information is not easy to detect by the human eye, so a larger scale should be used to make it as zero as possible; and a smaller scale should be used for low-frequency information so that the restored data will not differ too much from the original data.

In the existing video coder, the quantization table is fixed during the process of encoding and compressing video information. In other words, regardless of the content of the video information, the criteria for removing high and low frequency information are the same. This approach will lead to some problems. For special environments or scenes, it is impossible to provide a better compression ratio. If the compression is too high or not enough, the compression effect will be deteriorated, or the image quality will be deteriorated. For example, when the video information When there are high-frequency signals such as hair fluttering in the close-up view, the same quantization table, for this kind of scene, the compression ratio may be too high. After encoding and decoding, the fluttering hair may not be visible. For another example, when the background of the image is always static, the same quantization table may have a too small compression ratio, resulting in poor compression and wasting storage space. Therefore, it is necessary to improve on some shortcomings of commonly used technologies.

Contents of the invention

The purpose of the present invention is to establish a video coder that can update the quantization table, and use random access memory (RAM, Random Access Memory) to store the quantization table (Quantization Table), and the quantization table can be changed according to user needs or special circumstances It is updated to change the standard of quantization (Quantization), so that the result of video coding compression is more perfect.

Another object of the present invention is to propose a video encoding method, which can update the quantization table (Quantization Table) according to the content of the video information during the video encoding process, and quantize the video information in the frequency domain according to the quantization table. In order to more flexibly remove redundant signals in video information. At the same time, the compression rate can be increased without changing the image quality, or the image quality can be improved while maintaining the existing compression rate.

In order to achieve the above object, the present invention proposes a video encoder (Video Encoder) that can update the quantization table to compress a video information, and it has a random access memory (RAM, Random Access Memory) to store a quantization table ( Quantization Table), the quantization table can be updated during the process of the video encoder processing the video information, so as to change the selection standard of quantization (Quantization).

The video coder is Mpeg2 or Mpeg4 video coder.

The video encoder, wherein the random access memory is a volatile memory (Volatile Memory) or a non-volatile memory (NVRAM, Non-Volatile Random Access Memory).

The video encoder, wherein the volatile memory (Volatile Memory) is Static Random Access Memory (SRAM, Static Random Access Memory).

The video encoder further includes a memory device for storing quantization tables of several situations, so as to update the quantization tables of the Random Access Memory (RAM) according to various situations.

In order to achieve the above object, the present invention proposes a method for video encoding, the steps of which are: providing a video information; converting the video information into video information in a frequency domain; updating a Quantization Table (Quantization Table) according to the content of the video information ); and quantize the video information in the frequency domain according to the quantization table.

The video coding method is Mpeg2 or Mpeg4 video coding method.

In the video coding method, the quantization table is stored in a random access memory (RAM, Random Access Memory).

The video encoding method, wherein the random access memory is a volatile memory (VolatileMemory) or a non-volatile memory (NVRAM, Non-Volatile Random Access Memory).

The video encoding method, wherein the volatile memory (Volatile Memory) is Static Random Access Memory (SRAM, Static Random Access Memory).

The video encoding method further includes a storage device for storing quantization tables of several situations, so as to update the quantization tables of the Random Access Memory (RAM) according to various situations.

The above-mentioned invention proposes an improvement to the commonly used technology, utilizes a kind of video coder that can update the quantization table, stores the quantization table (Quantization Table) with random access memory (RAM, Random Access Memory), and the quantization table can be changed according to the user's demand Or special circumstances are updated, so as to change the quantization (Quantization) selection standard, so that the result of video coding compression is more perfect. The present invention further proposes a video encoding method. During the video encoding process, the quantization table (Quantization Table) can be updated according to the content of the video information, and the video information in the frequency domain can be quantized according to the quantization table, so as to more flexibly Remove unwanted signals from video messages.

Description of drawings

FIG. 1 is a schematic diagram of a conventional video encoder.

FIG. 2 is a schematic diagram of a video encoder with an updateable quantization table according to a preferred embodiment of the present invention.

FIG. 3 is a second schematic diagram of a video encoder with an updateable quantization table according to a preferred embodiment of the present invention.

Detailed ways

As shown in FIG. 2 , it is a schematic diagram of a video encoder with an updateable quantization table according to a preferred embodiment of the present invention. As shown in the figure, in the video coder 20, the video information in the spatial domain is input into the conversion unit 201, converted into video information in the frequency domain by means of DCT, etc., and then quantized by the quantization unit 202, and finally entropy coding 203 is performed to obtain Efficient ways such as Huffman or Arithmetic coding to store information.

The quantization table 2041 required for the quantization process is stored in a random access memory 204 (RAM, Random Access Memory) such as SRAM, and can be updated during the process of video information processing by the video encoder 20, so as to be updated according to the content of the video information. Change the removal criteria for high and low frequency information. For example, when there are many low-frequency signals and few high-frequency signals in the video information, you can use a larger scale for the high-frequency information to make it zero as much as possible; and use a smaller scale for the low-frequency information so that After its restoration, it will not differ too much from the original data. However, when the video information obviously only contains high-frequency signals, the quantization table can be updated to use a smaller scale for the high-frequency information, so as to avoid missing images after video decoding.

Fig. 3 is another embodiment of the present invention. In this embodiment, a memory device 30 is added, and the memory device can preset quantization tables of some frequently occurring situations. For example, the quantization table of the first case is the quantization table of static images, and the quantization table of the second case is the quantization table of nighttime images. , and the Nth case is the quantization table of the snow image. In addition, the quantization table 2041 can also be updated by the user. In this way, the application of the present invention can be made more flexible, and for some special images, the preset quantization table can be directly used to update.

Certainly, besides Mpeg2 or Mpeg4, the present invention can be applied to other compression methods useful for quantization table. Generally, there are two types of quantization operations: scalar and vector. The MPEG column uses scalar, but the vector quantization method is also a table, so this method can also be used for vector quantization. The purpose of changing the quantization table is one of the following: 1. Improve the compression rate without changing the image quality. 2. Improve the image quality while maintaining the existing compression rate.

To sum up, the present invention proposes an improvement to the commonly used technology, utilizes the video coder that establishes a kind of updatable quantization table, uses random access memory (RAM, Random Access Memory) to store the quantization table (Quantization Table), and the quantization table It can be updated according to user needs or special circumstances, so as to change the selection standard of quantization, so that the result of video coding compression is more perfect. The present invention further proposes a video encoding method. During the video encoding process, the quantization table (Quantization Table) can be updated according to the content of the video information, and the video information in the frequency domain can be quantized according to the quantization table, so as to more flexibly Remove unwanted signals from video messages.

Claims (11)

1. A video encoder (Video Encoder) capable of updating a quantization table, used to compress a video message, characterized in that it has a random access memory (RAM, Random Access Memory) to store a quantization table (Quantization Table) , the quantization table can be updated according to user needs or special circumstances, so as to change the selection criteria of quantization. 2. The video coder according to claim 1, wherein the video coder is an Mpeg2 or Mpeg4 video coder. 3. The video encoder according to claim 1, wherein the random access memory is a volatile memory (Volatile Memory) or a non-volatile memory (NVRAM, Non-Volatile Random Access Memory). 4. The video encoder according to claim 3, wherein the volatile memory (VolatileMemory) is Static Random Access Memory (SRAM, Static Random Access Memory). 5. The video coder as claimed in claim 3, further comprising a memory device for storing quantization tables of several situations, so as to update the random access memory (RAM, Random Access) according to various situations. Memory) the quantization table. 6. A method for video encoding, the steps of which are: provide a video message; converting the video information into a frequency domain video information; Updating a quantization table (Quantization Table) according to the content of the video information; and The video information in the frequency domain is quantized according to the quantization table. 7. The video coding method according to claim 6, characterized in that: said video coding method is Mpeg2 or Mpeg4 video coding method. 8. The video encoding method according to claim 6, wherein the quantization table is stored in a random access memory (RAM, Random Access Memory). 9. The video encoding method according to claim 8, wherein the random access memory can be a volatile memory (Volatile Memory) or a non-volatile memory (NVRAM, Non-Volatile Random Access Memory). 10. The video coding method according to claim 9, characterized in that: the volatile memory (Volatile Memory) is Static Random Access Memory (SRAM, Static Random Access Memory). 11. The video encoding method as claimed in claim 9, further comprising a memory device for storing quantization tables of several situations, so as to update the quantization tables according to various situations.

Images