CN101540916B - A codec method and device - Google Patents

Abstract

A codec method and device, which mainly includes: when encoding a macroblock in a current coded frame as an error recovery frame, identifying the macroblock in the error recovery frame and the corresponding position in at least one frame before the error recovery frame The transmission error sensitivity of the macroblock, the transmission error sensitivity refers to the degree of influence on the quality of the decoded image when a transmission error occurs in the macroblock; if it is determined that there is a macroblock whose transmission error sensitivity meets the predetermined condition, then the The macroblocks in the error recovery frame are coded in an intra-frame predictive coding manner. After receiving the code stream corresponding to the corresponding error recovery frame, the receiving end can terminate possible errors. The implementation of the embodiments of the present invention can not only effectively terminate the errors generated in the codec system, but also effectively reduce the impact of the insertion of the error recovery frame on the code stream of the transmission network, and can reduce the impact of the insertion of the error recovery frame on the codec system. impact on efficiency.

Description

A codec method and device

technical field

The present invention relates to the field of communication technology, in particular to a codec technology.

Background technique

In a video communication system, bit errors or packet loss are usually unavoidable. Moreover, after bit errors occur, not only the recovery of current data is affected, but also due to the predictive coding mechanism of video data, the corresponding bit errors are The time and space domains continue to diffuse, so that the error of a certain frame has the effect of error accumulation on subsequent frames. Therefore, it is necessary to take corresponding measures to improve the anti-error capability of video data, so that the corresponding errors can be terminated as soon as possible in the system.

At present, the commonly used solution for terminating corresponding errors is the I-frame (intra-frame predictive coded frame) refresh mechanism. In the I-frame refresh mechanism, an I-frame is inserted at regular intervals, and the P frame behind the I-frame is specified. Frames cannot be predicted with reference to image frames preceding the I-frame. In this way, after the I frame is received at the decoding end, errors that occur before will not affect the I frame and the frames after the I frame.

In the process of realizing the present invention, the inventors found that because the inserted I frame is coded by using the intra-frame prediction coding mode, there are at least the following problems in the above-mentioned prior art:

(1) Since the code rate of the I frame is relatively large, frequent insertion of the I frame will have a relatively large code stream impact on the transmission network, thereby affecting the corresponding transmission performance;

(2) Since the encoding efficiency of the I frame is low, if the I frame is frequently inserted, the efficiency of the entire codec system will be affected.

Contents of the invention

The embodiment of the present invention provides an encoding and decoding method, which can realize error termination under the premise of reducing the impact of the code stream on the transmission network and ensuring the efficiency of the encoding and decoding system.

An encoding method comprising:

When encoding a macroblock in a currently coded frame serving as an error recovery frame, identifying the transmission error sensitivity of the macroblock in the error recovery frame and correspondingly positioned macroblocks in at least one frame preceding the error recovery frame, said Transmission error sensitivity refers to the degree of influence on the quality of the decoded image when a transmission error occurs in the macroblock;

If it is determined that there is a macroblock whose transmission error sensitivity meets the predetermined condition, the macroblock in the error recovery frame is encoded by means of intra-frame prediction encoding.

An encoding device comprising:

The error sensitivity identification unit is used to identify the macroblock in the error recovery frame and the macroblock in the corresponding position in at least one frame before the error recovery frame when encoding the macroblock in the current coded frame as the error recovery frame Transmission error sensitivity, where the transmission error sensitivity refers to the degree of influence on the decoded image quality when a transmission error occurs in the macroblock;

The encoding unit is configured to perform an encoding operation on the macroblocks in the error recovery frame by means of intra-frame predictive encoding when the error sensitivity identification unit determines that there are macroblocks whose transmission error sensitivity meets a predetermined condition.

A decoding method comprising:

Determining that the current decoded frame is an error recovery frame according to the received code stream;

During the decoding process of the intra-frame predictive coded macroblock in the current decoding frame, search for the decoded intra-frame predictive coded macroblock adjacent to the currently decoded macroblock and in the same slice;

A decoding operation is performed on the currently decoded intra-frame predictively coded macroblock according to the found intra-frame predictively coded macroblock.

A decoding device, comprising:

An error recovery frame identification unit is used to determine whether the current decoded frame is an error recovery frame according to the received code stream;

An intra-frame prediction unit, configured to search for already-decoded intra-frame predictive coded macroblocks that are adjacent to the currently decoded intra-frame predictive coded macroblock and are in the same band when the error recovery frame identification unit identifies and determines whether the currently decoded frame is an error recovery frame Intra-prediction coded macroblocks;

The decoding unit is configured to perform a decoding operation on the currently decoded intra-frame prediction-coded macroblock according to the intra-frame prediction-coded macroblock found by the intra-frame prediction unit.

A decoding method comprising:

A flag symbol is included in the received code stream, which is used to indicate that the current frame is an error recovery frame. During the decoding process of the intra-frame predictive coded macroblock in the current decoded frame, search for the macroblock adjacent to the currently decoded macroblock and Decoded intra-frame predictive coded macroblocks in the same slice; perform decoding operation on the currently decoded intra-frame predictively coded macroblocks according to the found intra-frame predictively coded macroblocks;

In the process of decoding the first inter-frame predictive decoding frame after the error recovery frame, only the error recovery frame is used as a decoding reference frame.

It can be seen from the technical solutions provided by the above embodiments of the present invention that error recovery frames are used to terminate errors in the codec system, and the error recovery frames are only encoded in the corresponding intra-frame prediction coding mode when necessary Therefore, the implementation of the embodiment of the present invention can not only effectively terminate the errors generated in the codec system, but also effectively reduce the impact of the insertion of the error recovery frame on the code stream of the transmission network, and reduce the impact of the insertion of the error recovery frame on the transmission network. The impact on the efficiency of the codec system.

Description of drawings

FIG. 1 is a schematic diagram of an encoding process provided by an embodiment of the present invention;

FIG. 2 is a schematic diagram of a decoding process provided by an embodiment of the present invention;

FIG. 3 is a schematic structural diagram of a codec device provided by an embodiment of the present invention.

Detailed ways

In the embodiment of the present invention, at the coding end, when coding a macroblock in the current coded frame (that is, the current coded macroblock) as an error recovery frame, the macroblock in the error recovery frame and the previous macroblock in the error recovery frame can be identified. The transmission error sensitivity of the macroblock at the corresponding position in at least one frame, the transmission error sensitivity refers to the degree of influence on the quality of the decoded image when a transmission error occurs in the macroblock; and when it is determined that there is a transmission error sensitivity that meets the predetermined conditional macroblock, the macroblock in the error recovery frame (that is, the currently coded macroblock) is coded in the manner of intra-frame predictive coding, so as to avoid introducing errors that may occur in some coded macroblocks, Corresponding certain macroblocks may be macroblocks whose coded transmission error sensitivity meets a predetermined condition.

In the above process, it may specifically include: if any macroblock in the macroblock in the error recovery frame and the corresponding macroblock in at least one frame before the error recovery frame is an intra-coded macroblock or a motion vector greater than If there is an inter-coded macroblock with a predetermined threshold, it is determined that the macroblock is a macroblock whose transmission error sensitivity meets the predetermined condition (it can be called an error-sensitive macroblock), that is, it is determined that there is a macroblock whose transmission error sensitivity meets the predetermined condition. That is to say, the embodiment of the present invention can adopt multi-frame encoding mode and motion macroblock analysis to track the macroblock, and then determine the encoding mode of the error recovery frame, so as to end possible errors in the encoding process.

In order to facilitate the coding operation of the current coded macroblock and improve the coding efficiency, the transmission error sensitivity judgment result of the macroblock at the corresponding position in at least one frame before the error recovery frame can be saved in advance, that is, the predetermined number of coded frames can be saved The transmission error sensitivity flag corresponding to each macroblock is used as a basis for determining whether the transmission error sensitivity of each macroblock meets the predetermined condition; wherein, if the transmission error sensitivity flag corresponding to the macroblock is valid, it means that the The transmission error sensitivity of the macroblock meets the predetermined condition; in this way, the coding efficiency can be effectively improved.

In the embodiment of the present invention, when the predetermined error recovery frame insertion condition is met, the current encoding frame can be used as an error recovery frame for encoding operation; wherein, the corresponding error recovery frame insertion condition can include, but is not limited to: Interval between frames The number of frames, or, the interval to the last error recovery frame, or, the error sensitivity of the macroblocks contained in at least one frame before this coded frame, or, the Insert error recovery frame command sent by the device, etc.

In the embodiment of the present invention, in order to facilitate the decoding end to identify the corresponding error recovery frame, an error recovery frame flag (that is, an error recovery frame flag for indicating that the current frame is an error recovery frame) may also be inserted into the code stream corresponding to the error recovery frame. A flag symbol), through which the error recovery frame flag indicates that the current coded frame is an error recovery frame.

Corresponding to the processing procedure adopted by the encoding end above, in the embodiment of the present invention, after the decoding end determines that the current decoded frame is an error recovery frame according to the received code stream, the intra-frame predictive coding macroblock (such as predicted In the decoding process of the macroblock whose mode is DC mode), the decoded intra-predicted macroblock in the same slice as the currently decoded macroblock is searched, and the currently decoded macroblock is determined according to the found intra-predicted macroblock. Blocks are decoded to terminate possible errors in the system.

Wherein, the manner of determining that the current decoded frame is an error recovery frame at the decoding end may include but not limited to include: if there is an error recovery frame flag in the code stream corresponding to the current decoded frame, then determining that the current decoded frame is an error recovery frame.

The embodiments of the present invention can be specifically applied to a video surveillance system, or can also be applied to other similar codec systems, where corresponding error-sensitive areas and non-error-sensitive areas can be distinguished.

Taking the video surveillance system as an example, in the corresponding surveillance video sequence, due to less movement or switching of the camera, that is, a large part of the background area in the video sequence within a certain period of time has been static or has little movement. For the coding information of the background area in the video sequence, even if it is lost during transmission, it has little influence on the restoration quality of subsequent frames. Therefore, for the corresponding video surveillance system, the technical solutions provided by the embodiments of the present invention can be used to perform error termination processing.

In order to facilitate the understanding of the embodiments of the present invention, the processing procedures adopted by the encoding end and the decoding end in the embodiments of the present invention will be described in detail below.

(1) Coding end

Specifically, at the encoding end, a video encoding frame that prevents error accumulation can be set, that is, an error recovery frame (referred to as a RE frame), and the error-sensitive macroblock in the frame or a macroblock at a certain position is in several frames before the RE frame If it belongs to an error-sensitive macroblock, the macroblock is coded using intra-frame predictive coding, so that the RE frame can prevent error accumulation.

The specific processing methods that can be adopted at the encoding end are shown in Figure 1, and may include:

Step 101, set up M (M is a natural number greater than or equal to 1) storage spaces, and each storage space has at least the same basic storage unit as the number of macroblocks in a frame of video;

Step 102, for the M predicted frames before the RE frame, analyze information such as its motion vector or coding mode for each macroblock in each frame according to the order of the macroblocks, so as to analyze and determine the sensitivity of the macroblock to transmission errors (macro Block transmission error sensitivity), and then know whether the macroblock is an error-sensitive macroblock;

Specifically, the analysis of the corresponding macroblock error sensitivity is to analyze the degree of influence on the quality of the decoded image when a transmission error occurs in the macroblock; specifically, it can be based on at least one of the macroblock coding mode or the motion size Determine the transmission error sensitivity of the macroblock. For example, if the macroblock is an intra-coded macroblock, or if the macroblock is an inter-coded macroblock and the motion vector of the macroblock is greater than a certain threshold, then the macroblock can be considered is an error-sensitive macroblock;

Step 103, writing the transmission error sensitivity flag of the corresponding macroblock into the corresponding position in the storage space established in step 101, to indicate whether the macroblock is an error-sensitive macroblock;

Step 104, perform a corresponding encoding operation on each macroblock in the current RE frame according to the information in the storage space;

Specific encoding operations may include the following situations:

(1) When encoding a macroblock at a certain position in the RE frame, check the transmission error sensitivity flag of the macroblock at the corresponding position in the RE frame and the M frames before it, if any of the above frames correspond to If the transmission error sensitivity flag of the macroblock at the position is error sensitive (that is, the flag is valid), then intra-frame predictive coding is performed on the macroblock at this position, otherwise, normal intra-frame or inter-frame adaptive coding is performed; When the intra-predicted macroblock in the RE frame is predictively encoded, the reference data of the macroblock comes from the available adjacent macroblocks. The available adjacent macroblocks are: coded and in the same band as the current macroblock Neighboring macroblocks within and encoded with intra-frame prediction;

(2) When encoding an RE frame, an RE frame flag indicating that the frame type is an RE frame may be written in the image header stream of the frame.

Step 105, continue to obtain the corresponding coded code stream for the coding operation of other frames following the RE frame;

Wherein, it is forbidden to use the coded frame before the error recovery frame as a reference frame during the encoding operation of the interframe predictive coded frame after the error recovery frame, for example, for the case of using two reference frames, the RE frame follows For the first inter-frame prediction frame (that is, P frame), only one reference frame is used when encoding this frame and the reference frame is an RE frame, so that the frame after the encoded RE frame cannot refer to the frame before the RE frame for prediction.

In the process of performing the encoding operation at the encoding end, specifically, a corresponding RE frame flag can be set in the encoding code stream for each frame to indicate whether the current encoding frame is an RE frame, so that the decoding end can identify the corresponding RE frame , and perform corresponding decoding processing on the RE frame. For example, a flag bit can be used. If the current coded frame is not an RE frame, the flag bit is set to 0. If the current coded frame is an RE frame, the flag bit is set to is 1.

(2) Decoder

As shown in Figure 2, at the decoding end, the corresponding decoding process may include:

Step 201, receiving a section of video code stream;

Step 202, decoding the image header of the code stream;

Step 203, determining that the frame type is an RE frame according to the image header obtained through decoding;

For example, the frame can be determined to be an RE frame according to the error recovery frame flag in the image header;

Step 204, decoding the RE frame;

Specifically, decoding is performed on the current intra-frame predictive coded macroblock (such as a macroblock in DC prediction mode) in the RE frame, and the availability of adjacent macroblocks is judged, that is, whether the adjacent macroblocks have been decoded and are in the same position as the current macroblock In the same band and is an intra-predicted macroblock, if so, the adjacent macroblock is available, otherwise, the adjacent macroblock is not available; when it is determined that the adjacent macroblock is available, the current Decode macroblocks for decoding operations;

After the decoding operation of the RE frame is completed, the frame after the error recovery frame is also decoded, and in the process of decoding the inter-frame predictive decoding frame after the error recovery frame, it is forbidden to use the existing frame before the error recovery frame The decoded frame is used as the decoding reference frame. For example, in the case of using two decoding frames, only the error recovery frame is used as the decoding reference frame in the process of decoding the first inter-frame prediction decoding frame after the error recovery frame Perform decoding operation.

Through the above decoding process, errors that have occurred in the codec system, such as frame loss, can be effectively avoided.

The specific application manner of the embodiment of the present invention will be described below.

In the embodiment of the present invention, in the code stream obtained after encoding the RE frame, the specific syntax elements that can be included in the corresponding image header information are shown in Table 1:

Table 1

pb_picture_header(){ Descriptor pb_picture_start_code f(32) bbv_delay u(16) picture_coding_type u(2) picture_distance u(8) if(low_delay=='1') bbv_check_times ue(v) progressive_frame u(1) if(progressive_frame=='0'){ picture_structure u(1) if(picture_structure=='0') advanced_Pred_mode_disable u(1) }

top_field_first u(1) repeat_first_field u(1) fixed_picture_qp u(1) picture_qp u(6) if(!(picture_coding_type=='10'&&PictureStructure==1)) picture_reference_flag u(1) no_forward_reference_flag u(1) reserved_bits r(3) skip_mode_flag u(1)

loop_filter_disable u(1) if (! loop_filter_disable) { loop_filter_parameter_flag u(1) if(loop_filter_parameter_flag){ alpha_c_offset se(v) beta_offset se(v) } } next_start_code() }

In the image header information shown in Table 1 above, the corresponding image coding mode picture_coding_type is information added by the embodiment of the present invention, which can be specifically a 2-bit unsigned integer, which can be used to carry an error recovery frame flag to indicate the current frame It is an error recovery frame, and the encoding mode of the image corresponding to the picture_coding_type (that is, the image encoding type) can be referred to in Table 2 for details:

Table 2

picture_coding_type Encoding 00 error recovery frame 01 Forward predictive coding (P) 10 Bidirectional predictive coding (B) 11 reserve

In the embodiment of the present invention, for the case where two reference frames are used, the syntax elements of the first predicted frame P frame after the corresponding RE frame can be specifically shown in Table 3:

table 3

pb_picture_header(){ Descriptor pb_picture_start_code f(32)

bbv_delay u(16) picture_coding_type u(2) picture_distance u(8) if(low_delay=='1') bbv_check_times ue(v) progressive_frame u(1) if(progressive_frame=='0'){ picture_structure u(1) if(picture_structure=='0') advanced_pred_mode_disable u(1) }

top_field_first u(1) repeat_first_field u(1) fixed_picture_qp u(1) picture_qp u(6) if(!(picture_coding_type=='10'&&PictureStructure==1)) picture_reference_flag u(1) no_forward_reference_flag u(1) reserved_bits r(3) skip_mode_flag u(1) loop_filter_disable u(1) if (! loop_filter_disable) { loop_filter_parameter_flag u(1)

if(loop_filter_parameter_flag){ alpha_c_offset se(v) beta_offset se(v) } } next_start_code() }

In the information shown in Table 3, the corresponding picture reference flag picture_reference_flag is used to indicate that the preceding frame is an RE frame. Specifically, when the picture_reference_flag flag value is '1', it means that all macroblocks use the default reference image (that is, the RE frame is the reference image of the first P frame after this frame), and the picture_reference_flag flag value is '0' Indicates that each macroblock determines the corresponding reference image by itself at the macroblock layer.

Based on the information described in Table 1 to Table 3 above, if the picture_coding_type of the previous reference frame of the current frame is '00' and the picture_coding_type of the current frame is '01', then the picture_reference_flag of the current frame is 1 to indicate that the current frame only uses RE frame as a reference frame.

Still based on the information described in Table 1 to Table 3 above, after receiving a piece of video code stream at the decoding end, and obtaining the picture_coding_type parameter in the corresponding image header information from the decoded code stream, then:

(1) If picture_coding_type is equal to '00', it indicates that the current decoding frame is an RE frame, and the macroblock in the RE frame is decoded, wherein, if the macroblock is in an intra-frame coding mode (such as DC prediction mode, etc.) , to determine the availability of adjacent macroblocks. Availability refers to intra-frame predicted macroblocks that have been decoded and are in the same band as the current macroblock. If the macroblock is in other coding modes, the current There is a corresponding decoding method in the technology for normal decoding;

(2) If the picture_coding_type is not equal to '00', the frame can be decoded normally by using the corresponding decoding method in the prior art, but not limited to.

An embodiment of the present invention also provides an encoding device, the specific implementation structure of which is shown in Figure 3, and may specifically include the following processing units:

(1) An error sensitivity identification unit, used to identify the macroblock in the error recovery frame and the corresponding position in at least one frame before the error recovery frame when encoding the macroblock in the current coded frame as the error recovery frame The transmission error sensitivity of the macroblock, the transmission error sensitivity refers to the degree of influence on the quality of the decoded image when a transmission error occurs in the macroblock;

Wherein, the corresponding error sensitivity identification unit may specifically include at least one of the following units:

A coding mode identification unit, configured to determine that any macroblock in the macroblock of the error recovery frame and the macroblock at the corresponding position in at least one frame before the error recovery frame is an intra-coded macroblock, then determine that there is transmission Macroblocks whose error sensitivity meets predetermined conditions;

A motion vector identifying unit, configured to identify any macroblock in the macroblock of the error recovery frame and the macroblock at the corresponding position in at least one frame before the error recovery frame as an interframe coded macroblock with a motion vector greater than a predetermined threshold , it is determined that there is a macroblock whose transmission error sensitivity meets the predetermined condition.

(2) An encoding unit, configured to encode the macroblocks in the error recovery frame using intra-frame predictive encoding when the error sensitivity identification unit determines that there are macroblocks whose transmission error sensitivity meets the predetermined conditions. ;

The encoding unit can also be used to encode the frames after the error recovery frame, and during the encoding operation of the interframe predictive coded frame after the error recovery frame, it is forbidden to use the coded frame before the error recovery frame frame as a reference frame, for example, in the case of using two reference frames, during the encoding operation of the first interframe predictive coded frame after the error recovery frame, only the error recovery frame is referred to, that is, only the error recovery frame is used frame as a coding reference frame.

Optionally, in order to further improve the coding efficiency, the device may also include a storage unit for saving a predetermined number of transmission error sensitivity flags corresponding to each macroblock of the coded frame, and storing the stored macroblocks respectively The corresponding transmission error sensitivity flag is used as the basis for determining whether the transmission error sensitivity of each macroblock meets the predetermined condition; wherein, if the transmission error sensitivity flag corresponding to the macroblock is valid, it indicates that the transmission error sensitivity of the macroblock is The predetermined condition is met, that is, the macroblock is an error-sensitive macroblock.

Optionally, the device may also include an error recovery frame insertion control unit, configured to notify the error sensitivity identification unit to process the current coded frame as an error recovery frame when a predetermined error recovery frame insertion condition is met; Wherein, the error recovery frame insertion condition applied by the error recovery frame insertion control unit may include, but not limited to: the number of interval frames from the last error recovery frame, or the interval time from the last error recovery frame , or, the error sensitivity of the macroblock contained in at least one frame preceding the coded frame, or receiving an instruction to insert an error recovery frame sent from a controller or a decoder, and so on.

Optionally, the apparatus may further include a neighboring macroblock determining unit, configured to determine available neighboring macroblocks for the macroblocks in the error recovery frame, and provide them to the coding unit, so that according to the available Encode the macroblocks in the error recovery frame, the available adjacent macroblocks are coded adjacent macroblocks that are in the same slice as the current macroblock and are encoded by intra-frame prediction piece.

Optionally, in order to facilitate the identification of the corresponding error recovery frame by the decoding end, the device may further include a corresponding error recovery frame flag insertion unit, which is used to encode the code stream corresponding to the error recovery frame obtained by the coding unit An error recovery frame flag is inserted, and the error recovery frame flag indicates that the current coded frame is an error recovery frame.

The embodiment of the present invention also provides a decoding device, whose specific implementation structure is still shown in Figure 3, and may specifically include the following units:

(1) error recovery frame identification unit, used to determine whether the current decoded frame is an error recovery frame according to the received code stream; the error recovery frame identification unit may specifically include:

The identification unit is configured to determine that the current decoding frame is an error recovery frame when there is an error recovery frame flag in the code stream corresponding to the current decoding frame.

(2) An intra-frame prediction unit, configured to search for an intra-frame prediction encoded macroblock adjacent to the currently decoded macroblock and within the same band when the error recovery frame identification unit identifies and determines whether the currently decoded frame is an error recovery frame Intra-prediction coded macroblocks that have been decoded;

(3) A decoding unit, configured to perform decoding operations on the currently decoded intra-frame predictively coded macroblock according to the intra-frame predictively coded macroblock found by the intra-frame prediction unit, for example, it can be used to perform a decoding operation on a prediction mode that is a DC mode Macroblocks are decoded;

The decoding unit can also be used to decode the frame after the error recovery frame, and in the process of decoding the inter-frame prediction decoding frame after the error recovery frame, it is forbidden to use the decoded frame before the error recovery frame As a decoding reference frame, for example, in the case of using two decoding frames, only the error recovery frame is used as a decoding reference frame for decoding in the process of decoding the first inter-frame prediction decoding frame after the error recovery frame operate.

To sum up, the embodiment of the present invention can specifically analyze the motion of a video sequence to determine the areas that are sensitive to transmission errors in a video, and encode the sensitive areas using intra-frame coding. Therefore, the embodiment of the present invention can ensure the anti-error diffusion function of the frame on the basis of ensuring the coding performance, thereby improving the subjective and objective quality of the displayed video at the decoding end after the video is compressed and transmitted.

Through the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be realized by hardware, or by software plus a necessary general hardware platform. Based on this understanding, the technical solution of the present invention can be written into a computer executable program, and the computer device can realize the method provided by the present invention by executing the program. The computer executable program can be stored in a non-volatile storage medium (including CD-ROM, U disk, mobile hard disk, etc.).

The above is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present invention can easily think of changes or Replacement should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims (21)

1. a coding method is characterized in that, comprising:

When encoding as the macro block in the current encoded frame of error recovery frame, discern the transmission error sensitiveness of the macro block of correspondence position at least one frame before this macro block of this error recovery frame and this error recovery frame, described transmission error sensitiveness is meant when the macro block generation transmission error of correspondence position in described at least one frame the influence degree to decoded image quality;

As if the macro block of determining to exist transmission error sensitiveness to conform to a predetermined condition, then adopt the mode of intraframe predictive coding to carry out encoding operation the macro block in the described error recovery frame.

2. method according to claim 1 is characterized in that, describedly determines that the macro block that exists transmission error sensitiveness to conform to a predetermined condition comprises:

If the arbitrary macro block at least one frame before this macro block of error recovery frame and the error recovery frame in the macro block of correspondence position is inter-coded macroblocks or is the inter-coded macroblocks of motion vector greater than predetermined threshold, then determines the macro block that exists transmission error sensitiveness to conform to a predetermined condition.

3. method according to claim 1 and 2 is characterized in that, this method also comprises:

Preserve the corresponding respectively transmission error sensitivity label of each macro block of the coded frame of predetermined quantity, be used for the foundation that whether conforms to a predetermined condition as the transmission error sensitiveness of determining described each macro block; Wherein, if the transmission error sensitivity label of macro block correspondence is effective, represent that then the transmission error sensitiveness of this macro block conforms to a predetermined condition.

4. method according to claim 1 and 2 is characterized in that, this method also comprises:

When meeting predetermined error recovery frame insertion condition, then current encoded frame is carried out encoding operation as error recovery frame; Wherein, described error recovery frame insertion condition comprises: and the interval number of frames between the last error recovery frame, perhaps, and the blanking time between the last error recovery frame, perhaps, the mistake sensitiveness of the macro block that at least one frame before this coded frame comprises perhaps, receives the insertion error recovery frame instruction that slave controller or decoder send.

5. method according to claim 1 and 2 is characterized in that, the step that the mode of described employing intraframe predictive coding is carried out encoding operation comprises:

According to available adjacent macroblocks the macro block in the described error recovery frame is carried out intraframe predictive coding, described available adjacent macroblocks is: encoded, be in the same band and adjacent macroblocks that adopt intraframe predictive coding with current macro.

6. method according to claim 1 and 2 is characterized in that, this method also comprises:

Inserting the error recovery frame sign in the code stream of described error recovery frame correspondence, is error recovery frame by this error recovery frame sign indication current encoded frame.

7. method according to claim 1 and 2 is characterized in that, this method also comprises:

Forbid with the coded frame before this error recovery frame as the reference frame in the encoding operation process to the inter prediction encoding frame after the described error recovery frame.

8. a code device is characterized in that, comprising:

Mistake sensitiveness recognition unit, be used at coding during as the macro block of the current encoded frame of error recovery frame, discern the transmission error sensitiveness of the macro block of correspondence position at least one frame before this macro block of this error recovery frame and this error recovery frame, described transmission error sensitiveness is meant when the macro block generation transmission error of correspondence position in described at least one frame the influence degree to decoded image quality;

Coding unit is used for adopting the mode of intraframe predictive coding to carry out encoding operation the macro block in the described error recovery frame when described mistake sensitiveness recognition unit determines to exist the macro block that transmission error sensitiveness conforms to a predetermined condition.

9. device according to claim 8 is characterized in that, described mistake sensitiveness recognition unit specifically comprises following at least one unit:

The coding mode recognition unit, be used for when identification determines that arbitrary macro block in the macro block of this macro block of error recovery frame and at least one the frame correspondence position before the error recovery frame is inter-coded macroblocks, then determining the macro block that exists transmission error sensitiveness to conform to a predetermined condition;

The motion vector recognition unit, the arbitrary macro block that is used in the macro block of this macro block of discerning definite error recovery frame and at least one the frame correspondence position before the error recovery frame is the inter-coded macroblocks of motion vector greater than predetermined threshold, then determines the macro block that exists transmission error sensitiveness to conform to a predetermined condition.

10. according to Claim 8 or 9 described devices, it is characterized in that this device also comprises:

Memory cell is used to preserve the corresponding respectively transmission error sensitivity label of each macro block of the coded frame of predetermined quantity, is used for the foundation that whether conforms to a predetermined condition as the transmission error sensitiveness of determining described each macro block; Wherein, if the transmission error sensitivity label of macro block correspondence is effective, represent that then the transmission error sensitiveness of this macro block conforms to a predetermined condition.

11. according to Claim 8 or 9 described devices, it is characterized in that this device also comprises:

Error recovery frame is inserted control unit, be used for when meeting predetermined error recovery frame insertion condition, notify described mistake sensitiveness recognition unit that current encoded frame is handled as error recovery frame, the error recovery frame insertion condition that described error recovery frame is inserted the control unit application comprises: and the interval number of frames between the last error recovery frame, perhaps, and the blanking time between the last error recovery frame, perhaps, the mistake sensitiveness of the macro block that at least one frame before this coded frame comprises, perhaps, receive the insertion error recovery frame instruction that slave controller or decoder send.

12. device according to claim 11 is characterized in that, this device also comprises:

The adjacent macroblocks determining unit, be used to the macro block in the described error recovery frame to determine available adjacent macroblocks, and offer described coding unit, so that according to available adjacent macroblocks the macro block in the described error recovery frame is encoded, described available adjacent macroblocks is: encoded, be in the same band and adjacent macroblocks that adopt intraframe predictive coding with current macro.

13. according to Claim 8 or 9 described devices, it is characterized in that this device also comprises:

The error recovery frame sign inserts the unit, is used for the code stream insertion error recovery frame sign in the error recovery frame correspondence of described coding unit coding acquisition, is error recovery frame by this error recovery frame sign indication current encoded frame.

14. according to Claim 8 or 9 described devices, it is characterized in that, described coding unit also is used for the frame after the described error recovery frame is encoded, and in encoding operation process, forbid with the coded frame before this error recovery frame as the reference frame to the inter prediction encoding frame after the described error recovery frame.

15. a coding/decoding method is characterized in that, comprising:

Determine that according to the code stream that receives current decoded frame is an error recovery frame;

Intraframe predictive coding macro block in current decoded frame carries out in the decode procedure, and it is adjacent and be in the intraframe predictive coding macro block of having decoded in the same band to search described macro block with current decoding;

According to the intraframe predictive coding macro block that finds the intraframe predictive coding macro block of current decoding is carried out decode operation.

16. method according to claim 15 is characterized in that, the process that described definite current decoded frame is an error recovery frame comprises:

If have the error recovery frame sign in the code stream of current decoded frame correspondence, determine that then current decoded frame is an error recovery frame.

17., it is characterized in that this method also comprises according to claim 15 or 16 described methods:

In the process that the interframe prediction decoding frame after the described error recovery frame is decoded, forbid with the decoded frame before the described error recovery frame as the decoding reference frame.

18. a decoding device is characterized in that, comprising:

The error recovery frame recognition unit is used for determining according to the code stream that receives whether current decoded frame is error recovery frame;

Intraprediction unit, be used for when the identification of described error recovery frame recognition unit determines whether current decoded frame is error recovery frame, it is adjacent and be in the intraframe predictive coding macro block of having decoded in the same band to search intraframe predictive coding macro block with current decoding;

Decoding unit is used for the intraframe predictive coding macro block that finds according to described intraprediction unit the intraframe predictive coding macro block of current decoding is carried out decode operation.

19. device according to claim 18 is characterized in that, described error recovery frame recognition unit specifically comprises:

Recognition unit is used for when there is the error recovery frame sign in the code stream of current decoded frame correspondence, determines that current decoded frame is an error recovery frame.

20. according to claim 18 or 19 described devices, it is characterized in that, described decoding unit also is used for the frame after the described error recovery frame is decoded, and in the process that the interframe prediction decoding frame after this error recovery frame is decoded, forbid with the decoded frame before this error recovery frame as the decoding reference frame.

21. a coding/decoding method is characterized in that, comprising:

In the code stream that receives, comprise a tag mark, being used to indicate present frame is error recovery frame, intraframe predictive coding macro block in current decoded frame carries out in the decode procedure, and it is adjacent and be in the intraframe predictive coding macro block of having decoded in the same band to search described macro block with current decoding; According to the intraframe predictive coding macro block that finds the intraframe predictive coding macro block of current decoding is carried out decode operation;

In the process that the first interframe prediction decoding frame after the described error recovery frame is decoded, only with described error recovery frame as the decoding reference frame.

Images