CN102006473B - Video encoder and encoding method, and video decoder and decoding method - Google Patents
Video encoder and encoding method, and video decoder and decoding method Download PDFInfo
- Publication number
- CN102006473B CN102006473B CN 201010549778 CN201010549778A CN102006473B CN 102006473 B CN102006473 B CN 102006473B CN 201010549778 CN201010549778 CN 201010549778 CN 201010549778 A CN201010549778 A CN 201010549778A CN 102006473 B CN102006473 B CN 102006473B
- Authority
- CN
- China
- Prior art keywords
- background
- macro block
- image
- foreground
- prospect
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The invention provides a video encoding device comprising a video input module, a background updating module, a background modeling module, a background encoding module, a foreground dividing module, a foreground encoding module and a transmitting module, the video input module is used for inputting video images; the background updating module is used for updating the background in the video images when the background in the video images is changed; the background modeling module is used for establishing a background model according to the background in the updated video images; the background encoding module is used for encoding the background model so as to obtain a background code stream; the foreground dividing module is used for dividing the foreground region in the video images according to the input video images and the background model; the foreground encoding module is used for encoding the obtained foreground region so as to obtain a foreground code stream; and the transmitting module is used for respectively transmitting the background code stream and the foreground code stream.
Description
[technical field]
The present invention relates to field of video compression, particularly relate to a kind of video encoder and coding method and Video Decoder and coding/decoding method.
[background technology]
Along with multimedia application is more and more extensive, video technique also is applied in increasing field, such as video conference, video monitoring etc.Because the data volume of video is very large, can take too much broadband resource in transmission course, and produce very large time delay.Therefore, the Coding Compression Technology that has proposed video takies the problem of broadband resource and time delay to overcome the video data volume.
Video coding and transmission technology develop into today, have had a lot of coded systems and standard.But, also do not have a kind of coded system and standard to design for static camera.In actual applications, in a lot of video acquisition transmission systems, camera all is static.Such as, video conference and video monitoring etc.Video monitoring particularly, in recent years, video monitoring has been obtained tremendous development, and the laying amount of camera sharply increases.The increase of camera number has brought an important problem, and that is exactly transmission and the storage of video.Adopt traditional coded system, in order to reach very low code check, picture quality is compressed can't standing often, and simultaneously, code check is real the reduction not, causes the waste of resource.Simultaneously, in video, often important target limits, and existing method does not have yet and can distinguish image-region, gives prior target with higher Data Rate Distribution.
Therefore being necessary to propose a kind of new technical scheme solves the problems referred to above.
[summary of the invention]
The purpose of this part is to summarize some aspects of embodiments of the invention and briefly introduces some preferred embodiments.In this part and the application's specification digest and denomination of invention, may do a little simplification or omit to avoid making the purpose of this part, specification digest and denomination of invention fuzzy, and this simplification or omit and can not be used for limiting the scope of the invention.
One of purpose of the present invention is to provide a kind of video encoder and coding method, its with background and prospect separately, in the constant situation of background, a transmission primaries background, thereby when transmission of video, can greatly reduce the occupancy in broadband.
Two of purpose of the present invention is to provide a kind of Video Decoder and method, and it utilizes the method based on blending that prospect and background are merged stack, has well reduced original video image.
According to an aspect of the present invention, the invention provides a kind of video encoder, it comprises: video input module, inputted video image; The background in the new video image more when context update module, change of background in described video image; The background modeling module is set up background model according to the background in the described video image that upgrades; The background coding module is encoded with the background extraction code stream to described background model; The foreground segmentation module is partitioned into foreground area in the described video image according to video image and the described background model of input; The prospect coding module encodes to obtain the prospect code stream to the foreground area that obtains; Transport module transmits respectively described background code stream and prospect code stream.
Further, described transport module comprises background image transport module and foreground image transport module, and described background is stated the background code stream after image transmission module is used for transfer encoding, and described foreground image transport module is used for the prospect code stream after the transfer encoding.
Further, described method based on the max-flow dividing method comprises: the Gauss model that obtains the color component of each background pixel in the serial background image; Pixel in the video image and background image are carried out pixel value difference, to described difference greater than the pixel of threshold value then as foreground pixel; Obtain the mixed Gauss model of the color component of described foreground pixel; Adopt the described mixed Gauss model of the described Gauss model of background and prospect to calculate the probability that all pixels in the video image belong to prospect or background, obtain this pixel corresponding node after the normalization and be connected with the background node with prospect respectively in the drawings and be connected energy; Described video image is asked respectively horizontal edge intensity and vertical edge intensity between its neighbor; Being connected horizontal edge intensity with neighbor according to each pixel is connected with vertical edge intensity at the energy that is connected between pixel described in the video image and neighbor; According to all internodal connection energy that obtain, adopt the max-flow dividing method to carry out smallest partition, obtain prospect and background pixel in the described video image.
Further, described prospect coding module encodes to form the prospect code stream to the binary mask of the effective macro block in the foreground area and identical foreground area, wherein said binary mask is used for which position of sign video image and belongs to foreground area, each pixel that belongs to prospect is marked as 1 on the mask, otherwise is labeled as 0; Effective macro block acquisition methods in the described foreground area is: obtain its minimum boundary rectangle according to the binary mask of foreground area; image-region in the described boundary rectangle is divided into foursquare macro block; if comprise foreground pixel in the macro block then judge that this macro block is effective macro block; need coding; otherwise be judged to be invalid macro block, do not encode.
Further, described prospect coding module is encoded to the effective macro block in the foreground area and is comprised: to effective foreground macro block of the first frame appearance, order according to macro block is processed successively, described macro block is carried out discrete cosine transform, and discrete cosine transform coefficient quantized, adopt the differential coding mode that discrete cosine transform coefficient is predicted, and then carry out the entropy coding; Each frame to follow-up appearance, judge whether each macro block is effective in the foreground area, effective macro block is carried out following processing in order: to current macro, in the hunting zone of former frame image, seek and its effective foreground macro block of mating most, it is poor that effective foreground macro block of the described coupling in the former frame image and described current macro are done, the difference that obtains is carried out discrete cosine transform, carry out the entropy coding after its discrete cosine transform coefficient is quantized.
Further, described prospect coding module is encoded to the binary mask of foreground area and is comprised: the boundary rectangle scope of foreground area number and each foreground area in the document image, whether also will record in the image-region in the boundary rectangle scope each macro block effective, be 1 if any this tick lables effectively then, otherwise be masked as 0, adopt the entropy coded system to encode the sign of each macro block, and form bit stream with the macroblock coding result.
The present invention also provides a kind of method for video coding, and it comprises: inputted video image; When the change of background of described video image, upgrade its background; The background of described renewal is set up background model and described background model is encoded to obtain the background code stream; Be partitioned into the foreground area in the described video image and described foreground area encoded to obtain the prospect code stream according to described video image and described background model; Transmit respectively described background code stream and prospect code stream.
Further, utilize the max-flow dividing method that the foreground area in the described video image is cut apart, it comprises: the Gauss model that obtains the color component of each background pixel in the serial background image; Pixel in the video image and background image are carried out pixel value difference, to described difference greater than the pixel of threshold value then as foreground pixel; Obtain the mixed Gauss model of the color component of described foreground pixel; Adopt the described mixed Gauss model of the described Gauss model of background and prospect to calculate the probability that all pixels in the video image belong to prospect or background, obtain this pixel corresponding node after the normalization and be connected with the background node with prospect respectively in the drawings and be connected energy; Described video image is asked respectively horizontal edge intensity and vertical edge intensity between its neighbor; Being connected horizontal edge intensity with neighbor according to each pixel is connected with vertical edge intensity at the energy that is connected between pixel described in the video image and neighbor; According to all internodal connection energy that obtain, adopt the max-flow dividing method to carry out smallest partition, obtain prospect and background pixel in the described video image.
Further, the binary mask of the effective macro block in the foreground area and identical foreground area is encoded to obtain the prospect code stream, wherein said binary mask is used for which position of sign video image and belongs to foreground area, each pixel that belongs to prospect is marked as 1 on the mask, otherwise is labeled as 0; Effective macro block acquisition methods in the described foreground area is: obtain its minimum boundary rectangle according to the binary mask of foreground area; image-region in the described boundary rectangle is divided into foursquare macro block; if comprise foreground pixel in the macro block then judge that this macro block is effective macro block; need coding; otherwise be judged to be invalid macro block, do not encode.
Further, effective macro block in the foreground area encoded comprise: effective foreground macro block that the first frame is occurred, order according to macro block is processed successively, described macro block is carried out discrete cosine transform, and discrete cosine transform coefficient quantized, adopt the differential coding mode that discrete cosine transform coefficient is predicted, and then carry out the entropy coding; Each frame to follow-up appearance, judge whether each macro block is effective in the foreground area, effective macro block is carried out following processing in order: to current macro, in the hunting zone of former frame image, seek and its effective foreground macro block of mating most, it is poor that effective foreground macro block of the described coupling in the former frame image and described current macro are done, the difference that obtains is carried out discrete cosine transform, carry out the entropy coding after its discrete cosine transform coefficient is quantized.
Further, the binary mask of identical foreground area encoded comprise: the boundary rectangle scope of foreground area number and each foreground area in the document image, whether also will record in the image-region in the boundary rectangle scope each macro block effective, be 1 if any this tick lables effectively then, otherwise be masked as 0, the sign employing entropy coded system of each macro block is encoded.
Further, the coding result of effective macro block in the foreground area and the coding result of the binary mask of the foreground area of correspondence are formed code stream to transmit.
According to a further aspect in the invention, the invention provides a kind of Video Decoder, it comprises: parser modules, the code stream that receives is resolved, and the background code stream after will resolving and prospect code stream are separately; The background decoder module if receive the background code stream after the parsing then it is decoded with background image updating, otherwise does not process; The prospect decoder module obtains the foreground area number the prospect code stream after resolving, and obtains coordinate, macroblock coding information and the macro block effective marker information of the boundary rectangle of each foreground area; The video synthesis module, obtain the background image of present image, arrive the upper left corner of foreground area according to the boundary rectangle coordinate setting of each foreground area, according to the order identical with coding, judge whether the foreground macro block that macro block is corresponding in the boundary rectangle is effective macro block, if be invalid macro block then jump to next macro block and process, otherwise read the coded message of effective macro block, obtain discrete cosine transform coefficient after the decoding, the Cost fun ction image, from last two field picture, find its coupling macro block, recover the current macro image, just like that you cross current macroblock image and cover correspondence position on the present image; Or effective macro block is added on the background image.
Further, described former frame image is the former frame image after the video decoder decodes.
The present invention also provides a kind of video encoding/decoding method, and it comprises: the code stream that receives is resolved, and the background code stream after will resolving and prospect code stream are separately; If receive the background code stream after the parsing then it is decoded with background image updating, otherwise do not process; Obtain the foreground area number the prospect code stream after resolving, and obtain coordinate, macroblock coding information and the macro block effective marker information of the boundary rectangle of each foreground area; Obtain the background image of present image, arrive the upper left corner of foreground area according to the boundary rectangle coordinate setting of each foreground area, according to the order identical with coding, judge whether the foreground macro block that macro block is corresponding in the boundary rectangle is effective macro block, if be invalid macro block then jump to next macro block and process, otherwise read the coded message of effective macro block, obtain discrete cosine transform coefficient after the decoding, the Cost fun ction image, from last two field picture, find its coupling macro block, recover the current macro image, just like that you cross current macroblock image and cover correspondence position on the present image; Or effective macro block is added on the background image.
Further, described former frame image is the former frame image after the video decoder decodes.
Compared with prior art, video encoder is for the Video coding of camera static scene among the present invention, adopt foreground segmentation algorithm separation prospect from background, the background that does not change is only transmitted once, and the foreground area that comprises important content is carried out coding transmission.For with prospect and cleaner the cutting apart of background, described video encoder has adopted the method for cutting apart based on max-flow, can be meticulousr obtain segmenting edge.In addition, in order better to obtain the quality of composograph, Video Decoder adopts the method based on blending that prospect and background are merged stack.Owing to do not need to transmit background image when background is constant, thereby when transmission of video, greatly reduce the broadband occupancy.
[description of drawings]
In order to be illustrated more clearly in the technical scheme of the embodiment of the invention, the accompanying drawing of required use was done to introduce simply during the below will describe embodiment, apparently, accompanying drawing in the following describes only is some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain according to these accompanying drawings other accompanying drawing.Wherein:
Fig. 1 is a kind of video encoder structural representation in one embodiment among the present invention;
Fig. 2 is the flow chart of the foreground segmentation algorithm cut apart based on max-flow in one embodiment of the present invention;
Fig. 3 is a kind of method for video coding flow chart in one embodiment among the present invention;
Fig. 4 is a kind of Video Decoder structural representation in one embodiment among the present invention;
Fig. 5 is a kind of video encoding/decoding method flow chart in one embodiment among the present invention.
[embodiment]
Detailed description of the present invention is mainly come the running of direct or indirect simulation technical solution of the present invention by program, step, logical block, process or other symbolistic descriptions.Be the thorough the present invention that understands, in ensuing description, stated a lot of specific detail.And when not having these specific detail, the present invention then may still can realize.Affiliated those of skill in the art use herein these descriptions and statement essential to the work that the others skilled in the art in the affiliated field effectively introduce them.In other words, be the purpose of the present invention of avoiding confusion, because the easily understanding of the method for knowing and program, so they are not described in detail.
Alleged " embodiment " or " embodiment " refer to be contained in special characteristic, structure or the characteristic at least one implementation of the present invention herein.Different local in this manual " in one embodiment " that occur not are all to refer to same embodiment, neither be independent or the embodiment mutually exclusive with other embodiment optionally.In addition, represent the sequence of modules in method, flow chart or the functional block diagram of one or more embodiment and revocablely refer to any particular order, also be not construed as limiting the invention.
The invention provides a kind of video encoder, it is divided into video image background and prospect and carries out respectively coding transmission, and when context update the transmission primaries background image, and when background is not upgraded, do not carry out the transmission of background image.The concrete structure of described video encoder can be referring to shown in Figure 1.
Fig. 1 is video encoder structural representation in one embodiment among the present invention.See also shown in Figure 1ly, described video encoder 100 comprises video input module 110, context update module 120, background modeling module 130, background coding module 140, foreground segmentation module 150, prospect coding module 160 and transport module 170.Described video encoder 100 adopts foreground segmentation algorithm separation prospect from background for the Video coding of camera static scene, the background that does not change is only transmitted once or several, and the foreground region image that comprises important content is carried out coding transmission.
Described video input module 110 is used for the video image that input needs coding transmission, and described video image can also can be many continuous video images for individual.
Described context update module 120 background in the new video image more then when the change of background of described video image.In actual applications, the first time during inputted video image then with its initial background as the first time context update process, next detect video image background whether variation occured, then upgrade the background of primary video image if change.
Background after 130 pairs of described video images of described background modeling module upgrade is carried out background modeling to obtain background model.Described background is relatively motionless part of video image midium or long term, as in video monitoring, and the long-term relatively motionless part such as the number in the monitoring camera visual field, house, street.Corresponding, in concrete the application, described background modeling module 130 is then carried out modeling to the initial background image when video image is inputted for the first time, then the background after upgrading is being carried out modeling behind the context update.
140 pairs of described background models of described background coding module are encoded.
In one embodiment, a series of background images for the needs transmission can adopt I frame P frame mode that background image is encoded.At first, background image in the first two field picture that obtains transmits after encoding as the I frame, and the follow-up background that obtains is processed transmission as the P frame, in the I frame, find the macro block that each macro block mates most in the adjacent subsequent frame (being follow-up P frame) of time with it in the corresponding hunting zone, then it is poor to obtain one group of difference value vector the position of each corresponding in each macro block in the I frame and follow-up P frame macro block to be done, described difference value vector is all carried out DCT(Discrete Cosine Transform, be discrete cosine transform) conversion, and to the rear Vector Groups of order composition according to original macro block of its quantification, described Vector Groups is transmitted after encoding through entropy.General, because change of background is very little, so the code stream of above-mentioned P frame is very low, because in the static situation of video camera, common background image is stablized constant, and the method for above-mentioned definite coupling macro block can directly select the macro block of correspondence position to get final product with the macro block as the P frame in the I frame.
Certainly, the above-mentioned method of only having told about a kind of background coding, described background coding module 140 can also be encoded to background image by other method, its specific implementation can realize by the mode of software, hardware or software and hardware combining the coding to background model, the implementation method of wherein said dct transform and entropy coding belongs to all that those of ordinary skill in the field can both realize, and not as important technology scheme of the present invention, so just described in detail no longer one by one.
Described foreground segmentation module 150 is partitioned into foreground area in the described video image according to video image and the described background model of input.Described prospect is the moving target in the image, i.e. moving target in the image is with respect to background part more importantly in the video.
Because important with respect to the information of prospect, so can adopt the higher coded system of quality comes described prospect is encoded, and background is owing to be in for a long time constant state, in constant, only need the transmission primaries background to get final product, so the present invention selects prospect and background are separated coding and transmission.
Described foreground segmentation module 150 is in order to split prospect better from background, minimizing prospect and background edge place unsmooth, the method of foreground segmentation module 150 has a lot, and the method for existing extraction prospect is such as background subtraction method, mixed Gauss model method and Density Estimator method.But described method is not utilized the area information of image, thus so that the poor effect that the edge of the juncture area of prospect and background is cut apart.Here can adopt a better method, namely cut apart the method for (graph-cut) based on max-flow.
In one embodiment, the method for utilizing max-flow to cut apart is carried out foreground segmentation to image.It specifically can be referring to shown in Figure 2, and Fig. 2 is the flow chart of the foreground segmentation algorithm cut apart based on max-flow in one embodiment of the present invention.The described foreground segmentation algorithm of cutting apart based on max-flow comprises:
It is pointed out that when the context update in the video image and need to upgrade to guarantee that the background image that step 210 is obtained is latest image to background.
The method has not only utilized the difference between prospect and the background to cut apart prospect, and, also utilized the color model of prospect and background area, and the marginal information of prospect and background on the image, thereby can access better segmentation effect.And all pixels in the foreground image can be labeled as identical numeral according to said method, such as 1, and all pixels in the background image be labeled as identical numeral, such as 0, just obtained like this binary mask that can identify foreground image areas.That is to say that described binary mask is identical with the image size, the binary value of each location of pixels has identified whether pixel belongs to prospect in this image, if pixel belongs to prospect, then is 1 to value that should pixel on the mask then, otherwise is 0.
The prospect that described prospect coding module 160 goes out described foreground segmentation encodes to be used for transmission.For foreground area is carried out high-quality coding, the present invention encodes to form the prospect code stream to the binary mask of the effective macro block in the foreground area and identical foreground area.
Effective macro block acquisition methods of its described foreground area is specially: the binary mask of foreground area is obtained its minimum boundary rectangle, image-region in the described minimum boundary rectangle is divided into fixed-size macro block, if comprised foreground pixel in the macro block, think that then this macro block is effective, need coding, otherwise think that the pixel in this macro block is background pixel, invalid, do not encode; The above-mentioned macro block of coding that needs is formed a macro block vector in order, can adopt the method for zig-zag scanning to determine order.
Described macro block can be the square of N * N, and as being set as 8 * 8 or 16 * 16 macro block, the setting of described macroblock size is the convenience for follow-up DCT coding.
In one embodiment, the method that described prospect coding module is encoded to the effective macro block in the foreground area comprises: to effective foreground macro block of the first frame appearance, order according to macro block is processed successively, described macro block is carried out dct transform, and the DCT coefficient quantized, adopt the differential coding mode that the DCT coefficient is predicted, and then carry out the entropy coding; Each frame to follow-up appearance, judge whether each macro block is effective in the foreground area, effective macro block is carried out following processing in order: to current macro, in the hunting zone of former frame image, seek and its effective foreground macro block of mating most, it is poor that effective foreground macro block of the described coupling in the former frame image and described current macro are done, the difference that obtains is carried out dct transform, will carry out the entropy coding behind its DCT coefficient quantization.
The method that described prospect coding module is encoded to the binary mask of foreground area comprises: the boundary rectangle scope of foreground area number and each foreground area in the document image, whether also will record in the image-region in the boundary rectangle scope each macro block effective, be 1 if any this tick lables effectively then, otherwise be masked as 0, the sign employing entropy coded system of each macro block is encoded.
Described prospect coding module 160 forms code stream to be used for transmission with the data that effectively indicate on the coding result of macro block and the corresponding binary mask in the described foreground area.
Like this, avoided prospect at each video image to change and background when not changing to the transmission of background area, simultaneously, for foreground area, also utilize inter-frame information that redundant information is got rid of, reduced code check.
The image of the foreground area that the background image behind the coding that 170 pairs of described background coding modules 140 of described transport module obtain and prospect coding module 160 obtain transmits.Described transport module 170 comprises background image transport module 1701 and foreground image transport module 1702, background code stream behind the coding that 1701 pairs of background coding modules 140 of wherein said background image transport module obtain transmits, and the prospect code stream behind the coding that 1702 pairs of prospect coding modules 160 of described foreground image transport module obtain transmits.
In actual applications, 1702 pairs of background code streams of described background image transport module 1701 and foreground image transport module and prospect code stream transmit respectively.Because background image is in long-term in the constant situation, described background image transport module 1701 at this moment between the interior transmission primaries background code stream of section.
In sum, video encoder among the present invention separates background and prospect, and utilize the edge segmentation effect that has improved prospect and background juncture area based on the max-flow dividing method, and prospect carried out higher-quality coding, and nearly transmission primaries background when constant to background, thereby improved the compression quality of video, and reduced widely the video of transmission to the occupancy in broadband.
Fig. 3 is a kind of method for video coding flow chart in one embodiment among the present invention, sees also shown in Figure 3ly, and described method for video coding comprises:
Step 320 is set up background model according to described video image, when the background in the described video image changes, and update background module then.
Step 330 is according to video image and background model segmenting foreground image.
In order better prospect to be split from background, minimizing prospect and background edge place unsmooth, method to the foreground segmentation of image has a lot, and the method for existing extraction prospect is such as background subtraction method, mixed Gauss model method and Density Estimator method.But described method is not utilized the area information of image, thus so that the poor effect that the edge of the juncture area of prospect and background is cut apart.Here can adopt a better method, namely cut apart the method for (graph-cut) based on max-flow.The method that described max-flow is cut apart specifically can be referring to shown in Figure 2, and its content is described in detail in above-mentioned part, so just no longer described here.
Step 340 is encoded to background model and foreground image respectively.
In one embodiment, when background model was encoded, a series of background images for the needs transmission can adopt I frame P frame mode that background image is encoded.At first, background image in the first two field picture that obtains transmits after encoding as the I frame, and the follow-up background that obtains is processed transmission as the P frame, in the I frame, find the macro block that each macro block mates most in the adjacent subsequent frame (being follow-up P frame) of time with it in the corresponding hunting zone, then it is poor to obtain one group of difference value vector the position of each corresponding in each macro block in the I frame and follow-up P frame macro block to be done, described difference value vector is all carried out DCT(Discrete Cosine Transform, be discrete cosine transform) conversion, and to the rear Vector Groups of order composition according to original macro block of its quantification, described Vector Groups is transmitted after encoding through entropy.General, because change of background is very little, so the code stream of above-mentioned P frame is very low, because in the static situation of video camera, common background image is stablized constant, and the method for above-mentioned definite coupling macro block can directly select the macro block of correspondence position to get final product with the macro block as the P frame in the I frame.
Said process can be realized by the mode of software, hardware or software and hardware combining the coding to background model, the implementation method of wherein said dct transform and entropy coding belongs to all that those of ordinary skill in the field can both realize, and not as important technology scheme of the present invention, so just described in detail no longer one by one.
In another embodiment, when foreground area was encoded, for foreground area is carried out high-quality coding, the present invention encoded to form the prospect code stream to the binary mask of the effective macro block in the foreground area and identical foreground area.
Effective macro block acquisition methods of its described foreground area is specially: the binary mask of foreground area is obtained its minimum boundary rectangle, image-region in the described minimum boundary rectangle is divided into fixed-size macro block, if comprised foreground pixel in the macro block, think that then this macro block is effective, need coding, otherwise think that the pixel in this macro block is background pixel, invalid, do not encode; The above-mentioned macro block of coding that needs is formed a macro block vector in order, can adopt the method for zig-zag scanning to determine order.
Described macro block can be the square of N * N, and as being set as 8 * 8 or 16 * 16 macro block, the setting of described macroblock size is the convenience for follow-up DCT coding.
In one embodiment, the method that described prospect coding module is encoded to the effective macro block in the foreground area comprises: to effective foreground macro block of the first frame appearance, order according to macro block is processed successively, described macro block is carried out dct transform, and the DCT coefficient quantized, adopt the differential coding mode that the DCT coefficient is predicted, and then carry out the entropy coding; Each frame to follow-up appearance, judge whether each macro block is effective in the foreground area, effective macro block is carried out following processing in order: to current macro, in the hunting zone of former frame image, seek and its effective foreground macro block of mating most, it is poor that effective foreground macro block of the described coupling in the former frame image and described current macro are done, the difference that obtains is carried out dct transform, will carry out the entropy coding behind its DCT coefficient quantization.
The method that described prospect coding module is encoded to the binary mask of foreground area comprises: the boundary rectangle scope of foreground area number and each foreground area in the document image, whether also will record in the image-region in the boundary rectangle scope each macro block effective, be 1 if any this tick lables effectively then, otherwise be masked as 0, the sign employing entropy coded system of each macro block is encoded.
At last the data that effectively indicate on the coding result of macro block and the corresponding binary mask in the described foreground area are formed code stream to be used for transmission.
Like this, avoided prospect at each video image to change and background when not changing to the transmission of background area, simultaneously, for foreground area, also utilize inter-frame information that redundant information is got rid of, reduced code check.
The present invention also provides a kind of Video Decoder, and it will connect the code stream analyzing that searches out and become background code stream and prospect code stream, and respectively it is decoded and be reduced into original image.It specifically can be referring to shown in Figure 4.
Fig. 4 is a kind of Video Decoder structural representation in one embodiment among the present invention.Described Video Decoder 400 comprises parsing module 410, background decoder module 420, prospect decoder module 430 and video synthesis module 440.
Code stream after 410 pairs of code streams that receive of described parsing module are resolved and will be resolved is divided into background code stream and prospect code stream.
If described background decoder module 420 receives the background code stream after the parsing then it is decoded with background image updating, if do not receive the background code stream after the parsing then do not process.
In one embodiment, when receiving the I frame data, 420 of described background decoder modules adopt the decoder of I frame that it is decoded, and record.If what receive is the P frame data, then recovers residual image corresponding to its DCT coefficient, and from the I frame, obtain the coupling macroblock image, the reduction original video image.
Prospect code stream after 430 pairs of parsings that receive of described prospect decoder module is decoded.
Concrete, obtaining foreground features the prospect code stream of described prospect decoder module 430 after resolving, described foreground features comprises coordinate, macroblock coding information and the macro block effective marker information of the boundary rectangle of foreground area number, each foreground area.From macroblock coding information, recover respectively the information of each macro block.Send above-mentioned foreground area number, foreground area boundary rectangle coordinate, foreground area macroblock coding information, foreground area macro block validity flag and background image to video synthesis module 440.
Described video synthesis module 440 synthesizes to obtain original image with decoded background image and foreground image.
In actual applications, its synthetic being specially: setting up present image is background image; The foreground area of 440 pairs of existence of video synthesis module is processed respectively: arrive the foreground area upper left corner according to boundary rectangle coordinate setting, according to the order identical with coding, macro block in the boundary rectangle is carried out respectively following processing: obtain whether effective marker of foreground macro block corresponding to current macro, if macro block is invalid, then jump to and process next macro block, otherwise read the coded message of effective macro block, decoding obtains the DCT coefficient, the Cost fun ction image, from prior image frame, find its coupling macro block, recover the current macro image; The current macro image is covered correspondence position on the present image.
A kind of better method is to adopt the method for blending, and effective macro block background image that is added to is got on.Thereby so that the image that obtains true nature more can not left a trace in prospect and background intersection.
When mating with video encoder, adopt the Video Decoder Recovery image with as the coupling macroblock image, rather than adopt upper two field picture, thus the raising coding efficiency.
In sum, the present invention utilizes corresponding composition algorithm that prospect and background image are synthesized or utilizes the blending method that prospect is added on the image, thereby has avoided prospect and background intersection to leave a trace.
Fig. 5 is the method flow chart in one embodiment of a kind of video decode among the present invention, sees also shown in Figure 5ly, and described video decode comprises:
Step 510 is resolved with generation background code stream and prospect code stream the code stream that receives.
Step 520 is decoded to background code stream and prospect code stream respectively.
In concrete the application, if receive the background code stream after the parsing then it is decoded with background image updating, if do not receive the background code stream after the parsing then do not process.In one embodiment, when receiving the I frame data, then adopt the decoder of I frame that it is decoded, and record.If what receive is the P frame data, then recovers residual image corresponding to its DCT coefficient, and from the I frame, obtain the coupling macroblock image, the reduction original image.
In another embodiment, obtain the foreground area number the prospect code stream after resolving, and obtain coordinate, macroblock coding information and the macro block effective marker information of the boundary rectangle of each foreground area.From macroblock coding information, recover respectively the information of each macro block.Send above-mentioned foreground area number, foreground area boundary rectangle coordinate, foreground area macroblock coding information, foreground area macro block validity flag and background image to the video synthesis module.
Step 530 is synthesized decoded background image and foreground image.
In actual applications, its synthetic being specially: setting up present image is background image; The foreground area that exists is processed respectively: arrive the foreground area upper left corner according to boundary rectangle coordinate setting, according to the order identical with coding, macro block in the boundary rectangle is carried out respectively following processing: obtain whether effective marker of foreground macro block corresponding to current macro, if macro block is invalid, then jumps to the next macro block of processing, otherwise read the coded message of effective macro block, decoding obtains the DCT coefficient, the Cost fun ction image finds its coupling macro block from prior image frame, recover the current macro image; The current macro image is covered correspondence position on the present image.
A kind of better method is to adopt the method for blending, and effective macro block background image that is added to is got on.Thereby so that the image that obtains true nature more can not left a trace in prospect and background intersection.
In sum, the present invention utilizes corresponding composition algorithm that prospect and background image are synthesized or utilizes the blending method that prospect is added on the image, thereby has avoided prospect and background intersection to leave a trace.
Above-mentioned explanation has fully disclosed the specific embodiment of the present invention.It is pointed out that and be familiar with the scope that any change that the person skilled in art does the specific embodiment of the present invention does not all break away from claims of the present invention.Correspondingly, the scope of claim of the present invention also is not limited only to previous embodiment.
Claims (11)
1. video encoder is characterized in that it comprises:
Video input module, inputted video image;
The background in the new video image more when context update module, change of background in described video image;
The background modeling module is set up background model according to the background in the described video image that upgrades;
The background coding module is encoded with the background extraction code stream to described background model;
The foreground segmentation module is partitioned into foreground area in the described video image according to video image and the described background model of input;
The prospect coding module encodes to obtain the prospect code stream to the foreground area that obtains; With
Transport module transmits respectively described background code stream and prospect code stream,
Wherein said prospect coding module encodes to form the prospect code stream to the binary mask of the effective macro block in the foreground area and identical foreground area, wherein said binary mask is used for which position of sign video image and belongs to foreground area, each pixel that belongs to prospect is marked as 1 on the mask, otherwise is labeled as 0; Effective macro block acquisition methods in the described foreground area is: obtain its minimum boundary rectangle according to the binary mask of foreground area; image-region in the described boundary rectangle is divided into foursquare macro block; if comprise foreground pixel in the macro block then judge that this macro block is effective macro block; need coding; otherwise be judged to be invalid macro block; do not encode
Described prospect coding module is encoded to the binary mask of foreground area and is comprised:
The boundary rectangle scope of foreground area number and each foreground area in the document image, whether also will record in the image-region in the boundary rectangle scope each macro block effective, be masked as 1 such as this macro block effectively then, otherwise be masked as 0, adopt the entropy coded system to encode the sign of each macro block, and form bit stream with the macroblock coding result.
2. video encoder according to claim 1, it is characterized in that: described transport module comprises background image transport module and foreground image transport module, described background image transport module is used for the background code stream after the transfer encoding, and described foreground image transport module is used for the prospect code stream after the transfer encoding.
3. video encoder according to claim 1, it is characterized in that: the method that described foreground segmentation module is cut apart based on max-flow is partitioned into the foreground area in the described video image, and the method that described max-flow is cut apart comprises:
Obtain the Gauss model of the color component of each background pixel in the serial background image;
Pixel in the video image and background image are carried out pixel value difference, to described difference greater than the pixel of threshold value then as foreground pixel;
Obtain the mixed Gauss model of the color component of described foreground pixel;
Adopt the described mixed Gauss model of the described Gauss model of background and prospect to calculate the probability that all pixels in the video image belong to prospect or background, obtain this pixel corresponding node after the normalization and be connected with the background node with prospect respectively in the drawings and be connected energy;
Described video image is asked respectively horizontal edge intensity and vertical edge intensity between its neighbor;
Being connected horizontal edge intensity with neighbor according to each pixel is connected with vertical edge intensity at the energy that is connected between pixel described in the video image and neighbor; With
According to all internodal connection energy that obtain, adopt the max-flow dividing method to carry out smallest partition, obtain prospect and background pixel in the described video image.
4. video encoder according to claim 1, it is characterized in that: described prospect coding module is encoded to the effective macro block in the foreground area and is comprised:
Effective foreground macro block to the appearance of the first frame, order according to macro block is processed successively, and described macro block is carried out discrete cosine transform, and discrete cosine transform coefficient is quantized, adopt the differential coding mode that discrete cosine transform coefficient is predicted, and then carry out the entropy coding;
Each frame to follow-up appearance, judge whether each macro block is effective in the foreground area, effective macro block is carried out following processing in order: to current macro, in the hunting zone of former frame image, seek and its effective foreground macro block of mating most, it is poor that effective foreground macro block of the described coupling in the former frame image and described current macro are done, the difference that obtains is carried out discrete cosine transform, carry out the entropy coding after its discrete cosine transform coefficient is quantized.
5. method for video coding is characterized in that it comprises:
Inputted video image;
When the change of background of described video image, upgrade its background;
The background of described renewal is set up background model and described background model is encoded to obtain the background code stream;
Be partitioned into the foreground area in the described video image and described foreground area encoded to obtain the prospect code stream according to described video image and described background model; With
Transmit respectively described background code stream and prospect code stream,
Wherein the binary mask of the effective macro block in the foreground area and identical foreground area is encoded to obtain the prospect code stream, wherein said binary mask is used for which position of sign video image and belongs to foreground area, each pixel that belongs to prospect is marked as 1 on the mask, otherwise is labeled as 0; Effective macro block acquisition methods in the described foreground area is: obtain its minimum boundary rectangle according to the binary mask of foreground area; image-region in the described boundary rectangle is divided into foursquare macro block; if comprise foreground pixel in the macro block then judge that this macro block is effective macro block; need coding; otherwise be judged to be invalid macro block; do not encode
The binary mask of identical foreground area encoded comprises:
The boundary rectangle scope of foreground area number and each foreground area in the document image, whether also will record in the image-region in the boundary rectangle scope each macro block effective, be masked as 1 such as this macro block effectively then, otherwise be masked as 0, adopt the entropy coded system to encode the sign of each macro block, the coding result of effective macro block in the foreground area and the coding result of the binary mask of the foreground area of correspondence are formed code stream to transmit.
6. method for video coding according to claim 5, it is characterized in that: utilize the max-flow dividing method that the foreground area in the described video image is cut apart, it comprises:
Obtain the Gauss model of the color component of each background pixel in the serial background image;
Pixel in the video image and background image are carried out pixel value difference, to described difference greater than the pixel of threshold value then as foreground pixel;
Obtain the mixed Gauss model of the color component of described foreground pixel;
Adopt the described mixed Gauss model of the described Gauss model of background and prospect to calculate the probability that all pixels in the video image belong to prospect or background, obtain this pixel corresponding node after the normalization and be connected with the background node with prospect respectively in the drawings and be connected energy;
Described video image is asked respectively horizontal edge intensity and vertical edge intensity between its neighbor;
Being connected horizontal edge intensity with neighbor according to each pixel is connected with vertical edge intensity at the energy that is connected between pixel described in the video image and neighbor; With
According to all internodal connection energy that obtain, adopt the max-flow dividing method to carry out smallest partition, obtain prospect and background pixel in the described video image.
7. method for video coding according to claim 5, it is characterized in that: the effective macro block in the foreground area is encoded to be comprised:
Effective foreground macro block to the appearance of the first frame, order according to macro block is processed successively, and described macro block is carried out discrete cosine transform, and discrete cosine transform coefficient is quantized, adopt the differential coding mode that discrete cosine transform coefficient is predicted, and then carry out the entropy coding;
Each frame to follow-up appearance, judge whether each macro block is effective in the foreground area, effective macro block is carried out following processing in order: to current macro, in the hunting zone of former frame image, seek and its effective foreground macro block of mating most, it is poor that effective foreground macro block of the described coupling in the former frame image and described current macro are done, the difference that obtains is carried out discrete cosine transform, carry out the entropy coding after its discrete cosine transform coefficient is quantized.
8. Video Decoder is characterized in that it comprises:
Parser modules is resolved the code stream that receives, and the background code stream after will resolving and prospect code stream are separately;
The background decoder module if receive the background code stream after the parsing then it is decoded with background image updating, otherwise does not process;
The prospect decoder module obtains the foreground area number the prospect code stream after resolving, and obtains coordinate, macroblock coding information and the macro block effective marker information of the boundary rectangle of each foreground area; With
The video synthesis module, obtain the background image of present image, arrive the upper left corner of foreground area according to the boundary rectangle coordinate setting of each foreground area, according to the order identical with coding, judge whether the foreground macro block that macro block is corresponding in the boundary rectangle is effective macro block, if be invalid macro block then jump to next macro block and process, otherwise read the coded message of effective macro block, obtain discrete cosine transform coefficient after the decoding, the Cost fun ction image, from last two field picture, find its coupling macro block, recover the current macro image, the current macro image is covered correspondence position on the present image; Or effective macro block is added on the background image.
9. Video Decoder according to claim 8, it is characterized in that: described former frame image is the former frame image after the video decoder decodes.
10. video encoding/decoding method is characterized in that it comprises:
The code stream that receives is resolved, and the background code stream after will resolving and prospect code stream are separately;
If receive the background code stream after the parsing then it is decoded with background image updating, otherwise do not process;
Obtain the foreground area number the prospect code stream after resolving, and obtain coordinate, macroblock coding information and the macro block effective marker information of the boundary rectangle of each foreground area; With
Obtain the background image of present image, arrive the upper left corner of foreground area according to the boundary rectangle coordinate setting of each foreground area, according to the order identical with coding, judge whether the foreground macro block that macro block is corresponding in the boundary rectangle is effective macro block, if be invalid macro block then jump to next macro block and process, otherwise read the coded message of effective macro block, obtain discrete cosine transform coefficient after the decoding, the Cost fun ction image, from last two field picture, find its coupling macro block, recover the current macro image, and the current macro image is covered correspondence position on the present image; Or effective macro block is added on the background image.
11. video encoding/decoding method according to claim 10 is characterized in that: described former frame image is the former frame image after the video decoder decodes.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201010549778 CN102006473B (en) | 2010-11-18 | 2010-11-18 | Video encoder and encoding method, and video decoder and decoding method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201010549778 CN102006473B (en) | 2010-11-18 | 2010-11-18 | Video encoder and encoding method, and video decoder and decoding method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102006473A CN102006473A (en) | 2011-04-06 |
CN102006473B true CN102006473B (en) | 2013-03-13 |
Family
ID=43813490
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 201010549778 Active CN102006473B (en) | 2010-11-18 | 2010-11-18 | Video encoder and encoding method, and video decoder and decoding method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102006473B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9578224B2 (en) | 2012-09-10 | 2017-02-21 | Nvidia Corporation | System and method for enhanced monoimaging |
Families Citing this family (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103108197A (en) | 2011-11-14 | 2013-05-15 | 辉达公司 | Priority level compression method and priority level compression system for three-dimensional (3D) video wireless display |
CN103187083B (en) * | 2011-12-29 | 2016-04-13 | 深圳中兴力维技术有限公司 | A kind of storage means based on time domain video fusion and system thereof |
AU2012200319B2 (en) * | 2012-01-19 | 2015-11-26 | Canon Kabushiki Kaisha | Method, apparatus and system for encoding and decoding the significance map for residual coefficients of a transform unit |
US9829715B2 (en) | 2012-01-23 | 2017-11-28 | Nvidia Corporation | Eyewear device for transmitting signal and communication method thereof |
CN103634554B (en) * | 2012-08-20 | 2017-06-27 | 联想(北京)有限公司 | A kind of method of data transfer, the method for data receiver and electronic equipment |
CN103179402A (en) * | 2013-03-19 | 2013-06-26 | 中国科学院半导体研究所 | Video compression coding and decoding method and device |
CN103313061B (en) * | 2013-06-28 | 2016-09-07 | 汕头大学 | The video compress transmission method of a kind of adaptive RTS threshold adjustment and system |
CN103475882B (en) * | 2013-09-13 | 2017-02-15 | 北京大学 | Surveillance video encoding and recognizing method and surveillance video encoding and recognizing system |
US10935788B2 (en) | 2014-01-24 | 2021-03-02 | Nvidia Corporation | Hybrid virtual 3D rendering approach to stereovision |
CN103986882A (en) * | 2014-05-21 | 2014-08-13 | 福建歌航电子信息科技有限公司 | Method for image classification, transmission and processing in real-time monitoring system |
CN104243994B (en) * | 2014-09-26 | 2017-06-27 | 厦门亿联网络技术股份有限公司 | A kind of enhanced method of real time kinematics perceptual image |
CN105992010B (en) * | 2015-02-13 | 2020-02-18 | 中星技术股份有限公司 | Binary foreground compression coding method |
CN104702956B (en) * | 2015-03-24 | 2017-07-11 | 武汉大学 | A kind of background modeling method towards Video coding |
CN106210612A (en) * | 2015-04-30 | 2016-12-07 | 杭州海康威视数字技术股份有限公司 | Method for video coding, coding/decoding method and device thereof |
CN106331700B (en) | 2015-07-03 | 2019-07-19 | 华为技术有限公司 | Method, encoding device and the decoding device of reference picture coding and decoding |
CN106454348B (en) * | 2015-08-05 | 2019-08-13 | 中国移动通信集团公司 | A kind of video coding-decoding method and device |
CN105049818A (en) * | 2015-08-25 | 2015-11-11 | 北京丰华联合科技有限公司 | Method for optimizing video data transmission |
CN105451021B (en) * | 2015-10-05 | 2019-06-25 | 上海大学 | A kind of low-consumption wireless micro display system |
US9906981B2 (en) | 2016-02-25 | 2018-02-27 | Nvidia Corporation | Method and system for dynamic regulation and control of Wi-Fi scans |
CN105657429A (en) * | 2016-03-09 | 2016-06-08 | 成都爆米花信息技术有限公司 | Video compression method |
CN108337515B (en) * | 2018-01-19 | 2020-08-04 | 浙江大华技术股份有限公司 | Video coding method and device |
CN109246433B (en) * | 2018-09-26 | 2020-09-11 | 北京红云融通技术有限公司 | Video encoding method and device, decoding method and device and video transmission system |
CN109495749A (en) * | 2018-12-24 | 2019-03-19 | 上海国茂数字技术有限公司 | A kind of coding and decoding video, search method and device |
CN110677688A (en) * | 2019-10-31 | 2020-01-10 | 重庆唐码传媒有限公司 | Streaming media image pushing system |
CN111526417B (en) * | 2020-04-20 | 2021-09-17 | 河北达实信息技术有限公司 | Video image transmission method and transmission system |
CN113630609B (en) * | 2020-05-06 | 2024-03-12 | Tcl科技集团股份有限公司 | Video encoding method, decoding method, storage medium and terminal equipment |
CN111800653B (en) * | 2020-07-29 | 2021-06-11 | 北京大学深圳研究生院 | Video decoding method, system, device and computer readable storage medium |
CN116634178B (en) * | 2023-07-26 | 2023-10-31 | 清华大学 | Security scene monitoring video coding and decoding method and system with extremely low code rate |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1713729A (en) * | 2004-06-24 | 2005-12-28 | 华为技术有限公司 | Video frequency compression |
CN1980334A (en) * | 2005-12-09 | 2007-06-13 | 英业达股份有限公司 | Video-picture composition code transmission method |
CN101510304A (en) * | 2009-03-30 | 2009-08-19 | 北京中星微电子有限公司 | Method, device and pick-up head for dividing and obtaining foreground image |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7110596B2 (en) * | 2002-04-25 | 2006-09-19 | Microsoft Corporation | System and method facilitating document image compression utilizing a mask |
US7425981B2 (en) * | 2002-06-12 | 2008-09-16 | British Telecommunications Plc | Video pre-processing |
-
2010
- 2010-11-18 CN CN 201010549778 patent/CN102006473B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1713729A (en) * | 2004-06-24 | 2005-12-28 | 华为技术有限公司 | Video frequency compression |
CN1980334A (en) * | 2005-12-09 | 2007-06-13 | 英业达股份有限公司 | Video-picture composition code transmission method |
CN101510304A (en) * | 2009-03-30 | 2009-08-19 | 北京中星微电子有限公司 | Method, device and pick-up head for dividing and obtaining foreground image |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9578224B2 (en) | 2012-09-10 | 2017-02-21 | Nvidia Corporation | System and method for enhanced monoimaging |
Also Published As
Publication number | Publication date |
---|---|
CN102006473A (en) | 2011-04-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102006473B (en) | Video encoder and encoding method, and video decoder and decoding method | |
CN102006475B (en) | Video coding and decoding device and method | |
CN102067608B (en) | Fragmented reference in temporal compression for video coding | |
CN101981934B (en) | Methods, devices and systems for parallel video encoding and decoding | |
CN103475882B (en) | Surveillance video encoding and recognizing method and surveillance video encoding and recognizing system | |
CN102630043B (en) | Object-based video transcoding method and device | |
CN101193261B (en) | Video communication system and method | |
CN103152578B (en) | H.264 video watermark based on mixed encoding and decoding embeds and extracting method | |
CN103636215A (en) | Applying non-square transforms to video data | |
CN101924943B (en) | Real-time low-bit rate video transcoding method based on H.264 | |
EP2526699A1 (en) | Data pruning for video compression using example-based super-resolution | |
CN102450010A (en) | Directed interpolation and data post-processing | |
CN104754361A (en) | Image encoding and decoding method and device | |
CN103281527A (en) | Method for coding/decoding image and relevant device | |
CN101291436B (en) | Video coding/decoding method and device thereof | |
CN104159117A (en) | Method for hiding HEVC video information | |
CN106101714A (en) | One and the tightly coupled H.264 video information hiding method of compression encoding process | |
CN1926879A (en) | A video signal encoder, a video signal processor, a video signal distribution system and methods of operation therefor | |
CN101313582A (en) | Encoder assisted frame rate up conversion using various motion models | |
CN105681803A (en) | High capacity HEVC video information hiding method | |
CN1774931A (en) | Content analysis of coded video data | |
CN102823241A (en) | Method and apparatus for low bandwidth content preserving compression of stereoscopic three dimensional images | |
CN104104948B (en) | Video transcoding method and video code translator | |
CN103888770A (en) | Efficient self-adaptive video transcoding system based on data mining | |
CN102984525A (en) | Method for hiding video code stream error |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
C56 | Change in the name or address of the patentee | ||
CP01 | Change in the name or title of a patent holder |
Address after: 214028 Jiangsu New District of Wuxi, Taihu international science and Technology Park Jia Qing 530 building 10 layer Patentee after: WUXI ZHONGGAN MICROELECTRONIC CO., LTD. Address before: 214028 Jiangsu New District of Wuxi, Taihu international science and Technology Park Jia Qing 530 building 10 layer Patentee before: Wuxi Vimicro Co., Ltd. |