CN103716622A - Image processing method and device - Google Patents

Image processing method and device Download PDF

Info

Publication number
CN103716622A
CN103716622A CN201210374978.5A CN201210374978A CN103716622A CN 103716622 A CN103716622 A CN 103716622A CN 201210374978 A CN201210374978 A CN 201210374978A CN 103716622 A CN103716622 A CN 103716622A
Authority
CN
China
Prior art keywords
filter
image
training
target
sampling
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.)
Granted
Application number
CN201210374978.5A
Other languages
Chinese (zh)
Other versions
CN103716622B (en
Inventor
杨海涛
杨名远
张金雷
李斌
李厚强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Science and Technology of China USTC
Huawei Technologies Co Ltd
Original Assignee
University of Science and Technology of China USTC
Huawei Technologies Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by University of Science and Technology of China USTC, Huawei Technologies Co Ltd filed Critical University of Science and Technology of China USTC
Priority to CN201210374978.5A priority Critical patent/CN103716622B/en
Priority to PCT/CN2013/084531 priority patent/WO2014048374A1/en
Publication of CN103716622A publication Critical patent/CN103716622A/en
Application granted granted Critical
Publication of CN103716622B publication Critical patent/CN103716622B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/30—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using hierarchical techniques, e.g. scalability
    • H04N19/33—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using hierarchical techniques, e.g. scalability in the spatial domain
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
    • H04N19/117—Filters, e.g. for pre-processing or post-processing
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
    • H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
    • H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)
  • Image Processing (AREA)

Abstract

Disclosed is an image processing method which includes: determining a first training filter according to a non-smooth-area image block of an reinforced-layer image and a non-smooth-area image block of a basic-layer image so as to make the first training filter satisfies: the similarity between first prediction information, which is determined according to the first training filter and the non-smooth-area image block of the basic-layer image, and first original information, which is determined according to the non-smooth-area image block of the reinforce-layer image meets a first preset condition, wherein the basic-layer image is corresponding to the reinforced-layer image; determining alternative upsampling filters according to the first training filter, wherein the alternative upsampling filters include the first training filter; determining a target upsampling filter from the alternative upsampling filters; determining prediction information according to the target upsampling filter and the basic-layer image block; and performing coding processing on a target image block according to the prediction information so as to generate a target code stream.

Description

The method and apparatus of processing for image
Technical field
The present invention relates to field of video processing, and more specifically, relate to a kind of method and apparatus of processing for image of processing for image.
Background technology
Along with the fast development of the Internet and becoming increasingly abundant of people's material spirit culture, application demand for video in the Internet is especially more and more for the application demand of HD video, and the data volume of HD video is very large, wanting HD video can transmit in band-limited the Internet, and the problem that must first solve is exactly HD video compressed encoding problem.
In network environment (such as the Internet), because the network bandwidth is limited, terminal equipment and user's demand is all different, so for certain specific application, the code stream of first compression is not satisfactory and effective, for some specific user or equipment, or even nonsensical.The effective method addressing this problem utilizes scalable video (SVC, scalable video coding) technology exactly.In this SVC technology, the mass parameter according to comprising spatial resolution, temporal resolution or signal to noise ratio intensity etc., is divided into a plurality of image layer by an image.The target of SVC is exactly to allow the high image layer of quality utilize fully the information of low-quality image layer as far as possible, improves the efficiency of inter-layer prediction, can efficiency when making the high image of coding quality higher.
When the high image of quality is carried out to hierarchical coding, because resolution is different, need to carry out up-sampling to low-quality image, to reach after the consistent resolution of the image high with quality, the information that recycles low-quality image is predicted enhancement layer.Therefore, the effect of up-sampling can produce directly impact to coding efficiency, and the difference due to the characteristics such as texture of image, by using a plurality of filters, can effectively promote the effect of up-sampling.
At present, known a kind of technology of using a plurality of filters to carry out up-sampling, its to each image block calculating location of 4 * 41,2,3,4 the horizontal direction of four pixels and the second differnce of vertical direction, the second differnce of four pixel level directions is added, as level (horizontal), the second differnce of four pixel vertical directions is added, as vertical (vertical).Thereby, can determine grain direction (direction).For example, if vertical is greater than the twice of horizontal, determine that direction is 1; If horizontal is greater than the twice of vertical, determine that direction is 2; If vertical equals the twice of horizontal, determine that direction is 0.Thereafter, calculate the mean value of vertical and horizontal, to determine type under this block of pixels, and train respectively a filter for each type, like this can be so that training gained filter conforms to corresponding classification very much, but the index (index) of the filter that the method need to be used the coefficient of filter and every class piece at coding side is encoded and is transferred to decoding end, so just can be correctly decoded in decoding end.If make in this way basic layer to be rebuild to image, carry out up-sampling, can cause the expression filter coefficient of needs transmission and the bit of index too many, thereby affect coding efficiency.
Summary of the invention
The embodiment of the present invention provides a kind of method and apparatus of processing for image, can improve effect and performance that image is processed.
First aspect, a kind of method of processing for image is provided, the method comprises: according to the non-flat skating area area image piece of the non-flat skating area area image piece of enhancement layer image and basic tomographic image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image, according to this first training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, from this alternative up-sampling filter, determine target up-sampling filter, according to this target up-sampling filter and basic tomographic image piece, determine information of forecasting, according to this information of forecasting, to the processing of encoding of target image piece, to generate target code stream, wherein, this target image piece is arranged in this enhancement layer image, this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image.
In a kind of possible execution mode, the method also comprises: according to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image; And this is according to this first training filter, determine alternative up-sampling filter, further comprise: according to this first training filter and this second training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
In conjunction with first aspect and the possible execution mode of the first, in the possible execution mode of the second, should be from alternative up-sampling filter, determine target up-sampling filter, comprise: according to the characteristic information of this basic tomographic image piece, determine the smoothness of this target image, wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece; According to the smoothness of this target image piece, determine this target up-sampling filter.
In conjunction with first aspect, execution mode and the possible execution mode of the second that the first is possible, in the third possible execution mode, according to this first training filter and this second training filter, determine alternative up-sampling filter, further comprise: according to this first training filter, this second training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, this second training filter and this conventional filter; And this is according to this information of forecasting, to the processing of encoding of this target image piece, to generate target code stream, comprise: according to this information of forecasting, to the processing of encoding of this target image piece, to generate target code stream, this target code stream comprises the first indication information that is used to indicate this target up-sampling filter, this first indication information during in order to target image piece after the above-mentioned coding of decoding as the foundation of obtaining this target up-sampling filter.
In conjunction with first aspect, possible execution mode and the third the possible execution mode of execution mode, the second that the first is possible, in the 4th kind of possible execution mode, this is according to this first training filter, determine alternative up-sampling filter, further comprise: according to this first training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this conventional filter; And this determines target up-sampling filter, comprising: the texture degree of determining this target image piece; According to the texture degree of this target image piece, determine this target up-sampling filter.
Second aspect, a kind of method of processing for image is provided, the method comprises: according to the non-flat skating area area image piece of the non-flat skating area area image piece of enhancement layer image and basic tomographic image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image, according to this first training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, from alternative up-sampling filter, determine target up-sampling filter, this alternative up-sampling filter comprises this first training filter, according to this target up-sampling filter and basic tomographic image piece, determine information of forecasting, according to this information of forecasting and the residual information that obtains from target code stream, to the processing of decoding of this object code stream, to obtain this target image piece, wherein, this target image piece is arranged in this enhancement layer image, this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image.
In a kind of possible execution mode, the method also comprises: according to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image; And this is according to this first training filter, determine alternative up-sampling filter, further comprise: according to this first training filter and this second training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
In conjunction with second aspect and the possible execution mode of the first, in the possible execution mode of the second, should be from alternative up-sampling filter, should be from alternative up-sampling filter, determine target up-sampling filter, comprising: according to the characteristic information of this basic tomographic image piece, determine the smoothness of this target image, wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece; According to the smoothness of this target image piece, determine this target up-sampling filter.
In conjunction with second aspect, execution mode and the possible execution mode of the second that the first is possible, in the third possible execution mode, according to this first training filter and this second training filter, determine alternative up-sampling filter, further comprise: according to this first training filter, this second training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, this second training filter and this conventional filter; And should from alternative up-sampling filter, determine target up-sampling filter, comprising: from this target code stream, obtain the first indication information that is used to indicate this target up-sampling filter, according to this first indication information, determine this target up-sampling filter.
In conjunction with second aspect, possible execution mode and the third the possible execution mode of execution mode, the second that the first is possible, in the 4th kind of possible execution mode, this is according to this first training filter, determine alternative up-sampling filter, further comprise: according to this first training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this conventional filter; And this determines target up-sampling filter, comprising: the texture degree of determining this target image piece; According to the texture degree of this target image piece, determine this target up-sampling filter.
The third aspect, a kind of device of processing for image is provided, this device comprises: acquiring unit, for the non-flat skating area area image piece with basic tomographic image according to the non-flat skating area area image piece of enhancement layer image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image, determining unit, according to this first training filter, determines alternative up-sampling filter, and wherein, this alternative up-sampling filter comprises this first training filter, for from this alternative up-sampling filter, determine target up-sampling filter, coding unit, for obtaining this up-sampling filter from this determining unit, and according to this target up-sampling filter and basic tomographic image piece, determines information of forecasting, be used for according to this information of forecasting, to the processing of encoding of target image piece, to generate target code stream, wherein, this target image piece is arranged in this enhancement layer image, this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image.
In a kind of possible execution mode, this acquiring unit is also for according to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image; And this determining unit is further used for, according to this first training filter and this second training filter, determining alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
In conjunction with the third aspect and the possible execution mode of the first, in the possible execution mode of the second, this determining unit is specifically for according to the characteristic information of this basic tomographic image piece, determine the smoothness of this target image, wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece; For according to the smoothness of this target image piece, determine this target up-sampling filter.
In conjunction with the third aspect, execution mode and the possible execution mode of the second that the first is possible, in the third possible execution mode, this determining unit is further used for according to this first training filter, this second training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, this second training filter and this conventional filter; And this coding unit is specifically for according to this information of forecasting, to the processing of encoding of this target image piece, to generate target code stream, this target code stream comprises the first indication information that is used to indicate this target up-sampling filter, this first indication information in order in decoding during the target image piece after above-mentioned coding as the foundation of obtaining this target up-sampling filter.
In conjunction with the third aspect, possible execution mode and the third the possible execution mode of execution mode, the second that the first is possible, in the 4th kind of possible execution mode, this determining unit is further used for according to this first training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this conventional filter; Specifically for determining the texture degree of this target image piece; For according to the texture degree of this target image piece, determine this target up-sampling filter.
Fourth aspect, a kind of device of processing for image is provided, it is characterized in that, this device comprises: acquiring unit, for the non-flat skating area area image piece with basic tomographic image according to the non-flat skating area area image piece of enhancement layer image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image, determining unit, for according to this first training filter, determines alternative up-sampling filter, and wherein, this alternative up-sampling filter comprises this first training filter, for from alternative up-sampling filter, determine target up-sampling filter, this alternative up-sampling filter comprises this first training filter, decoding unit, for obtain this target up-sampling filter from this determining unit, and according to this target up-sampling filter and basic tomographic image piece, determines information of forecasting, for according to this information of forecasting and the residual information that obtains from target code stream, to the processing of decoding of this object code stream, to obtain this target image piece, wherein, this target image piece is arranged in this enhancement layer image, this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image.
In a kind of possible execution mode, this acquiring unit is also for according to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image; And this determining unit is further used for, according to this first training filter and this second training filter, determining alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
In conjunction with fourth aspect and the possible execution mode of the first, in the possible execution mode of the second, this determining unit is specifically for according to the characteristic information of this basic tomographic image piece, determine the smoothness of this target image, wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece; According to the smoothness of this target image piece, determine this target up-sampling filter.
In conjunction with fourth aspect, execution mode and the possible execution mode of the second that the first is possible, in the third possible execution mode, this determining unit is further used for according to this first training filter, this second training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, this second training filter and this conventional filter; For from this target code stream, obtain the first indication information that is used to indicate this target up-sampling filter, for according to this first indication information, determine this target up-sampling filter.
In conjunction with fourth aspect, possible execution mode and the third the possible execution mode of execution mode, the second that the first is possible, in the 4th kind of possible execution mode, this determining unit is further used for according to this first training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this conventional filter; For determining the texture degree of this target image piece; For according to the texture degree of this target image piece, determine this target up-sampling filter.
The 5th aspect, provides a kind of encoder of processing for image, and this encoder comprises: bus, the processor being connected with this bus, the memory being connected with this bus, wherein, this processor is by this bus, call the program of storing in this memory, for the non-flat skating area area image piece with basic tomographic image according to the non-flat skating area area image piece of enhancement layer image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image, according to this first training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, from this alternative up-sampling filter, determine target up-sampling filter, according to this target up-sampling filter and basic tomographic image piece, determine information of forecasting, according to this information of forecasting, to the processing of encoding of target image piece, to generate target code stream, wherein, this target image piece is arranged in this enhancement layer image, this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image.
In a kind of possible execution mode, this processing unit is also for according to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image; For according to this first training filter and this second training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
In conjunction with the 5th aspect and the possible execution mode of the first, in the possible execution mode of the second, this processing unit is specifically for according to the characteristic information of this basic tomographic image piece, determine the smoothness of this target image, wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece; For according to the smoothness of this target image piece, determine this target up-sampling filter.
In conjunction with the 5th aspect, execution mode and the possible execution mode of the second that the first is possible, in the third possible execution mode, this processor is specifically for according to this first training filter, this second training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, this second training filter and this conventional filter; Be used for according to this information of forecasting, to the processing of encoding of this target image piece, to generate target code stream, this target code stream comprises the first indication information that is used to indicate this target up-sampling filter, this first indication information in order in decoding during the target image piece after above-mentioned coding as obtaining complying with of this target up-sampling filter
In conjunction with the 5th aspect, possible execution mode and the third the possible execution mode of execution mode, the second that the first is possible, in the 4th kind of possible execution mode, this processor is specifically for according to this first training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this conventional filter; Specifically for determining the texture degree of this target image piece; For according to the texture degree of this target image piece, determine this target up-sampling filter.。
The 6th aspect, provides a kind of decoder of processing for image, and this decoder comprises: bus, the processor being connected with this bus, the memory being connected with this bus, wherein, this processor is by this bus, call the program of storing in this memory, for the non-flat skating area area image piece with basic tomographic image according to the non-flat skating area area image piece of enhancement layer image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image, according to this first training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, from alternative up-sampling filter, determine target up-sampling filter, this alternative up-sampling filter comprises this first training filter, according to this target up-sampling filter and basic tomographic image piece, determine information of forecasting, according to this information of forecasting and the residual information that obtains from target code stream, to the processing of decoding of this object code stream, to obtain this target image piece, wherein, this target image piece is arranged in this enhancement layer image, this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image.
In a kind of possible execution mode, this processing unit is also for according to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image; For according to this first training filter and this second training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
In conjunction with the 6th aspect and the possible execution mode of the first, in the possible execution mode of the second, this processing unit is specifically for according to the characteristic information of this basic tomographic image piece, determine the smoothness of this target image, wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece; According to the smoothness of this target image piece, determine this target up-sampling filter.
In conjunction with the 6th aspect, execution mode and the possible execution mode of the second that the first is possible, in the third possible execution mode, this processor is further used for according to this first training filter, this second training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, this second training filter and this conventional filter; For from this target code stream, obtain the first indication information that is used to indicate this target up-sampling filter, for according to this first indication information, determine this target up-sampling filter.
In conjunction with the 6th aspect, possible execution mode and the third the possible execution mode of execution mode, the second that the first is possible, in the 4th kind of possible execution mode, this processor is further used for according to this first training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this conventional filter; For determining the texture degree of this target image piece; For according to the texture degree of this target image piece, determine this target up-sampling filter.
According to method, device code device and the decoder processed for image of the embodiment of the present invention, by according to smooth region and non-smooth region, determine up-sampling filter, can be when improving the effect of up-sampling, reduce the quantity of up-sampling filter and without transmission filter coefficient and index, thereby can improve coding efficiency, improve effect and performance that image is processed.
Accompanying drawing explanation
In order to be illustrated more clearly in the technical scheme of the embodiment of the present invention, to the accompanying drawing of required use in the embodiment of the present invention be briefly described below, apparently, below described accompanying drawing be only some embodiments of the present invention, for those of ordinary skills, do not paying under the prerequisite of creative work, can also obtain according to these accompanying drawings other accompanying drawing.
Fig. 1 is the indicative flowchart of the method processed for image according to an embodiment of the invention.
Fig. 2 is according to an embodiment of the invention for determining the schematic diagram of the method for training filter.
Fig. 3 is another indicative flowchart of the method processed for image according to an embodiment of the invention.
Fig. 4 is the schematic block diagram of the device processed for image according to an embodiment of the invention.
Fig. 5 is the schematic block diagram of the device processed for image according to another embodiment of the present invention.
Fig. 6 is the schematic block diagram of the encoder processed for image according to an embodiment of the invention.
Fig. 7 is the schematic block diagram of the decoder processed for image according to another embodiment of the present invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is the present invention's part embodiment, rather than whole embodiment.Embodiment based in the present invention, those of ordinary skills, not making the every other embodiment obtaining under creative work prerequisite, belong to the scope of protection of the invention.
Fig. 1 show from coding side angle, describe according to the indicative flowchart of the method 100 of processing for image of the embodiment of the present invention.As shown in Figure 1, the method 100 comprises:
S110, according to the non-flat skating area area image piece of the non-flat skating area area image piece of enhancement layer image and basic tomographic image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image;
S120, according to this first training filter, determines alternative up-sampling filter, and wherein, this alternative up-sampling filter comprises this first training filter;
S130, from this alternative up-sampling filter, determines target up-sampling filter;
S140, according to this target up-sampling filter and basic tomographic image piece, determines information of forecasting.
S150, according to this information of forecasting, to the processing of encoding of target image piece, to generate target code stream, wherein, this target image piece is arranged in this enhancement layer image, and this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image
Image is being carried out to hierarchical coding, for example, when spatial scalable is encoded, image can be carried out resolution processes to obtain low-resolution image, as a comparison original image is called to high-definition picture, encoder is respectively to the processing of encoding of this low-resolution image and this high-definition picture.For convenience of describing, the high image to be encoded of quality is called to enhancement layer image herein, the image low-quality to be encoded of correspondence (for example described low-resolution image) is called to basic tomographic image.
In embodiments of the present invention, target image is the image that uses hierarchical coding technology to process, basic layer refers to the lower layer of quality (comprising the parameters such as frame rate, spatial resolution, temporal resolution, signal to noise ratio intensity or credit rating) in hierarchical coding, and enhancement layer refers to the higher layer of quality (comprising the parameters such as frame rate, spatial resolution, temporal resolution, signal to noise ratio intensity or credit rating) in hierarchical coding.It should be noted that, in embodiments of the present invention, in embodiments of the present invention, for a given enhancement layer, basic layer corresponding thereto can be quality lower than arbitrary layer of this enhancement layer, for example, if five layers of current existence, (coding quality improves successively, ground floor quality is minimum, and layer 5 quality is the highest), if enhancement layer is the 4th layer, basic layer can be ground floor, can be also the second layer, also can be the 3rd layer, also can be the 4th layer.In like manner, for a given basic layer, enhancement layer corresponding thereto can be the arbitrary layer of quality lower than this basic layer.
Enhancement layer image is the image in the enhancement layer of pre-treatment, basic tomographic image be in basic layer with the image of enhancement layer image at synchronization.
In sum, in embodiments of the present invention, the quality of this basic tomographic image is lower than the quality of this enhancement layer image.
The image block of target image piece for processing in this enhancement layer image.
Basic tomographic image piece is in basic tomographic image, to have on locus the image block of corresponding relation with this target image piece.
In embodiments of the present invention, the corresponding relation of basic tomographic image piece and target image piece can calculate according to the resolution proportionate relationship between basic tomographic image and enhancement layer image.For example, within comprising the system of x direction and y direction, if enhancement layer image is respectively 2 times of basic tomographic image in the resolution of x direction and y direction, the pixel coordinate for the upper left corner in enhancement layer is (2x, 2y) and size be the image block of (2m) * (2n), corresponding blocks in its basic tomographic image can be that the pixel coordinate in the upper left corner is that (x, y) and size are the image block of m * n.
In embodiments of the present invention, image (comprising enhancement layer image and basic tomographic image) is divided into simple two class regions, that is, and smooth region and non-smooth region.At smooth region, during coding, conventionally can obtain predicted value comparatively accurately, thereby encoding block mark (CBF, CodedBlock Flag) definite while finally encoding is 0, residual error is 0.At non-smooth region, during coding because characteristics of signals correlation is little, cannot accurately predicting, therefore can not obtain predicted value comparatively accurately, cause predicting that rear residual error is larger or inhomogeneous, thereby the CBF of final coding is not 0.
Therefore, in embodiments of the present invention, can determine that the image block that in basic tomographic image, CBF is 0 (the smooth region image block of basic tomographic image) is positioned at smooth region, thereby, can think that the image block (the smooth region image block of enhancement layer image) in the enhancement layer image corresponding with this image block is positioned at smooth region.In like manner, can determine that CBF in basic tomographic image is not that 0 image block (the non-flat skating area area image piece of basic tomographic image) is positioned at non-smooth region, thereby, can think that the image block (the non-flat skating area area image piece of enhancement layer image) in the enhancement layer image corresponding with this image block is positioned at non-smooth region.As mentioned above, can determine the smooth region image block of enhancement layer image and the smooth region image block of this basic tomographic image, and, the non-flat skating area area image piece of enhancement layer image and the non-flat skating area area image piece of this basic tomographic image.
Thereafter, at S110, can be according to the non-flat skating area area image piece training first training filter of the non-flat skating area area image piece of enhancement layer image and this basic tomographic image, as example and non-limiting, can by the following method, train this first training filter.
As shown in Figure 2, y1 (n) represents the smooth region image block (set that a plurality of images are fast of enhancement layer image (original image), below, be called original smooth region image block), x (n) represents the smooth region image block of basic tomographic image, H (n) represents filter to be determined (the first training filter), and e (n) represents the smooth region image block of later basic tomographic image (being denoted as predicted picture piece) and original smooth region image block after filtering.For example, so that the mean square error of the pixel value of original smooth region image block and predicted picture piece is minimum, as optimal conditions (first pre-conditioned an example), solve H (n).
The optimal solution of this H (n) can be expressed as with following formula (1),
H = R xx - 1 × r yx - - - ( 1 )
Wherein, R xxthe auto-correlation function that represents original smooth region image block, r yxrepresent the smooth region image block of basic tomographic image and the cross-correlation function of original smooth region image block.
Should be understood that the above method of obtaining training filter of enumerating is only exemplary illustration, the present invention is not limited thereto, for example, can be by changing above-mentioned optimal condition etc., the method that this is obtained to training filter is carried out various changes.
Alternatively, the method also comprises:
According to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image; And
This,, according to this first training filter, determines alternative up-sampling filter, further comprises:
According to this first training filter and this second training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
In embodiments of the present invention, except the sample of training is different, (the first training filter is that the non-flat skating area area image piece of enhancement layer image and the non-flat skating area area image piece training of basic tomographic image obtain, the second training filter is that the smooth region image block of this enhancement layer image and the training of the smooth region image block of this basic tomographic image obtain), outer other similar process, here, for fear of repeating, the description thereof will be omitted.
Pass through said method, can obtain respectively the first training filter and the second training filter, therefore, in embodiments of the present invention, the coefficient of training filter is fixed, in cataloged procedure, do not need to consume bit transmission filter coefficient, simultaneously when decoding end selective filter, for example, can judge type (that is, being smooth region image block or non-flat skating area area image piece) under current block according to the CBF value of basic layer corresponding blocks, thereby, the index of the filter using without each image block of transmission, has improved the performance that image is processed.
Thereby, in alternative filter, can comprise this first training filter and the second training filter, that is, and situation 1.
Situation 1
Alternatively, should from alternative up-sampling filter, determine target up-sampling filter, comprise:
According to the characteristic information of this basic tomographic image piece, determine the smoothness of this target image, wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece;
According to the smoothness of this target image piece, determine this target up-sampling filter.
Specifically, in the situation that this alternative up-sampling filter comprises the first training filter and the second training filter obtaining as mentioned above, in S 130, can be according to the type of target image piece (being smooth region image block or non-flat skating area area image piece), select target up-sampling filter.
In embodiments of the present invention, when the CBF of the basic tomographic image piece corresponding with this target image piece is 0, can determine that this target image piece is positioned at smooth region.In like manner, when the CBF of the basic tomographic image piece corresponding with this target image piece is not 0, can determine that this target image piece is positioned at non-smooth region.
Should understand; the method of smoothness and the parameter of use of more than enumerating the image block that really sets the goal are only exemplary illustration; the present invention is not limited thereto; other can determine the smoothness of target image piece;; determine the smooth region of target image piece in enhancement layer image or method and the parameter of non-smooth region, all fall into protection scope of the present invention.
In the case, at S140, can determine as mentioned above according to this target up-sampling filter, the reconstruction image of basic tomographic image piece is carried out to up-sampling, thereby determines the predicted value with respect to target image piece.
Thereby, at S150, can calculate the distortion rate cost of each predictive mode, and determine the optimal prediction modes using when this target image piece is encoded.It should be noted that, when the target image piece to after this coding is decoded processing, can adopt with above-mentioned same method and determine target up-sampling filter.
Alternatively, this alternative up-sampling filter also comprises conventional filter, and
This is according to this information of forecasting, to the processing of encoding of this target image piece, to generate target code stream, comprising:
According to this information of forecasting, to the processing of encoding of this target image piece, to generate target code stream, this target code stream comprises the first indication information that is used to indicate this target up-sampling filter, this first indication information in order in decoding during the target image piece after above-mentioned coding as the foundation of obtaining this target up-sampling filter.
In embodiments of the present invention, as conventional filter, for example, can comprise existing cosine interpolation filter.Below, in explanation, for convenience of explanation, take this conventional filter as cosine interpolation filter is example, describe.
Specifically, in embodiments of the present invention, the above-mentioned method (training process) of obtaining training filter is carried out in pixel domain, and, can carry out different training for different down-sampling ratios (the resolution ratio between enhancement layer image and basic tomographic image), for example, existence is for the different situation of filter of whole location of pixels and half-pixel position acquisition, therefore, may cause the prediction of whole location of pixels and half-pixel position unbalanced, thereby the stationarity of impact prediction residual error, therefore, in embodiments of the present invention, can also in alternative up-sampling filter, can also add conventional filter, for example, cosine interpolation filter, the acquisition methods of this cosine interpolation filter can be same as the prior art, here, for fear of repeating, the description thereof will be omitted.
Thereby, in alternative filter, can comprise this first training filter and the second training filter and cosine interpolation filter (example of conventional filter), that is, and situation 2.
Situation 2
In S130, for example, can utilization rate aberration optimizing (RDO, Rate-Distortion Optimization), choice for use cosine interpolation filter or the first training filter or the second training filter, as target up-sampling filter.
In the case, thereafter, at S140, can to the reconstruction image of basic tomographic image piece, carry out up-sampling according to this target up-sampling filter, thereby determine the predicted value with respect to target image piece.
Thereby, at S150, can calculate the distortion rate cost of each predictive mode, and determine the optimal prediction modes using when this target image piece is encoded.In the situation that this optimal prediction modes is inter-layer texture prediction pattern, can be to being used to indicate this target up-sampling filter (specifically, indicating this target up-sampling filter is cosine interpolation filter, the first training filter or which filter of the first training in filter) the processing of encoding of the first indication information, and the first indication information after coding is processed adds code stream, while usining target image piece after the above-mentioned coding of decoding as obtaining this target up-sampling filter foundation.It should be noted that, in alternative filter, comprise this cosine interpolation filter or the first training filter or the second training filter, during as target up-sampling filter, alternative filter is three, and, coding side or decoding end can be according to the smoothnesses of target image piece as mentioned above, determine and use the first training filter or the second training filter, therefore, can in code stream, only with a bit identifier, carry this first indication information, take indicate this target up-sampling filter as conventional filter (cosine interpolation filter) still train filter (first training filter or second training filter).In alternative filter, the number of conventional filter during at least two, can, by increasing the figure place of the first indication information, realize the object of indicating target image block.
When carrying out up-sampling by cosine interpolation filter, be always the corresponding pixel points that directly copies basic tomographic image, but, for example, in scalable video coding method, the pixel of basic tomographic image is not directly to copy the pixel of corresponding enhancement layer image, now, if the whole pixel of target image is positioned at catastrophe point, while carrying out up-sampling, directly copies basic layer corresponding pixel points and there will be very large error.Therefore make to carry out in this way up-sampling, may make the whole location of pixels of enhancement layer image have larger error, therefore, in embodiments of the present invention, the training filter (the first training filter or the second training filter) that uses the method training of the embodiment of the present invention to obtain by the pixel to said mutation point, can improve the effect of up-sampling, improve the performance that coding and image are processed.
Alternatively, this alternative up-sampling filter also comprises conventional filter, and
This determines target up-sampling filter, comprising:
Determine the texture degree of this target image piece;
According to the texture degree of this target image piece, determine this target up-sampling filter.
As mentioned above, in the invention process, in the situation that this alternative up-sampling filter comprises cosine interpolation filter, the first training filter and the second training filter.As mentioned above, according to after RDO select target up-sampling filter, need in code stream, transmit for carrying the marker bit of the first indication information after coding, during with target image piece after the above-mentioned coding of decoding, determine which filter target up-sampling filter is.Therefore, in this programme, can adopt two candidate, but by RDO, not select, but select according to the different characteristic of different filters, thereby avoid a marker bit of many transmission to represent target up-sampling filter.That is, this alternative filter comprises the first training filter and conventional filter, that is, and and situation 3
Situation 3
In S130, for example, can be according to the characteristic information of basic tomographic image piece select target up-sampling filter from the first training filter and cosine interpolation filter (example of conventional filter), thereby avoid transmitting marker bit and represent used filter, for example, can adopt the filter of determining choice for use according to the texture strength of basic tomographic image piece, in embodiments of the present invention, can adopt, for example, rim detection, the methods such as texture strength analysis or intraframe coding predictive mode, determine the texture degree of basic tomographic image piece, and, for example, when texture degree is stronger, choice for use training gained filter is as target up-sampling filter, all the other all use cosine interpolation filter.That more than enumerates passes through rim detection, and the methods such as texture strength analysis or intraframe coding predictive mode determine that the method for the texture degree of basic tomographic image piece can be same as the prior art, and here, for fear of repeating, the description thereof will be omitted.
In the case, thereafter, at S140, can to the reconstruction image of basic tomographic image piece, carry out up-sampling according to this target up-sampling filter, thereby determine the predicted value with respect to target image piece.
Thereby, at S150, can calculate the distortion rate cost of each predictive mode, and determine the optimal prediction modes using when this target image piece is encoded.And to the processing of encoding of target image piece.
When carrying out up-sampling by cosine interpolation filter, be always the corresponding pixel points that directly copies basic tomographic image, but, for example, in scalable video coding method, the pixel of basic tomographic image is not directly to copy the pixel of corresponding enhancement layer image, now, if the whole pixel of target image is positioned at catastrophe point, while carrying out up-sampling, directly copies basic layer corresponding pixel points and there will be very large error.The first training filter can recover preferably to the pixel of catastrophe point position, therefore can well make up the deficiency of cosine interpolation filter.
It should be noted that, when the target image piece to after this coding is decoded processing, can adopt with above-mentioned same method and determine target up-sampling filter.
According to the method for processing for image of the embodiment of the present invention, by according to smooth region and non-smooth region, determine up-sampling filter, can be when improving the effect of up-sampling, reduce the quantity of up-sampling filter and without transmission filter coefficient and index, thereby can improve coding efficiency, improve effect and performance that image is processed.
Fig. 3 show from coding side angle, describe according to the indicative flowchart of the method 200 of processing for image of the embodiment of the present invention.As shown in Figure 3, the method 200 comprises:
S210, according to the non-flat skating area area image piece of the non-flat skating area area image piece of enhancement layer image and basic tomographic image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image;
S220, according to this first training filter, determines alternative up-sampling filter, and wherein, this alternative up-sampling filter comprises this first training filter;
S230, from alternative up-sampling filter, determines target up-sampling filter, and this alternative up-sampling filter comprises this first training filter;
S240, according to this target up-sampling filter and basic tomographic image piece, determines information of forecasting.
S250, according to this information of forecasting and the residual information that obtains from target code stream, to the processing of decoding of this object code stream, to obtain this target image piece, wherein, this target image piece is arranged in this enhancement layer image, and this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image
Image is being carried out to hierarchical coding, for example, when spatial scalable is encoded, image can be carried out resolution processes to obtain low-resolution image, as a comparison original image is called to high-definition picture, encoder is respectively to the processing of encoding of this low-resolution image and this high-definition picture.For convenience of describing, the high image to be encoded of quality is called to enhancement layer image herein, the image low-quality to be encoded of correspondence (for example described low-resolution image) is called to basic tomographic image.
In embodiments of the present invention, target image is the image that uses hierarchical coding technology to process, basic layer refers to the lower layer of quality (comprising the parameters such as frame rate, spatial resolution, temporal resolution, signal to noise ratio intensity or credit rating) in hierarchical coding, and enhancement layer refers to the higher layer of quality (comprising the parameters such as frame rate, spatial resolution, temporal resolution, signal to noise ratio intensity or credit rating) in hierarchical coding.It should be noted that, in embodiments of the present invention, in embodiments of the present invention, for a given enhancement layer, basic layer corresponding thereto can be quality lower than arbitrary layer of this enhancement layer, for example, if five layers of current existence, (coding quality improves successively, ground floor quality is minimum, and layer 5 quality is the highest), if enhancement layer is the 4th layer, basic layer can be ground floor, can be also the second layer, also can be the 3rd layer, also can be the 4th layer.In like manner, for a given basic layer, enhancement layer corresponding thereto can be the arbitrary layer of quality lower than this basic layer.
Enhancement layer image is the image in the enhancement layer of pre-treatment, basic tomographic image be in basic layer with the image of enhancement layer image at synchronization.
In sum, in embodiments of the present invention, the quality of this basic tomographic image is lower than the quality of this enhancement layer image.
The image block of target image piece for processing in this enhancement layer image.
Basic tomographic image piece is in basic tomographic image, to have on locus the image block of corresponding relation with this target image piece.
In embodiments of the present invention, the corresponding relation of basic tomographic image piece and target image piece can calculate according to the resolution proportionate relationship between basic tomographic image and enhancement layer image.For example, within comprising the system of x direction and y direction, if enhancement layer image is respectively 2 times of basic tomographic image in the resolution of x direction and y direction, the pixel coordinate for the upper left corner in enhancement layer is (2x, 2y) and size be the image block of (2m) * (2n), corresponding blocks in its basic tomographic image can be that the pixel coordinate in the upper left corner is that (x, y) and size are the image block of m * n.
In embodiments of the present invention, image (comprising enhancement layer image and basic tomographic image) is divided into simple two class regions, that is, and smooth region and non-smooth region.At smooth region, during coding, conventionally can obtain predicted value comparatively accurately, thereby encoding block mark (CBF, Coded Block Flag) definite while finally encoding is 0, residual error is 0.At non-smooth region, during coding because characteristics of signals correlation is little, cannot accurately predicting, therefore can not obtain predicted value comparatively accurately, cause predicting that rear residual error is larger or inhomogeneous, thereby the CBF of final coding is not 0.
Therefore, in embodiments of the present invention, can determine that the image block that in basic tomographic image, CBF is 0 (the smooth region image block of basic tomographic image) is positioned at smooth region, thereby, can think that the image block (the smooth region image block of enhancement layer image) in the enhancement layer image corresponding with this image block is positioned at smooth region.In like manner, can determine that CBF in basic tomographic image is not that 0 image block (the non-flat skating area area image piece of basic tomographic image) is positioned at non-smooth region, thereby, can think that the image block (the non-flat skating area area image piece of enhancement layer image) in the enhancement layer image corresponding with this image block is positioned at non-smooth region.As mentioned above, can determine the smooth region image block of enhancement layer image and the smooth region image block of this basic tomographic image, and, the non-flat skating area area image piece of enhancement layer image and the non-flat skating area area image piece of this basic tomographic image.
Thereafter, at S210, can be according to the non-flat skating area area image piece training first training filter of the non-flat skating area area image piece of enhancement layer image and this basic tomographic image, as example and non-limiting, can by the following method, train this first training filter.
As shown in Figure 2, y1 (n) represents the smooth region image block (set that a plurality of images are fast of enhancement layer image (original image), below, be called original smooth region image block), x (n) represents the smooth region image block of basic tomographic image, H (n) represents filter to be determined (the first training filter), and e (n) represents the smooth region image block of later basic tomographic image (being denoted as predicted picture piece) and original smooth region image block after filtering.For example, so that the mean square error of the pixel value of original smooth region image block and predicted picture piece is minimum, as optimal conditions (first pre-conditioned an example), solve H (n).
The optimal solution of this H (n) can be expressed as with above formula (1),
Should be understood that the above method of obtaining training filter of enumerating is only exemplary illustration, the present invention is not limited thereto, for example, can be by changing above-mentioned optimal condition etc., the method that this is obtained to training filter is carried out various changes.
Alternatively, the method also comprises:
According to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image; And
This,, according to this first training filter, determines alternative up-sampling filter, further comprises:
According to this first training filter and this second training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
In embodiments of the present invention, except the sample of training is different, (the first training filter is that the non-flat skating area area image piece of enhancement layer image and the non-flat skating area area image piece training of basic tomographic image obtain, the second training filter is that the smooth region image block of this enhancement layer image and the training of the smooth region image block of this basic tomographic image obtain), outer other similar process, here, for fear of repeating, the description thereof will be omitted.
Pass through said method, can obtain respectively the first training filter and the second training filter, therefore, in embodiments of the present invention, the coefficient of training filter is fixed, in cataloged procedure, do not need to consume bit transmission filter coefficient, simultaneously when decoding end selective filter, for example, can judge type (that is, being smooth region image block or non-flat skating area area image piece) under current block according to the CBF value of basic layer corresponding blocks, thereby, the index of the filter using without each image block of transmission, has improved the performance that image is processed.
Thereby, in alternative filter, can comprise this first training filter and the second training filter, that is, and situation 4.
Situation 4
Alternatively, should from alternative up-sampling filter, determine target up-sampling filter, comprise:
According to the characteristic information of this basic tomographic image piece, determine the smoothness of this target image, wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece;
According to the smoothness of this target image piece, determine this target up-sampling filter.
Specifically, in the situation that this alternative up-sampling filter comprises the first training filter and the second training filter obtaining as mentioned above, in S230, can be according to the type of target image piece (being smooth region image block or non-flat skating area area image piece), select target up-sampling filter.
In embodiments of the present invention, when the CBF of the basic tomographic image piece corresponding with this target image piece is 0, can determine that this target image piece is positioned at smooth region.In like manner, when the CBF of the basic tomographic image piece corresponding with this target image piece is not 0, can determine that this target image piece is positioned at non-smooth region.
Should understand; the method of smoothness and the parameter of use of more than enumerating the image block that really sets the goal are only exemplary illustration; the present invention is not limited thereto; other can determine the smoothness of target image piece;; determine the smooth region of target image piece in enhancement layer image or method and the parameter of non-smooth region, all fall into protection scope of the present invention.
In the case, at S240, can determine as mentioned above according to this target up-sampling filter, the reconstruction image of basic tomographic image piece is carried out to up-sampling, thereby determines the predicted value with respect to target image piece.
Thereby, at S250, can target image piece be decoded according to this predicted value and the residual information obtaining from target code stream, in embodiments of the present invention, this process can be same as the prior art, and for fear of repeating, the description thereof will be omitted here.
Alternatively, this alternative up-sampling filter also comprises conventional filter, and
Should from alternative up-sampling filter, determine target up-sampling filter, comprise:
From this target code stream, obtain the first indication information that is used to indicate this target up-sampling filter,
According to this first indication information, determine this target up-sampling filter.
In embodiments of the present invention, as conventional filter, for example, can comprise existing cosine interpolation filter.Below, in explanation, for convenience of explanation, take this conventional filter as cosine interpolation filter is example, describe.
Specifically, in embodiments of the present invention, the above-mentioned method (training process) of obtaining training filter is carried out in pixel domain, and, can carry out different training for different down-sampling ratios (the resolution ratio between enhancement layer image and basic tomographic image), for example, existence is for the different situation of filter of whole location of pixels and half-pixel position acquisition, therefore, may cause the prediction of whole location of pixels and half-pixel position unbalanced, thereby the stationarity of impact prediction residual error, therefore, in embodiments of the present invention, can also in alternative up-sampling filter, can also add conventional filter, for example, cosine interpolation filter, the acquisition methods of this cosine interpolation filter can be same as the prior art, here, for fear of repeating, the description thereof will be omitted.
Thereby, in alternative filter, can comprise this first training filter and the second training filter and cosine interpolation filter (example of conventional filter), that is, and situation 5.
Situation 5
In S230, for example, can from target code stream, obtain the first indication information, in embodiments of the present invention, in alternative filter, comprise this cosine interpolation filter or the first training filter or the second training filter, during as target up-sampling filter, alternative filter is three, and, coding side or decoding end can be according to the smoothnesses of target image piece as mentioned above, determine and use the first training filter or the second training filter, therefore, can in code stream, only with a bit identifier, carry this first indication information, take indicate this target up-sampling filter as conventional filter (cosine interpolation filter) still train filter (first training filter or second training filter).For example, when indication information is 1, selects training filter (comprising the first training filter or the second training filter), and according to the smoothness of this target image piece, from this first training filter or the second training filter, determine target up-sampling filter.For example, when indication information is 0, select cosine interpolation filter as target up-sampling filter.
Should be understood that the above method of passing through the first indication information indicating target up-sampling filter of enumerating is only exemplary illustration, the present invention is not limited thereto.
It should be noted that, when the number of conventional filter is at least two in alternative filter, can, by increasing the figure place of the first indication information, realize the object of indicating target image block.
In the case, at S240, can determine as mentioned above according to this target up-sampling filter, the reconstruction image of basic tomographic image piece is carried out to up-sampling, thereby determines the predicted value with respect to target image piece.
Thereby, at S250, can target image piece be decoded according to this predicted value and the residual information obtaining from target code stream, in embodiments of the present invention, this process can be same as the prior art, and for fear of repeating, the description thereof will be omitted here.
When carrying out up-sampling by cosine interpolation filter, be always the corresponding pixel points that directly copies basic tomographic image, but, for example, in scalable video coding method, the pixel of basic tomographic image is not directly to copy the pixel of corresponding enhancement layer image, now, if the whole pixel of target image is positioned at catastrophe point, while carrying out up-sampling, directly copies basic layer corresponding pixel points and there will be very large error.Therefore make to carry out in this way up-sampling, may make the whole location of pixels of enhancement layer image have larger error, therefore, in embodiments of the present invention, the training filter (the first training filter or the second training filter) that uses the method training of the embodiment of the present invention to obtain by the pixel to said mutation point, can improve the effect of up-sampling, improve the performance that coding and image are processed.
Alternatively, this alternative up-sampling filter also comprises conventional filter, and
This determines target up-sampling filter, comprising:
Determine the texture degree of this target image piece;
According to the texture degree of this target image piece, determine this target up-sampling filter.
As mentioned above, in the invention process, in the situation that this alternative up-sampling filter comprises cosine interpolation filter, the first training filter and the second training filter.As mentioned above, according to after RDO select target up-sampling filter, need in code stream, transmit for carrying the marker bit of the first indication information after coding, during with target image piece after the above-mentioned coding of decoding, determine which filter target up-sampling filter is.Therefore, in this programme, can adopt two candidate, but by RDO, not select, but select according to the different characteristic of different filters, thereby avoid a marker bit of many transmission to represent target up-sampling filter.That is, this alternative filter comprises the first training filter and conventional filter, that is, and and situation 6
Situation 6
In S230, for example, can be according to the characteristic information of basic tomographic image piece select target up-sampling filter from the first training filter and cosine interpolation filter (example of conventional filter), thereby avoid transmitting marker bit and represent used filter, for example, can adopt the filter of determining choice for use according to the texture strength of basic tomographic image piece, in embodiments of the present invention, can adopt, for example, rim detection, the methods such as texture strength analysis or intraframe coding predictive mode, determine the texture degree of basic tomographic image piece, and, for example, when texture degree is stronger, choice for use training gained filter is as target up-sampling filter, all the other all use cosine interpolation filter.That more than enumerates passes through rim detection, and the methods such as texture strength analysis or intraframe coding predictive mode determine that the method for the texture degree of basic tomographic image piece can be same as the prior art, and here, for fear of repeating, the description thereof will be omitted.
In the case, at S240, can determine as mentioned above according to this target up-sampling filter, the reconstruction image of basic tomographic image piece is carried out to up-sampling, thereby determines the predicted value with respect to target image piece.
Thereby, at S250, can target image piece be decoded according to this predicted value and the residual information obtaining from target code stream, in embodiments of the present invention, this process can be same as the prior art, and for fear of repeating, the description thereof will be omitted here.
When carrying out up-sampling by cosine interpolation filter, be always the corresponding pixel points that directly copies basic tomographic image, but, for example, in scalable video coding method, the pixel of basic tomographic image is not directly to copy the pixel of corresponding enhancement layer image, now, if the whole pixel of target image is positioned at catastrophe point, while carrying out up-sampling, directly copies basic layer corresponding pixel points and there will be very large error.The first training filter can recover preferably to the pixel of catastrophe point position, therefore can well make up the deficiency of cosine interpolation filter.
According to the method for processing for image of the embodiment of the present invention, by according to smooth region and non-smooth region, determine up-sampling filter, can be when improving the effect of up-sampling, reduce the quantity of up-sampling filter and without transmission filter coefficient and index, thereby can improve coding efficiency, improve effect and performance that image is processed.
Above, in conjunction with Fig. 1 to Fig. 3, describe the method for processing for image according to the embodiment of the present invention in detail, below, in connection with Fig. 4 to Fig. 5, describe the device of processing for image according to the embodiment of the present invention in detail.
Fig. 4 shows according to the schematic block diagram of the device 300 of processing for image of the embodiment of the present invention.As shown in Figure 7, this device 300 comprises:
Acquiring unit 310, for the non-flat skating area area image piece with basic tomographic image according to the non-flat skating area area image piece of enhancement layer image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image;
Determining unit 320, for according to this first training filter, determines alternative up-sampling filter, and wherein, this alternative up-sampling filter comprises this first training filter;
For from this alternative up-sampling filter, determine target up-sampling filter, and to coding unit 330 these target up-sampling filters of transmission, this alternative up-sampling filter comprises this first training filter that this acquiring unit 310 obtains;
Coding unit 330, for obtaining this up-sampling filter from this determining unit, and according to this target up-sampling filter and basic tomographic image piece, determines information of forecasting;
Be used for according to this information of forecasting, to the processing of encoding of target image piece, to generate target code stream, wherein, this target image piece is arranged in this enhancement layer image, this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image.
Alternatively, this acquiring unit 310 is also for according to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image; And
This determining unit 320 is further used for, according to this first training filter and this second training filter, determining alternative up-sampling filter, and wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
Alternatively, this determining unit 320, specifically for according to the characteristic information of this basic tomographic image piece, is determined the smoothness of this target image, and wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece;
For according to the smoothness of this target image piece, determine this target up-sampling filter.
Alternatively, this determining unit 320 is further used for according to this first training filter, this second training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, this second training filter and this conventional filter; And
This coding unit 330 is specifically for according to this information of forecasting, to the processing of encoding of this target image piece, to generate target code stream, this target code stream comprises the first indication information that is used to indicate this target up-sampling filter, this first indication information in order in decoding during the target image piece after above-mentioned coding as the foundation of obtaining this target up-sampling filter.
Alternatively, this alternative up-sampling filter also comprises conventional filter, and
This determining unit 320 is specifically for determining the texture degree of this target image piece;
For according to the texture degree of this target image piece, determine this target up-sampling filter.
According to the device of processing for image of the embodiment of the present invention, by according to smooth region and non-smooth region, determine up-sampling filter, can be when improving the effect of up-sampling, reduce the quantity of up-sampling filter and without transmission filter coefficient and index, thereby can improve coding efficiency, improve effect and performance that image is processed.
Can be corresponding to the coding side of the method for the embodiment of the present invention according to the device 300 of processing for image of the embodiment of the present invention, and, each unit of the device 300 that should process for image is that module and above-mentioned other operations and/or function are respectively in order to realize the corresponding flow process of the method 100 in Fig. 1, for simplicity, do not repeat them here.
Fig. 5 shows according to the schematic block diagram of the device 400 of processing for image of the embodiment of the present invention.As shown in Figure 5, this device 500 comprises:
Acquiring unit 510, for the non-flat skating area area image piece with basic tomographic image according to the non-flat skating area area image piece of enhancement layer image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image;
Determining unit 520, for according to this first training filter, determines alternative up-sampling filter, and wherein, this alternative up-sampling filter comprises this first training filter;
Be used for from alternative up-sampling filter, determine target up-sampling filter, this alternative up-sampling filter comprises this first training filter, and to decoding unit 530 these target up-sampling filters of transmission, this alternative up-sampling filter comprises that this acquiring unit obtains this first training filter of 510;
Decoding unit 530, for obtaining this target up-sampling filter from this determining unit 520, and according to this target up-sampling filter and basic tomographic image piece, determines information of forecasting;
For according to this information of forecasting and the residual information that obtains from target code stream, to the processing of decoding of target image piece, wherein, this target image piece is arranged in this enhancement layer image, this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image.
Alternatively, this acquiring unit 510 is also for according to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image; And
This determining unit 520 is further used for, according to this first training filter and this second training filter, determining alternative up-sampling filter, and wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
Alternatively, this determining unit 520, specifically for according to the characteristic information of this basic tomographic image piece, is determined the smoothness of this target image, and wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece;
According to the smoothness of this target image piece, determine this target up-sampling filter.
Alternatively, this alternative up-sampling filter also comprises conventional filter, and
This determining unit 520, specifically for from this target code stream, is obtained the first indication information that is used to indicate this target up-sampling filter,
For according to this first indication information, determine this target up-sampling filter.
Alternatively, this alternative up-sampling filter also comprises conventional filter, and
This determining unit 520 is specifically for determining the texture degree of this target image piece;
For according to the texture degree of this target image piece, determine this target up-sampling filter.
According to the device of processing for image of the embodiment of the present invention, by according to smooth region and non-smooth region, determine up-sampling filter, can be when improving the effect of up-sampling, reduce the quantity of up-sampling filter and without transmission filter coefficient and index, thereby can improve coding efficiency, improve effect and performance that image is processed.
Can be corresponding to the coding side of the method for the embodiment of the present invention according to the device 400 of processing for image of the embodiment of the present invention, and, each unit of the device 400 that should process for image is that module and above-mentioned other operations and/or function are respectively in order to realize the corresponding flow process of the method 200 in Fig. 3, for simplicity, do not repeat them here.
Fig. 6 shows according to the schematic block diagram of the encoder 500 of processing for image of the embodiment of the present invention.As shown in Figure 6, this decoder 500 comprises:
Bus 510;
The processor 520 being connected with this bus;
The memory 530 being connected with this bus;
Wherein, this processor 520 is by this bus 510, call the program of storage in this memory 530, with the non-flat skating area area image piece with basic tomographic image according to the non-flat skating area area image piece of enhancement layer image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image,
According to this first training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter;
From this alternative up-sampling filter, determine target up-sampling filter;
According to this target up-sampling filter and basic tomographic image piece, determine information of forecasting;
According to this information of forecasting, to the processing of encoding of target image piece, to generate target code stream, wherein, this target image piece is arranged in this enhancement layer image, this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image
Alternatively, this processing unit 520 is also for according to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image;
For according to this first training filter and this second training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
Alternatively, this processing unit 520, specifically for according to the characteristic information of this basic tomographic image piece, is determined the smoothness of this target image, and wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece;
According to the smoothness of this target image piece, determine this target up-sampling filter.
Alternatively, this processing unit 520 is specifically for according to this first training filter, this second training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, this second training filter and this conventional filter;
Be used for according to this information of forecasting, to the processing of encoding of this target image piece, to generate target code stream, this target code stream comprises the first indication information that is used to indicate this target up-sampling filter, this first indication information in order in decoding during the target image piece after above-mentioned coding as the foundation of obtaining this target up-sampling filter.
Alternatively, this processing unit 520, specifically for according to this first training filter and conventional filter, is determined alternative up-sampling filter, and wherein, this alternative up-sampling filter comprises this first training filter and this conventional filter;
For determining the texture degree of this target image piece;
For according to the texture degree of this target image piece, determine this target up-sampling filter.
Be used for the encoder of processing for image according to the embodiment of the present invention, by according to smooth region and non-smooth region, determine up-sampling filter, can be when improving the effect of up-sampling, reduce the quantity of up-sampling filter and without transmission filter coefficient and index, thereby can improve coding efficiency, improve effect and performance that image is processed.
Can be corresponding to the coding side of the method for the embodiment of the present invention according to the encoder 500 of processing for image of the embodiment of the present invention, and, each unit of the encoder 500 that should process for image is that module and above-mentioned other operations and/or function are respectively in order to realize the corresponding flow process of the method 100 in Fig. 1, for simplicity, do not repeat them here.
Fig. 7 shows according to the schematic block diagram of the decoder 600 of processing for image of the embodiment of the present invention.As shown in Figure 7, this decoder 600 comprises:
Bus 610;
The processor 620 being connected with this bus;
The memory 630 being connected with this bus;
Wherein, this processor 620 is by this bus 610, call the program of storage in this memory 630, for the non-flat skating area area image piece with basic tomographic image according to the non-flat skating area area image piece of enhancement layer image, determine the first training filter, so that this first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of this first training filter and this basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of this enhancement layer image are satisfied first pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image,
According to this first training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter;
From alternative up-sampling filter, determine target up-sampling filter, this alternative up-sampling filter comprises this first training filter;
According to this target up-sampling filter and basic tomographic image piece, determine information of forecasting;
According to this information of forecasting and the residual information that obtains from target code stream, to the processing of decoding of this object code stream, to obtain this target image piece, wherein, this target image piece is arranged in this enhancement layer image, this basic tomographic image piece is arranged in this basic tomographic image, and the locus of this primary image piece in this basic tomographic image is corresponding with this target image piece locus in this enhancement layer image.
Alternatively, this processing unit 620 is also for according to the smooth region image block of the smooth region image block of this enhancement layer image and this basic tomographic image, determine the second training filter, so that this second training filter meets: the second information of forecasting of determining according to the smooth region image block of this second training filter and this basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of this enhancement layer image are satisfied second pre-conditioned, wherein, the basic tomographic image being somebody's turn to do is corresponding with this enhancement layer image; And
For according to this first training filter and this second training filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter and this second training filter.
Alternatively, this processing unit 620, specifically for according to the characteristic information of this basic tomographic image piece, is determined the smoothness of this target image, and wherein, this characteristic information comprises encoding block label information or the residual information to this basic tomographic image piece of this basic tomographic image piece;
According to the smoothness of this target image piece, determine this target up-sampling filter.
Alternatively, this processing unit 620 is specifically for according to this first training filter, this second training filter and conventional filter, determine alternative up-sampling filter, wherein, this alternative up-sampling filter comprises this first training filter, this second training filter and this conventional filter;
For from this target code stream, obtain the first indication information that is used to indicate this target up-sampling filter, according to this first indication information, determine this target up-sampling filter.。
Alternatively, this processing unit 620, specifically for according to this first training filter and conventional filter, is determined alternative up-sampling filter, and wherein, this alternative up-sampling filter comprises this first training filter and this conventional filter;
For determining the texture degree of this target image piece;
According to the texture degree of this target image piece, determine this target up-sampling filter.
The decoder of processing according to the image of the embodiment of the present invention, by according to smooth region and non-smooth region, determine up-sampling filter, can be when improving the effect of up-sampling, reduce the quantity of up-sampling filter and without transmission filter coefficient and index, thereby can improve coding efficiency, improve effect and performance that image is processed.
Can be corresponding to the decoding end of the method for the embodiment of the present invention according to the decoder 600 of processing for image of the embodiment of the present invention, and, each unit of the decoder 600 that should process for image is that module and above-mentioned other operations and/or function are respectively in order to realize the corresponding flow process of the method 200 in Fig. 3, for simplicity, do not repeat them here.
It should be noted that, in order to make coding side consistent with the target up-sampling filter of decoding end (encoder, decoder) use, so require coding side consistent with the method for decoding end select target up-sampling filter.In other words, can be according to described coding side processing method correspondence definite decoding end processing method really, or the definite coding side processing method corresponding according to described decoding end processing method.
Should be understood that term "and/or" herein, is only a kind of incidence relation of describing affiliated partner, and expression can exist three kinds of relations, and for example, A and/or B, can represent: individualism A exists A and B, these three kinds of situations of individualism B simultaneously.In addition, character "/", generally represents that forward-backward correlation is to liking a kind of relation of "or" herein.
Should understand, in various embodiment of the present invention, the size of the sequence number of above-mentioned each process does not also mean that the priority of execution sequence, and the execution sequence of each process should determine with its function and internal logic, and should not form any restriction to the implementation process of the embodiment of the present invention.
Those of ordinary skills can recognize, unit and the algorithm steps of each example of describing in conjunction with embodiment disclosed herein, can realize with the combination of electronic hardware or computer software and electronic hardware.These functions are carried out with hardware or software mode actually, depend on application-specific and the design constraint of technical scheme.Professional and technical personnel can specifically should be used for realizing described function with distinct methods to each, but this realization should not thought and exceeds scope of the present invention.
Those skilled in the art can be well understood to, and for convenience and simplicity of description, the specific works process of the system of foregoing description, device and unit, can, with reference to the corresponding process in preceding method embodiment, not repeat them here.
In the several embodiment that provide in the application, should be understood that disclosed system, apparatus and method can realize by another way.For example, device embodiment described above is only schematic, for example, the division of described unit, be only that a kind of logic function is divided, during actual realization, can have other dividing mode, for example a plurality of unit or assembly can in conjunction with or can be integrated into another system, or some features can ignore, or do not carry out.Another point, shown or discussed coupling each other or direct-coupling or communication connection can be by some interfaces, indirect coupling or the communication connection of device or unit can be electrically, machinery or other form.
The described unit as separating component explanation can or can not be also physically to separate, and the parts that show as unit can be or can not be also physical locations, can be positioned at a place, or also can be distributed in a plurality of network element.Can select according to the actual needs some or all of unit wherein to realize the object of the present embodiment scheme.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, can be also that the independent physics of unit exists, and also can be integrated in a unit two or more unit.
If described function usings that the form of SFU software functional unit realizes and during as production marketing independently or use, can be stored in a computer read/write memory medium.Understanding based on such, the part that technical scheme of the present invention contributes to prior art in essence in other words or the part of this technical scheme can embody with the form of software product, this computer software product is stored in a storage medium, comprise that some instructions are with so that a computer equipment (can be personal computer, server, or the network equipment etc.) carry out all or part of step of method described in each embodiment of the present invention.And aforesaid storage medium comprises: various media that can be program code stored such as USB flash disk, portable hard drive, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disc or CDs.
The above; be only the specific embodiment of the present invention, but protection scope of the present invention is not limited to this, is anyly familiar with those skilled in the art in the technical scope that the present invention discloses; can expect easily changing or replacing, within all should being encompassed in protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection range of described claim.

Claims (20)

1. a method of processing for image, is characterized in that, described method comprises:
According to the non-flat skating area area image piece of the non-flat skating area area image piece of enhancement layer image and basic tomographic image, determine the first training filter, so that described the first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of described the first training filter and described basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of described enhancement layer image are satisfied first pre-conditioned, wherein, described basic tomographic image is corresponding with described enhancement layer image;
According to described the first training filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter;
From described alternative up-sampling filter, determine target up-sampling filter;
According to described target up-sampling filter and basic tomographic image piece, determine information of forecasting;
According to described information of forecasting, to the processing of encoding of target image piece, to generate target code stream, wherein, described target image piece is arranged in described enhancement layer image, described basic tomographic image piece is arranged in described basic tomographic image, and the locus of described primary image piece in described basic tomographic image is corresponding with the locus of described target image piece in described enhancement layer image.
2. method according to claim 1, is characterized in that, described method also comprises:
According to the smooth region image block of the smooth region image block of described enhancement layer image and described basic tomographic image, determine the second training filter, so that described the second training filter meets: the second information of forecasting of determining according to the smooth region image block of described the second training filter and described basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of described enhancement layer image are satisfied second pre-conditioned, wherein, described basic tomographic image is corresponding with described enhancement layer image; And
Described according to described first training filter, determine alternative up-sampling filter, further comprise:
According to described the first training filter and described the second training filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter and described the second training filter.
3. method according to claim 2, is characterized in that, described from alternative up-sampling filter, determines target up-sampling filter, comprising:
According to the characteristic information of described basic tomographic image piece, determine the smoothness of described target image, wherein, described characteristic information comprises encoding block label information or the residual information to described basic tomographic image piece of described basic tomographic image piece;
According to the smoothness of described target image piece, determine described target up-sampling filter.
4. method according to claim 2, is characterized in that, according to described the first training filter and described the second training filter, determines alternative up-sampling filter, further comprises:
According to described the first training filter, described the second training filter and conventional filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter, described the second training filter and described conventional filter; And
Described according to described information of forecasting, to the processing of encoding of described target image piece, to generate target code stream, comprising:
According to described information of forecasting, to the processing of encoding of described target image piece, to generate target code stream, described target code stream comprises the first indication information that is used to indicate described target up-sampling filter, described the first indication information in order in decoding during the target image piece after above-mentioned coding as the foundation of obtaining described target up-sampling filter.
5. method according to claim 1, is characterized in that, described according to described first training filter, determine alternative up-sampling filter, further comprise:
According to described the first training filter and conventional filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter and described conventional filter; And
Described definite target up-sampling filter, comprising:
Determine the texture degree of described target image piece;
According to the texture degree of described target image piece, determine described target up-sampling filter.
6. a method of processing for image, is characterized in that, described method comprises:
According to the non-flat skating area area image piece of the non-flat skating area area image piece of enhancement layer image and basic tomographic image, determine the first training filter, so that described the first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of described the first training filter and described basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of described enhancement layer image are satisfied first pre-conditioned, wherein, described basic tomographic image is corresponding with described enhancement layer image;
According to described the first training filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter;
From alternative up-sampling filter, determine target up-sampling filter, described alternative up-sampling filter comprises described the first training filter;
According to described target up-sampling filter and basic tomographic image piece, determine information of forecasting;
According to described information of forecasting and the residual information that obtains from target code stream, to the processing of decoding of described object code stream, to obtain described target image piece, wherein, described target image piece is arranged in described enhancement layer image, described basic tomographic image piece is arranged in described basic tomographic image, and the locus of described primary image piece in described basic tomographic image is corresponding with the locus of described target image piece in described enhancement layer image.
7. method according to claim 6, is characterized in that, described method also comprises:
According to the smooth region image block of the smooth region image block of described enhancement layer image and described basic tomographic image, determine the second training filter, so that described the second training filter meets: the second information of forecasting of determining according to the smooth region image block of described the second training filter and described basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of described enhancement layer image are satisfied second pre-conditioned, wherein, described basic tomographic image is corresponding with described enhancement layer image; And
Described according to described first training filter, determine alternative up-sampling filter, further comprise:
According to described the first training filter and described the second training filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter and described the second training filter.
8. method according to claim 7, is characterized in that, described from alternative up-sampling filter, determines target up-sampling filter, comprising:
According to the characteristic information of described basic tomographic image piece, determine the smoothness of described target image, wherein, described characteristic information comprises encoding block label information or the residual information to described basic tomographic image piece of described basic tomographic image piece;
According to the smoothness of described target image piece, determine described target up-sampling filter.
9. method according to claim 8, is characterized in that, according to described the first training filter and described the second training filter, determines alternative up-sampling filter, further comprises:
According to described the first training filter, described the second training filter and conventional filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter, described the second training filter and described conventional filter; And
Described from alternative up-sampling filter, determine target up-sampling filter, comprising:
From described target code stream, obtain the first indication information that is used to indicate described target up-sampling filter,
According to described the first indication information, determine described target up-sampling filter.
10. method according to claim 6, is characterized in that, described according to described first training filter, determine alternative up-sampling filter, further comprise:
According to described the first training filter and conventional filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter and described conventional filter; And
Described definite target up-sampling filter, comprising:
Determine the texture degree of described target image piece;
According to the texture degree of described target image piece, determine described target up-sampling filter.
11. 1 kinds of devices of processing for image, is characterized in that, described device comprises:
Acquiring unit, for the non-flat skating area area image piece with basic tomographic image according to the non-flat skating area area image piece of enhancement layer image, determine the first training filter, so that described the first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of described the first training filter and described basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of described enhancement layer image are satisfied first pre-conditioned, wherein, described basic tomographic image is corresponding with described enhancement layer image;
Determining unit, according to described the first training filter, determines alternative up-sampling filter, and wherein, described alternative up-sampling filter comprises described the first training filter;
For from described alternative up-sampling filter, determine target up-sampling filter;
Coding unit, for obtaining described up-sampling filter from described determining unit, and according to described target up-sampling filter and basic tomographic image piece, determines information of forecasting;
Be used for according to described information of forecasting, to the processing of encoding of target image piece, to generate target code stream, wherein, described target image piece is arranged in described enhancement layer image, described basic tomographic image piece is arranged in described basic tomographic image, and the locus of described primary image piece in described basic tomographic image is corresponding with the locus of described target image piece in described enhancement layer image.
12. devices according to claim 11, it is characterized in that, described acquiring unit is also for according to the smooth region image block of the smooth region image block of described enhancement layer image and described basic tomographic image, determine the second training filter, so that described the second training filter meets: the second information of forecasting of determining according to the smooth region image block of described the second training filter and described basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of described enhancement layer image are satisfied second pre-conditioned, wherein, described basic tomographic image is corresponding with described enhancement layer image, and
Described determining unit is further used for training filter according to described the first training filter and described second, determines alternative up-sampling filter, and wherein, described alternative up-sampling filter comprises described the first training filter and described the second training filter.
13. devices according to claim 12, it is characterized in that, described determining unit is specifically for according to the characteristic information of described basic tomographic image piece, determine the smoothness of described target image, wherein, described characteristic information comprises encoding block label information or the residual information to described basic tomographic image piece of described basic tomographic image piece;
For according to the smoothness of described target image piece, determine described target up-sampling filter.
14. devices according to claim 12, it is characterized in that, described determining unit is further used for according to described the first training filter, described the second training filter and conventional filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter, described the second training filter and described conventional filter; And
Described coding unit is specifically for according to described information of forecasting, to the processing of encoding of described target image piece, to generate target code stream, described target code stream comprises the first indication information that is used to indicate described target up-sampling filter, described the first indication information in order in decoding during the target image piece after above-mentioned coding as the foundation of obtaining described target up-sampling filter.
15. devices according to claim 11, it is characterized in that, described determining unit is further used for according to described the first training filter and conventional filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter and described conventional filter;
Specifically for determining the texture degree of described target image piece;
For according to the texture degree of described target image piece, determine described target up-sampling filter.
16. 1 kinds of devices of processing for image, is characterized in that, described device comprises:
Acquiring unit, for the non-flat skating area area image piece with basic tomographic image according to the non-flat skating area area image piece of enhancement layer image, determine the first training filter, so that described the first training filter meets: the first information of forecasting of determining according to the non-flat skating area area image piece of described the first training filter and described basic tomographic image and similarity between the first raw information of determining according to the non-flat skating area area image piece of described enhancement layer image are satisfied first pre-conditioned, wherein, described basic tomographic image is corresponding with described enhancement layer image;
Determining unit, for according to described the first training filter, determines alternative up-sampling filter, and wherein, described alternative up-sampling filter comprises described the first training filter;
For from alternative up-sampling filter, determine target up-sampling filter, described alternative up-sampling filter comprises described the first training filter;
Decoding unit, for obtain described target up-sampling filter from described determining unit, and according to described target up-sampling filter and basic tomographic image piece, determines information of forecasting;
For according to described information of forecasting and the residual information that obtains from target code stream, to the processing of decoding of described object code stream, to obtain described target image piece, wherein, described target image piece is arranged in described enhancement layer image, described basic tomographic image piece is arranged in described basic tomographic image, and the locus of described primary image piece in described basic tomographic image is corresponding with the locus of described target image piece in described enhancement layer image.
17. devices according to claim 16, it is characterized in that, described acquiring unit is also for according to the smooth region image block of the smooth region image block of described enhancement layer image and described basic tomographic image, determine the second training filter, so that described the second training filter meets: the second information of forecasting of determining according to the smooth region image block of described the second training filter and described basic tomographic image and similarity between the second raw information of determining according to the smooth region image block of described enhancement layer image are satisfied second pre-conditioned, wherein, described basic tomographic image is corresponding with described enhancement layer image, and
Described determining unit is further used for training filter according to described the first training filter and described second, determines alternative up-sampling filter, and wherein, described alternative up-sampling filter comprises described the first training filter and described the second training filter.
18. devices according to claim 17, it is characterized in that, described determining unit is specifically for according to the characteristic information of described basic tomographic image piece, determine the smoothness of described target image, wherein, described characteristic information comprises encoding block label information or the residual information to described basic tomographic image piece of described basic tomographic image piece;
According to the smoothness of described target image piece, determine described target up-sampling filter.
19. devices according to claim 17, it is characterized in that, described determining unit is further used for according to described the first training filter, described the second training filter and conventional filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter, described the second training filter and described conventional filter;
For from described target code stream, obtain the first indication information that is used to indicate described target up-sampling filter,
For according to described the first indication information, determine described target up-sampling filter.
20. devices according to claim 16, it is characterized in that, described determining unit is further used for according to described the first training filter and conventional filter, determine alternative up-sampling filter, wherein, described alternative up-sampling filter comprises described the first training filter and described conventional filter;
For determining the texture degree of described target image piece;
For according to the texture degree of described target image piece, determine described target up-sampling filter.
CN201210374978.5A 2012-09-29 2012-09-29 Image processing method and device Active CN103716622B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201210374978.5A CN103716622B (en) 2012-09-29 2012-09-29 Image processing method and device
PCT/CN2013/084531 WO2014048374A1 (en) 2012-09-29 2013-09-27 Method and device for image processing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201210374978.5A CN103716622B (en) 2012-09-29 2012-09-29 Image processing method and device

Publications (2)

Publication Number Publication Date
CN103716622A true CN103716622A (en) 2014-04-09
CN103716622B CN103716622B (en) 2017-02-08

Family

ID=50387016

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201210374978.5A Active CN103716622B (en) 2012-09-29 2012-09-29 Image processing method and device

Country Status (2)

Country Link
CN (1) CN103716622B (en)
WO (1) WO2014048374A1 (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2016192054A1 (en) * 2015-06-03 2016-12-08 富士通株式会社 Image coding method and apparatus using prediction information and image processing device
CN107800450A (en) * 2017-11-13 2018-03-13 韩劝劝 Radio plays intensity control system
CN109257605A (en) * 2017-07-13 2019-01-22 华为技术有限公司 Image processing method, equipment and system

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070217502A1 (en) * 2006-01-10 2007-09-20 Nokia Corporation Switched filter up-sampling mechanism for scalable video coding
US20080007438A1 (en) * 2006-07-10 2008-01-10 Sharp Laboratories Of America, Inc. Methods and Systems for Signaling Multi-Layer Bitstream Data
CN101478671A (en) * 2008-01-02 2009-07-08 中兴通讯股份有限公司 Video encoding apparatus applied on video monitoring and video encoding method thereof
CN101895748A (en) * 2010-06-21 2010-11-24 华为终端有限公司 Coding and decoding methods and coding and decoding devices
CN102084653A (en) * 2007-06-29 2011-06-01 弗劳恩霍夫应用研究促进协会 Scalable video coding supporting pixel value refinement scalability
CN102355583A (en) * 2011-09-29 2012-02-15 广西大学 Scalable video encoding (SVC) block-level interlayer intra prediction (ILIP) method
CN102387366A (en) * 2005-03-18 2012-03-21 夏普株式会社 Methods and systems for extended spatial scalability with picture-level adaptation
CN102598660A (en) * 2009-04-20 2012-07-18 杜比实验室特许公司 Adaptive interpolation filters for multi-layered video delivery

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5039142B2 (en) * 2006-10-25 2012-10-03 フラウンホーファー−ゲゼルシャフト・ツール・フェルデルング・デル・アンゲヴァンテン・フォルシュング・アインゲトラーゲネル・フェライン Quality scalable coding method

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102387366A (en) * 2005-03-18 2012-03-21 夏普株式会社 Methods and systems for extended spatial scalability with picture-level adaptation
US20070217502A1 (en) * 2006-01-10 2007-09-20 Nokia Corporation Switched filter up-sampling mechanism for scalable video coding
US20080007438A1 (en) * 2006-07-10 2008-01-10 Sharp Laboratories Of America, Inc. Methods and Systems for Signaling Multi-Layer Bitstream Data
CN102084653A (en) * 2007-06-29 2011-06-01 弗劳恩霍夫应用研究促进协会 Scalable video coding supporting pixel value refinement scalability
CN101478671A (en) * 2008-01-02 2009-07-08 中兴通讯股份有限公司 Video encoding apparatus applied on video monitoring and video encoding method thereof
CN102598660A (en) * 2009-04-20 2012-07-18 杜比实验室特许公司 Adaptive interpolation filters for multi-layered video delivery
CN101895748A (en) * 2010-06-21 2010-11-24 华为终端有限公司 Coding and decoding methods and coding and decoding devices
CN102355583A (en) * 2011-09-29 2012-02-15 广西大学 Scalable video encoding (SVC) block-level interlayer intra prediction (ILIP) method

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2016192054A1 (en) * 2015-06-03 2016-12-08 富士通株式会社 Image coding method and apparatus using prediction information and image processing device
CN109257605A (en) * 2017-07-13 2019-01-22 华为技术有限公司 Image processing method, equipment and system
CN109257605B (en) * 2017-07-13 2021-11-19 华为技术有限公司 Image processing method, device and system
CN107800450A (en) * 2017-11-13 2018-03-13 韩劝劝 Radio plays intensity control system

Also Published As

Publication number Publication date
CN103716622B (en) 2017-02-08
WO2014048374A1 (en) 2014-04-03

Similar Documents

Publication Publication Date Title
CN103220527B (en) Method for encoding images and device and its coding/decoding method and device
CN102970526B (en) A kind of method obtaining transform block size and module
CN104994386A (en) Method and apparatus for encoding and decoding image through intra prediction
CN109660800B (en) Motion estimation method, motion estimation device, electronic equipment and computer-readable storage medium
CN112929666B (en) Method, device and equipment for training coding and decoding network and storage medium
CN103081463A (en) Apparatus for encoding and decoding image by skip encoding and method for same
CN102939752A (en) Method and apparatus for encoding video by performing in-loop filtering based on tree-structured data unit, and method and apparatus for decoding video by performing the same
CN109688407B (en) Reference block selection method and device for coding unit, electronic equipment and storage medium
CN102300094B (en) Video coding method
CN102324037A (en) Shot boundary detection method based on support vector machine and genetic algorithm
CN103313047A (en) Video coding method and apparatus
CN102685505B (en) Intra-frame prediction method and prediction device
CN105120290A (en) Fast coding method for depth video
CN104243992A (en) Image encoding apparatus and image encoding method
CN103931181A (en) Adaptive upsampling for spatially scalable video coding
JP6065613B2 (en) Video encoding device
CN103716622A (en) Image processing method and device
CN103716629A (en) Image processing method, device, coder and decoder
KR101845622B1 (en) Adaptive rdpcm method for video coding, video encoding method based on adaptive rdpcm and video decoding method based on adaptive rdpcm
CN110381312B (en) HEVC (high efficiency video coding) -based depth partition range prediction method and device
CN103716630A (en) Upsampling filter generation method and device
JP2009212605A (en) Information processing method, information processor, and program
CN103716631A (en) Image processing method, device, coder and decoder
WO2019199513A1 (en) A method and apparatus for encoding and decoding metadata associated with patched projection of point clouds
CN104104966A (en) Video block dividing method and device

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