Critical sample filtering

Abstract

Video coding uses critical sample values to adjust reconstructed pixel values to improve a reconstructed image portion. A critical sample value is determined for a region of a frame of a video sequence. The region has multiple samples. The critical sample value is a value of a sample from the multiple samples that has a cardinality within the region that is above a critical sample threshold. After reconstructing the region of the frame to form a reconstructed region, a value of a reconstructed sample within the reconstructed region is modified to the critical sample value. At an encoder, the critical sample value may be transmitted to a decoder within a header that applies to the region.

Description

CRITICAL SAMPLE FILTERINGBACKGROUND

[0001] Digital video streams may represent video using a sequence of frames or still images. Digital video can be used for various applications including, for example, video conferencing, high-definition video entertainment, video advertisements, or sharing of usergenerated videos. A digital video stream can contain a large amount of data and consume a significant amount of computing or communication resources of a computing device for processing, transmission, or storage of the video data. Various approaches have been proposed to reduce the amount of data in video streams, including lossy and lossless coding techniques.SUMMARY

[0002] An aspect of the teachings herein is a method for using critical sample filtering in video coding. A critical sample value is determined for a region of a frame of a video sequence. The region of the frame of the video sequence includes multiple samples (e.g., pixels). The critical sample value is a value of a sample from the multiple samples that has a cardinality within the region that is above a first threshold. The region of the frame is reconstructed to form a reconstructed region, after which a value of a reconstructed sample (e.g., a reconstructed pixel) within the reconstructed region is modified to the critical sample value.

[0003] In some implementations, the method includes determining a critical sample range for the reconstructed region, the critical sample range comprising an upper range bound and a lower range bound relative to the critical sample value, and modifying the value of the reconstructed sample to the critical sample value is based on the critical sample range. In a variation, the header information includes an integer, the upper range bound of the critical sample range relative to a respective critical sample value from the multiple critical sample values is defined as the respective critical sample value plus the integer, and the lower range bound of the critical sample range relative the respective critical sample value is defined as a greater of the respective critical sample value minus the integer and an immediately preceding critical sample value plus one.

[0004] In some implementations, the reconstructed region includes multiple reconstructed samples, and the method includes determining the value for each of the multiple reconstructed samples within a sample area of the reconstructed region, wherein the sample area comprises the reconstructed sample and secondary reconstructed samples proximate to the reconstructed sample, and determining a cardinality of the multiple reconstructed samples within the sample area that have the value between the upper range bound and the lower range bound of the critical sample range, inclusive.

[0005] In some implementations, modifying the value of the reconstructed sample to the critical sample value is responsive to determining that the cardinality of the multiple reconstructed samples within the sample area that have the value between the upper range bound and the lower range bound of the critical sample range, inclusive, is above a sample area threshold. In a variation, the sample area includes at least one of a cross shape comprising five of the multiple reconstructed samples, a square shape comprising nine of the multiple reconstructed samples, or a diamond shape comprising thirteen of the multiple reconstructed samples, and the reconstructed sample is at a center of each of the at least one of the cross shape, the square shape, or the diamond shape. In a variation, the sample area threshold is equal to a cardinality of the multiple reconstructed samples within the region.

[0006]

[0007] In some implementations, the critical sample range and the critical sample value are determined at least in part by header information within an encoded bitstream.

[0008] In some implementations, the method includes determining, from the header information, a sequence for modifying the value of the reconstructed sample to the critical sample value and applying at least one of a deblocking filter, a sample offset, or an adaptive loop filter to the region.

[0009] In some implementations, the critical sample value is one of multiple critical sample values, and the multiple critical sample values are signaled in the header information in order of magnitude. In a variation, the multiple critical sample values are signaled in the header information by signaling a difference with respect to an immediately preceding critical sample value from the multiple critical sample values.

[0010] In some implementations, the critical sample value corresponds to a plane of color data of the region. In a variation, the critical sample range applies to all planes of color data of the region.

[0011] An aspect of the teachings herein is another method for using critical sample filtering in video coding. A critical sample value in a region of a frame of a video sequence isidentified. The region of the frame of the video sequence includes multiple samples (e.g., pixels). The critical sample value is a value of a sample from the multiple samples that has a cardinality within the region that is above a threshold. The critical sample value is encoded into an encoded bitstream.

[0012] In some implementations, the method includes encoding information relating to a critical sample range in the encoded bitstream, the critical sample range including an upper range bound and a lower range bound relative to the critical sample value. In a variation, the information relating to the critical sample range identifies an integer, and at least one of the upper range bound or the lower range bound is based on the integer. In a variation, the critical sample range applies to all planes of color data of the region.

[0013] In some implementations, the critical sample value is one of multiple critical sample values, and the method includes identifying a cardinality of the multiple critical sample values in the region, and encoding the multiple critical sample values and cardinality of the multiple critical sample values in the region in the encoded bitstream. In a variation, the multiple critical sample values are encoded in the encoded bitstream in order of magnitude. In a variation, the multiple critical sample values are encoded in the bitstream by encoding a difference with respect to an immediately preceding critical sample value from the multiple critical sample values.

[0014] An aspect of the teachings herein is a system or apparatus that can perform any of these methods and others described herein.

[0015] An aspect of the teachings herein is a non-transitory, computer-readable storage medium storing instructions that, when executed, can cause a processor to perform any of these methods and others described herein.

[0016] An aspect of the teaching herein is a non-transitory, computer-readable storage medium storing an encoded bitstream. The encoded bitstream includes compressed image data that corresponds to a region of a frame of a video sequence. The encoded bitstream also includes a critical sample value for the region. The region includes multiple samples (e.g., pixels). The critical sample value is a value of a sample from the multiple samples that has a cardinality within the region that is above a threshold.

[0017] In some implementations, the critical sample value is one of multiple critical sample values for the region, and the first header information includes the multiple critical sample values.

[0018] In some implementations, the encoded bitstream includes second header information including a cardinality of the multiple critical sample values for the region.

[0019] In some implementations, the encoded bitstream includes third header information including an integer that defines at least one of an upper range bound or a lower range bound for a critical sample range for the region, and upper range bound and the lower range bound are relative to the multiple critical sample values. In a variation, the third header information is signaled in the encoded bitstream after the first header information.

[0020] In some implementations, the first header information is signaled in the encoded bitstream after the second header information.

[0021] In some implementations, the encoded bitstream includes fourth header information that indicates a sequence for modifying the value of a reconstructed sample to the critical sample value and applying at least one of a deblocking filter, a sample offset, or an adaptive loop filter to the region.

[0022] In some implementations, the multiple critical sample values are signaled in the first header information in order of magnitude, and the multiple critical sample values are signaled in the first header information by signaling a difference with respect to an immediately preceding critical sample value from the multiple critical sample values.

[0023] These and other aspects of the present disclosure are disclosed in the following detailed description of the embodiments, the appended claims, and the accompanying figures.BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The description herein makes reference to the accompanying drawings described below, wherein like reference numerals refer to like parts throughout the several views.

[0025] FIG. 1 is a schematic of a video encoding and decoding system.

[0026] FIG. 2 is a block diagram of an example of a computing device that can implement a transmitting station or a receiving station.

[0027] FIG. 3 is a diagram of a typical video stream to be encoded and subsequently decoded.

[0028] FIG. 4 is a block diagram of an encoder according to implementations of this disclosure.

[0029] FIG. 5 is a block diagram of a decoder according to implementations of this disclosure.

[0030] FIG. 6 is a flowchart diagram of a technique for video coding using adaptive range clipping.

[0031] FIG. 7 is a block diagram of a reconstruction path in which the techniques described herein may be implemented.

[0032] FIG. 8 illustrates an example of implementing critical sample filtering according to the techniques described herein.

[0033] FIG. 9 is a flowchart diagram of a technique for video coding using critical sample filtering.

[0034] FIG. 10 is a flowchart diagram of another technique for video coding using critical sample filtering.

[0035] FIG. 11 is a block diagram of another decoder in which critical sample filtering may be implemented.DETAILED DESCRIPTION

[0036] Video compression schemes may include breaking respective images, or frames, of a video stream into smaller portions, such as coding tree blocks (CTBs) or coding tree units (CTUs) (sometimes referred to as superblocks), and generating an encoded bitstream using techniques to limit the information included for respective CTUs thereof. The bitstream can be decoded to re-create the source frames from the limited information. Encoding CTUs to or decoding CTUs from the bitstream can include predicting the values of pixels or CTUs based on similarities with other pixels or CTUs in the same frame or in one or more other frames that have already been coded.

[0037] Those similarities can be determined using intra prediction, which attempts to predict the pixel values of a coding unit (CU) (coding block, etc.) of a CTU using pixels peripheral to the CU (e.g., pixels that are in the same frame as the CU, but that are outside the CU). During encoding, the result of an intra-prediction mode performed against a CU is a prediction unit (PU) (prediction block, predictor block, etc.). A prediction residual can be determined based on a difference between the pixel values of the CU and the pixel values of the PU. The prediction residual and the intra-prediction mode used to ultimately obtain that prediction residual can then be encoded to a bitstream. During decoding, the prediction residual is reconstructed into a CU using a PU produced based on the intra-prediction mode and is thereafter included in an output video stream.

[0038] Similarly, inter prediction attempts to predict the pixel values of a CU of a CTU using pixels from one or more reference frames. During encoding, the result of an interprediction mode performed against a CU is also a PU. A prediction residual can be determined based on a difference between the pixel values of the CU and the pixel values of the PU. The prediction residual and the inter-prediction mode used to ultimately obtain that prediction residual can then be encoded to a bitstream. During decoding, the predictionresidual is reconstructed into a CU using a PU produced based on the inter-prediction mode and is thereafter included in an output video stream.

[0039] A frame, and hence its CTUs and CUs may include a luminance, also referred to as luma, component and two chrominance, also referred to as chroma, components. These luma and chroma components may in some cases be referred to as luma blocks and chroma blocks. The luma component may, for example, be expressed within a Y plane and the chroma components may be expressed either within U and V planes or Cr and Cb planes. The luma component is understood to include some number of luma samples and each chroma component is understood to include some number of chroma samples. Generally, the luma samples provide measures of brightness throughout a frame and thus represent the structural qualities of the video content of the frame, whereas the chroma samples provide measures of color throughout the frame. The number of luma samples can indicate the spatial resolution (or simply the resolution) of the frame.

[0040] A reconstructed frame (image, picture) may include distortion in one or more layers comprising luma components or chroma components that are introduced by lossy encoding techniques. For example, quantization of pixel values of a prediction residual may cause pixel values (e.g., values of Y, U, V components) of a resulting reconstructed CU to deviate from pixel values of the original CU. The deviations may result in distortion such as blocking, ringing, and / or other artifacts in a reconstructed frame.

[0041] To reduce distortion, different filtering techniques may be used with the reconstructed pixel values. In some pictures (e.g., a frame or image), there may be critical luma and / or chroma samples, e.g., pixels having values that appear many times in the picture. In some implementations, the critical sample values can be pixel values that result in relatively large mismatch between original and reconstructed values. For example, distortions in a reconstructed frame may be more apparent where an original frame includes sharp edges between contrasting pixel values such as in a screencast (e.g., an original frame that depicts black text on a white background or vice versa). To reduce distortion, critical sample filtering may be performed on the reconstructed frame or one or more portions thereof to correct the coding error introduced to critical samples during compression.

[0042] Critical sample filtering may reduce distortion by identifying one or more pixel values in the original frame, or portion of the original frame, that appear in the frame or portion more frequently than other pixel values. These pixel values may be referred to as critical sample values. One or more pixel values of the reconstructed frame, or portion of the reconstructed frame (e.g., the reconstructed CU, CTU, etc.), may be modified to the one ormore critical sample values based on other (e.g., surrounding) pixel values. Doing so may reduce distortion in the reconstructed frame.

[0043] Implementations of this disclosure describe critical sample filtering that reduces distortion in a reconstructed frame or portions thereof. Further details of techniques for video coding using critical sample filtering are described herein with initial reference to a system in which the disclosure may be implemented.

[0044] FIG. 1 is a schematic of a video encoding and decoding system 100. A transmitting station 102 can be, for example, a computer having an internal configuration of hardware such as that described in FIG. 2. However, other implementations of the transmitting station 102 are possible. For example, the processing of the transmitting station 102 can be distributed among multiple devices.

[0045] A network 104 can connect the transmitting station 102 and a receiving station 106 for encoding and decoding of the video stream. Specifically, the video stream can be encoded in the transmitting station 102, and the encoded video stream can be decoded in the receiving station 106. The network 104 can be, for example, the Internet. The network 104 can also be a local area network (LAN), wide area network (WAN), virtual private network (VPN), cellular telephone network, or any other means of transferring the video stream from the transmitting station 102 to, in this example, the receiving station 106.

[0046] The receiving station 106, in one example, can be a computer having an internal configuration of hardware such as that described in FIG. 2. However, other suitable implementations of the receiving station 106 are possible. For example, the processing of the receiving station 106 can be distributed among multiple devices.

[0047] Other implementations of the video encoding and decoding system 100 are possible. For example, an implementation can omit the network 104. In another implementation, a video stream can be encoded and then stored for transmission at a later time to the receiving station 106 or any other device having memory. In one implementation, the receiving station 106 receives (e.g., via the network 104, a computer bus, and / or some communication pathway) the encoded video stream and stores the video stream for later decoding. In an example implementation, a real-time transport protocol (RTP) is used for transmission of the encoded video over the network 104. In another implementation, a transport protocol other than RTP may be used, e.g., video streaming protocol based on the Hypertext Transfer Protocol (HTTP).

[0048] When used in a video conferencing system, for example, the transmitting station 102 and / or the receiving station 106 may include the ability to both encode and decode avideo stream as described below. For example, the receiving station 106 could be a video conference participant who receives an encoded video bitstream from a video conference server (e.g., the transmitting station 102) to decode and view and further encodes and transmits his or her own video bitstream to the video conference server for decoding and viewing by other participants.

[0049] FIG. 2 is a block diagram of an example of a computing device 200 that can implement a transmitting station or a receiving station. For example, the computing device 200 can implement one or both of the transmitting station 102 and the receiving station 106 of FIG. 1. The computing device 200 can be in the form of a computing system including multiple computing devices, or in the form of one computing device, for example, a mobile phone, a tablet computer, a laptop computer, a notebook computer, a desktop computer, and the like.

[0050] A processor 202 in the computing device 200 can be a conventional central processing unit. Alternatively, the processor 202 can be another type of device, or multiple devices, capable of manipulating or processing information now existing or hereafter developed. For example, although the disclosed implementations can be practiced with one processor as shown (e.g., the processor 202), advantages in speed and efficiency can be achieved by using more than one processor.

[0051] A memory 204 in computing device 200 can be a read only memory (ROM) device or a random access memory (RAM) device in an implementation. However, other suitable types of storage device can be used as the memory 204. The memory 204 can include code and data 206 that is accessed by the processor 202 using a bus 212. The memory 204 can further include an operating system 208 and application programs 210, the application programs 210 including at least one program that permits the processor 202 to perform the techniques described herein. For example, the application programs 210 can include applications 1 through N, which further include a video coding application that performs the techniques described herein. The computing device 200 can also include a secondary storage 214, which can, for example, be a memory card used with a mobile computing device.Because the video communication sessions may contain a significant amount of information, they can be stored in whole or in part in the secondary storage 214 and loaded into the memory 204 as needed for processing.

[0052] The computing device 200 can also include one or more output devices, such as a display 218. The display 218 may be, in one example, a touch sensitive display that combines a display with a touch sensitive element that is operable to sense touch inputs. The display218 can be coupled to the processor 202 via the bus 212. Other output devices that permit a user to program or otherwise use the computing device 200 can be provided in addition to or as an alternative to the display 218. When the output device is or includes a display, the display can be implemented in various ways, including by a liquid crystal display (LCD), a cathode-ray tube (CRT) display, or a light emitting diode (LED) display, such as an organic LED (OLED) display.

[0053] The computing device 200 can also include or be in communication with an image-sensing device 220, for example, a camera, or any other image-sensing device 220 now existing or hereafter developed that can sense an image such as the image of a user operating the computing device 200. The image-sensing device 220 can be positioned such that it is directed toward the user operating the computing device 200. In an example, the position and optical axis of the image-sensing device 220 can be configured such that the field of vision includes an area that is directly adjacent to the display 218 and from which the display 218 is visible.

[0054] The computing device 200 can also include or be in communication with a soundsensing device 222, for example, a microphone, or any other sound-sensing device now existing or hereafter developed that can sense sounds near the computing device 200. The sound-sensing device 222 can be positioned such that it is directed toward the user operating the computing device 200 and can be configured to receive sounds, for example, speech or other utterances, made by the user while the user operates the computing device 200.

[0055] Although FIG. 2 depicts the processor 202 and the memory 204 of the computing device 200 as being integrated into one unit, other configurations can be utilized. The operations of the processor 202 can be distributed across multiple machines (wherein individual machines can have one or more processors) that can be coupled directly or across a local area or other network. The memory 204 can be distributed across multiple machines such as a network-based memory or memory in multiple machines performing the operations of the computing device 200. Although depicted here as one bus, the bus 212 of the computing device 200 can be composed of multiple buses. Further, the secondary storage 214 can be directly coupled to the other components of the computing device 200 or can be accessed via a network and can comprise an integrated unit such as a memory card or multiple units such as multiple memory cards. The computing device 200 can thus be implemented in a wide variety of configurations.

[0056] FIG. 3 is a diagram of an example of a video stream 300 to be encoded and subsequently decoded. The video stream 300 includes a video sequence 302. At the nextlevel, the video sequence 302 includes a number of adjacent frames 304. While three frames are depicted as the adjacent frames 304, the video sequence 302 can include any number of adjacent frames 304. The adjacent frames 304 can then be further subdivided into individual frames, for example, a frame 306. At the next level, the frame 306 can be divided into a series of planes or segments 308. The segments 308 can be subsets of frames that permit parallel processing, for example. The segments 308 can also be subsets of frames that can separate the video data into separate colors. For example, a frame 306 of color video data can include a luminance plane and two chrominance planes. The segments 308 may be sampled at different resolutions.

[0057] Whether or not the frame 306 is divided into segments 308, the frame 306 may be further subdivided into blocks 310, which can contain data corresponding to, for example, 16x16 pixels in the frame 306. The blocks 310 can also be arranged to include data from one or more segments 308 of pixel data. The blocks 310 can also be of any other suitable size such as 4x4 pixels, 8x8 pixels, 16x8 pixels, 8x16 pixels, 16x16 pixels, or larger. Unless otherwise noted, the terms block and macroblock are used interchangeably herein.

[0058] FIG. 4 is a block diagram of an encoder 400 according to implementations of this disclosure. The encoder 400 can be implemented, as described above, in the transmitting station 102, such as by providing a computer software program stored in memory, for example, the memory 204. The computer software program can include machine instructions that, when executed by a processor such as the processor 202, cause the transmitting station 102 to encode video data in the manner described in FIG. 4. The encoder 400 can also be implemented as specialized hardware included in, for example, the transmitting station 102. In one particularly desirable implementation, the encoder 400 is a hardware encoder.

[0059] The encoder 400 has the following stages to perform the various functions in a forward path (shown by the solid connection lines) to produce an encoded or compressed bitstream 420 using the video stream 300 as input: an intra / inter prediction stage 402, a transform stage 404, a quantization stage 406, and an entropy encoding stage 408. The encoder 400 may also include a reconstruction path (shown by the dotted connection lines) to reconstruct a frame for encoding of future blocks. In FIG. 4, the encoder 400 has the following stages to perform the various functions in the reconstruction path: a dequantization stage 410, an inverse transform stage 412, a reconstruction stage 414, and a loop filtering stage 416. Other structural variations of the encoder 400 can be used to encode the video stream 300.

[0060] When the video stream 300 is presented for encoding, respective adjacent frames 304, such as the frame 306, can be processed in units of blocks. At the intra / inter prediction stage 402, respective blocks can be encoded using intra-frame prediction (also called intra prediction) or inter- frame prediction (also called inter prediction). In any case, a prediction block can be formed. In the case of intra prediction, a prediction block may be formed from samples in the current frame that have been previously encoded and reconstructed. In the case of inter prediction, a prediction block may be formed from samples in one or more previously constructed reference frames.

[0061] Next, the prediction block can be subtracted from the current block at the intra / inter prediction stage 402 to produce a residual block (also called a residual). The transform stage 404 transforms the residual into transform coefficients in, for example, the frequency domain using block-based transforms. The quantization stage 406 converts the transform coefficients into discrete quantum values, which are referred to as quantized transform coefficients, using a quantizer value or a quantization level. For example, the transform coefficients may be divided by the quantizer value and truncated.

[0062] The quantized transform coefficients are then entropy encoded by the entropy encoding stage 408. The entropy-encoded coefficients, together with other information used to decode the block (which may include, for example, syntax elements such as used to indicate the type of prediction used, transform type, motion vectors, a quantizer value, or the like), are then output to the compressed bitstream 420. The compressed bitstream 420 can be formatted using various techniques, such as variable length coding (VLC) or arithmetic coding. The compressed bitstream 420 can also be referred to as an encoded video stream or encoded video bitstream, and the terms will be used interchangeably herein.

[0063] The reconstruction path (shown by the dotted connection lines) can be used to ensure that the encoder 400 and a decoder 500 (described below with respect to FIG. 5) use the same reference frames to decode the compressed bitstream 420. The reconstruction path performs functions that are like functions that take place during the decoding process (described below with respect to FIG. 5), including dequantizing the quantized transform coefficients at the dequantization stage 410 and inverse transforming the dequantized transform coefficients at the inverse transform stage 412 to produce a derivative residual block (also called a derivative residual). At the reconstruction stage 414, the prediction block that was predicted at the intra / inter prediction stage 402 can be added to the derivative residual to create a reconstructed block. The loop filtering stage 416 can be applied to the reconstructed block to reduce distortion such as blocking artifacts.

[0064] Other variations of the encoder 400 can be used to encode the compressed bitstream 420. In some implementations, a non-transform based encoder can quantize the residual signal directly without the transform stage 404 for certain blocks or frames. In some implementations, an encoder can have the quantization stage 406 and the dequantization stage 410 combined in a common stage.

[0065] FIG. 5 is a block diagram of a decoder 500 according to implementations of this disclosure. The decoder 500 can be implemented in the receiving station 106, for example, by providing a computer software program stored in the memory 204. The computer software program can include machine instructions that, when executed by a processor such as the processor 202, cause the receiving station 106 to decode video data in the manner described in FIG. 5. The decoder 500 can also be implemented in hardware included in, for example, the transmitting station 102 or the receiving station 106.

[0066] The decoder 500, similar to the reconstruction path of the encoder 400 discussed above, includes in one example the following stages to perform various functions to produce an output video stream 516 from the compressed bitstream 420: an entropy decoding stage 502, a dequantization stage 504, an inverse transform stage 506, an intra / inter prediction stage 508, a reconstruction stage 510, a loop filtering stage 512, and a post filtering stage 514. Other structural variations of the decoder 500 can be used to decode the compressed bitstream 420.

[0067] When the compressed bitstream 420 is presented for decoding, the data elements within the compressed bitstream 420 can be decoded by the entropy decoding stage 502 to produce a set of quantized transform coefficients. The dequantization stage 504 dequantizes the quantized transform coefficients (e.g., by multiplying the quantized transform coefficients by the quantizer value), and the inverse transform stage 506 inverse transforms the dequantized transform coefficients to produce a derivative residual that can be identical to that created by the inverse transform stage 412 in the encoder 400. Using header information decoded from the compressed bitstream 420, the decoder 500 can use the intra / inter prediction stage 508 to create the same prediction block as was created in the encoder 400 (e.g., at the intra / inter prediction stage 402).

[0068] At the reconstruction stage 510, the prediction block can be added to the derivative residual to create a reconstructed block. The loop filtering stage 512 can be applied to the reconstructed block to reduce blocking artifacts. Other filtering can be applied to the reconstructed block. In this example, the post filtering stage 514 is applied to the reconstructed block to reduce blocking distortion, and the result is output as the output videostream 516. The output video stream 516 can also be referred to as a decoded video stream, and the terms will be used interchangeably herein. Other variations of the decoder 500 can be used to decode the compressed bitstream 420. In some implementations, the decoder 500 can produce the output video stream 516 without the post filtering stage 514 or otherwise omit the post filtering stage 514.

[0069] As mentioned initially, adaptive clipping may be used to keep pixel values within a range of values. The value ranges for luma and chroma color components may be separately signaled at the picture level and applied at different points in the reconstruction process at an encoder and decoder. For example, the different points may be after prediction at the intra / inter prediction stage 402, 508, reconstruction stage 414, 510, and loop filtering stage 416, 512. In some implementations of adaptive clipping, only a value range for the luma components is signaled, and the clipping may be applied at the same points or at different points such as before and after the loop filtering stage 416, 512. For example, in a video signal that conforms to the International Telecommunication Union Radiocommunication Sector (ITU-R) BT.2020 specification, only luma code values between 64 to 940 are allowed for a 10-bit narrow-range video. In either case (value ranges for luma only or for all color components), the signaling cost is high, especially because clipping is not needed in all situations.

[0070] FIG. 6 is a flowchart of a technique 600 for video coding using adaptive range clipping. The technique 600 can be implemented, for example, as a software program that may be executed by computing devices such as transmitting station 102 or receiving station 106. The software program can include machine-readable instructions that may be stored in a memory such as the memory 204 or the secondary storage 214, and that, when executed by a processor, such as the processor 202, may cause the computing device to perform the technique 600. The technique 600 may be implemented at least in part in the reconstruction stage of an encoder and / or the reconstruction path of a decoder. The technique 600 can be implemented using specialized hardware or firmware. Multiple processors, memories, or both, may be used.

[0071] At operation 602, range bound information is determined for at least one plane of color data. The range bound information may be a minimum value, a maximum value, or both, for one or more planes of color data. For example, the range bound information may be a minimum value (also referred to as a lower bound) and a maximum value (also referred to as an upper bound) for pixel values of a luma plane, a minimum value and a maximum value for pixels values of one or both chroma planes, or both. In some examples, the range boundinformation may include a minimum value or a maximum value for pixel values of a luma plane, a minimum value or a maximum value for pixels values of one or both chroma planes, or both, and the other value of the range may be inferred. For example, if a maximum value is transmitted, the minimum value may be inferred to be 0. Other examples are possible.

[0072] At a decoder, the range bound information is determined from a first header for multiple frames at operation 602. That is, the range bound information may be decoded from the first header (e.g., by entropy decoding). The first header for multiple frames is a header that applies to or is available to multiple frames. Stated differently, the first header includes coding parameters and other information that was used to encode and can be used to decode at least some frames of the multiple frames. The multiple frames can be any grouping of frames. For example, the frames may comprise a group of pictures (GOP). In some implementations, the first header is signaled with a sequence parameter set (SPS), an adaptive parameter set (APS), or a picture parameter set (PPS).

[0073] At operation 604, the technique 600 determines, from respective second headers of at least a portion of a frame of the multiple frames, a value of a clip flag that indicates whether to apply the range bound information to the at least the portion of the frame. That is, a second header is a header that applies to or is available to at least a portion of a frame of the multiple frames. For example, the second header may be a frame header of a frame of the multiple frames such that the coding parameters of the frame header (such as the range bound information) apply to or are available to the entire frame. The second headers may be frame headers for each frame of the multiple frames. In this way, a clip flag may be included for each frame such that the value of the clip flag indicates whether the corresponding frame was encoded and should be decoded using the range bound information. The second headers may be below the frame level. For example, a respective second header may be included in a slice header or a tile header such that signaling a frame of the multiple frames includes multiple second headers and hence multiple clip flags. In other implementations, a clip flag may be signaled at a region (a group of blocks) level, such as at coding tree unit (CTU) or super block level (such as respective blocks of 128 x 128 pixels that are partitioned into coding blocks for prediction), the level of 2x2 CTUs / super block levels, a CTU (or super block) row, a CTU (or super block) column, and so on.

[0074] At a decoder, the value of a clip flag may be determined from respective second headers within the encoded bitstream at operation 604. That is, the value of a clip flag may be decoded from a second header (e.g., by entropy decoding). The value indicates whether theframe portion (blocks of a frame, a slice, a tile, etc.) to which the clip flag applies was encoded and should be decoded using the range bound information.

[0075] In some implementations, the value of a clip flag may be inferred instead of signaled. For example, where the clip flag is a binary flag, a value of 1 determined from the encoded bitstream may indicate to apply the range bound information to the corresponding portion of a frame, and the omission of the clip flag indicates the value is 0 such that the range bound information is not applied.

[0076] At operation 606, the multiple frames are selectively decoded using the range bound information from the first header and the value of the clip flag from the respective second headers. For example, this may include decoding a respective frame on a block-by- block basis according to the techniques described regarding FIG. 5, while performing clipping before and / or after one or more stages of a reconstruction path as described previously. Selectively decoding thus means that clipping may be performed on some but not all frames of the multiple frames and on at least some regions of one or more frames as indicated by the value of respective clip flags. The multiple frames may be selectively decoded in a decoding order.

[0077] At an encoder, the techniques used to select the range bound information, if any, for sets of multiple frames of a video sequence and to select what portion(s), if any, of each frame of the multiple frames to assign a clip flag are not particularly limited. For example, this information may be decided arbitrarily or may be based on a particular standard / specification. Additionally, or alternatively, techniques that code and reconstruct frames according to different parameters, such as different values for the range bound information and clip flags, can be used and the results compared to determine the best set of parameters (e.g., the set of parameters that results in the lowest distortion, the fewest bits, or some combination thereof). For example, rate-distortion calculations can be used.

[0078] Regardless of how the range bound information and the values of the clip flags are selected, the encoder can encode, into a first header of an encoded bitstream, the range bound information. As described previously, the first header is a header that applies to multiple frames of a video sequence, and the range bound information comprises at least one of an upper bound or a lower bound for values of at least one plane of color data. The encoder can also encode, into respective second headers of the encoded bitstream, a value of a clip flag. As also described above, a second header is a header that applies to at least a portion of a frame of the multiple frames, and the value of the clip flag indicates whether to apply the range bound information to the at least the portion of the frame. Finally, the encoderselectively encodes the multiple frames using the range bound information of the first header and the value of the clip flag of the respective second headers. In an example where the output video should conform to the ITU-R BT.2020 specification, range bound information comprising a lower bound of 64 and an upper bound of 940 may be signaled for the luma plane in the first header of sets of multiple frames forming a video sequence, and the clip flag of all frames indicates to apply the range bound information.

[0079] As mentioned, the method, process, or technique 600 of FIG. 6 may be implemented in the reconstruction path of an encoder, such as the encoder 400, or a decoder, such as the decoder 500. FIG. 7 is a block diagram of another reconstruction path in which the techniques described herein, including the technique 600, may be implemented. The techniques can occur at a reconstruction path of both an encoder and a decoder, so FIG. 7 represents a portion of either an encoder or decoder. The structure of FIG. 7 shows a modification to the reconstruction paths of the encoder 400 and decoder 500 that separates a luma reconstruction path 700 from a chroma reconstruction path 702 to functionally illustrate use of the teachings herein with luma mapping and chroma sampling (LMCS). Luma mapping (LM) may performed separately or with chroma scaling (CS).

[0080] LM remaps the luma code values and CS allows flexible adjustment between luma and chroma signals. LM aims to improve coding efficiency by reallocating the luma code values of the input video signal within the complete codeword range. As in the example of a video signal that conforms to the ITU-R BT.2020 specification, only luma code values between 64 to 940 are allowed for a 10-bit narrow-range video. This inefficient codeword utilization can be addressed by remapping, which allows for coding performance improvements. CS, when used, is intended to re-balance the impact of luma remapping on the relative luma / chroma coding bit costs. The flexible adjustment between luma and chroma signals can be achieved by enabling or disabling CS at, e.g., the sequence or picture level, or further adjusting chroma scaling by applying a chroma scaling offset (deltaCRS), which can be signaled in a header.

[0081] LM maps the luma code values (or simply, luma values) of an input video signal from an original (unmapped) sample domain to a mapped sample domain. As shown in the FIG. 7, the processes in the mapped sample domain (gray-shaded blocks) include inverse quantization, inverse transform, luma intra prediction, and reconstruction (e.g., of a luma block) by summing the luma prediction values with the luma residual values. The processes in the original sample domain include in-loop filters (as described below), inter prediction, chroma intra prediction, reconstruction (e.g., of a chroma block) by summing chromaprediction values with the chroma residual values, and storage of pictures in a decoded picture buffer (DPB). Forward luma mapping maps the luma code values from the original sample domain to the mapped sample domain, and inverse luma mapping maps the luma code values from the mapped sample domain back to the original sample domain.

[0082] In the luma reconstruction path 700, the following steps are performed for LM. First, inverse quantization and inverse transform are applied to the decoded luma transform coefficients (e.g., obtained by entropy decoding) to produce the luma residues (or residuals) in the mapped sample domain, Y'reS. The processes of inverse quantization and inverse transform may be performed as described above with regards to the dequantization stage 410, 504 and the inverse transform stage 412, 506, respectively.

[0083] Thereafter, reconstructed luma sample values in the mapped sample domain, Y'r, are obtained by summing Y'reSwith the corresponding predicted luma values in the mapped sample domain, Y'pred. When intra mode is signaled for prediction of a portion (e.g., a block) of the frame, Y'pred is directly obtained by performing intra prediction in the mapped sample domain. In contrast, when inter mode is signaled for prediction of the portion of the frame, the predicted luma values in the original sample domain, Ypred, are first obtained by motion compensation using reference pictures from the DPB, and then forward luma mapping (discussed in more detail below) is applied to produce the luma values in the mapped sample domain, Y'pred. These prediction processes may be performed at a prediction stage like the intra / inter prediction stage 402, 508 described above.

[0084] The reconstructed values, which result from adding Y'pred and Y'reS, are then inverse-mapped (e.g., using inverse luma mapping as discussed in more detail below) and processed by loop filtering (using in-loop filters) before being stored in the DPB in the original sample domain. In-loop filters are discussed in more detail below. The stored picture is then used for inter prediction of one or more frames after the frame in the coding sequence.

[0085] CS, when performed, includes the following steps performed in the chroma reconstruction path 702. First, the inverse quantization and inverse transform processes are applied to the decoded chroma transform coefficients to produce chroma residue-scaled values, CresScaie. Chroma residue (or residual) values, Cres, are obtained by multiplying CresScaie by the inverse scaling factor invScaleC. The inverse scaling factor invScaleC may be determined based on a chroma scaling offset deltaCRS (e.g., a signaled value) and an average reconstructed luma value, avgY', from one or more neighboring coding units or blocks.Reconstructed chroma sample values, Cr, are obtained by summing Creswith the corresponding predicted chroma values, Cpred. A predicted chroma value, Cpred, is determinedas described with regards to the predicted luma values, namely based on the prediction mode at a prediction stage like the intra / inter prediction stage 402, 508 described above. As mentioned, CS is optional. For example, CS may be disabled for chroma blocks with area size less than or equal to 4 samples.

[0086] In this example, the in-loop filtering process may be performed at a loop filtering stage of an encoder or decoder, such as the loop filtering stage 416 of FIG. 4 or the loop filtering stage 512 of FIG. 5. In-loop filtering may include filtering techniques such as deblocking (DBF), sample adaptive offset (SAG), adaptive loop filter (ALF), or some combination thereof. Broadly, these filtering techniques are used to reduce distortion introduced by encoding.

[0087] More specifically, a deblocking filter is designed to smooth sharp edges between blocks (CTUs and / or CUs) of the frame. This is sometimes referred to as removing blocking artifacts. The deblocking filter is generally applied to the entire reconstructed picture. For example, the rules, parameters, etc., for deciding whether to modify a pixel value and how to modify a pixel value at an edge are established on a frame-level basis. For example, one of multiple (e.g., three) filter strengths may be signaled. The deblocking filter may be applied to an 8x8 sample grid, a 4x4 sample grid, or some other size grid. The deblocking filter may first apply horizontal filtering for vertical edges of the picture and thereafter apply vertical filtering for horizontal edges of the picture, or vice versa. In some implementations of a codec (i.e., an encoder and decoder combination), the deblocking filter is applied before other in-loop filtering, if any, but this is not required.

[0088] SAO filtering also reduces distortion by compensating the pixel value offset between reconstructed pixels and original pixels. In general, SAO filtering classifies reconstructed pixels and adds a respective offset to each class or group of pixels. SAO may use different offsets pixel sample by pixel sample in a region depending on the sample classification, and SAO parameters may vary from region to region. The offsets may be determined at a decoder using, e.g., a look up table that is based on a histogram analysis made by the encoder. The region size may be fixed to one coding tree block (CTB) (or coding tree unit). That is, the SAO parameters may be signaled at CTB level.

[0089] More than one SAO type may be used. For example, an edge offset (EO) type and a band offset (BO) type may be used. The EO type may have multiple sub-types corresponding to processing along different directions. For example, the EO type may have four sub-types corresponding to processing along the horizontal, vertical, 135-degree, and 45- degree directions. For an EO sub-type, the value of a pixel (also called a sample) is comparedto two of its neighbors using one of four different gradient patterns to classify the pixel. An offset is applied to pixels in each of the four gradient patterns. No offset may be applied for pixel values that do not match one of the gradient patterns. The BO type may be based where the sample values fall within multiple bands (such as 32 bands for values 0-255 in 8-bit coding). An offset is applied to pixels in at least some of the bands, which offset is determined for respective bands. Each color component of the picture may have its own SAO parameters. The SAO filtering can increase edge sharpness and reduce ringing and impulse artifacts.

[0090] Filtering performed by an ALF may be selectively performed before or preferably after SAO filtering. An AEF can minimize the mean squared error between an original picture and the reconstructed picture, such as the picture output from SAO filtering, to improve the quality of the reconstruction. The AEF is generally referred to as adaptive because the coefficients may be signaled in the bitstream so they can be designed to reflect the content and distortion of the reconstructed picture. Filtering may be performed generally by partitioning sample locations into classes and applying (e.g., Wiener) filters on a class basis. The filter shapes may differ for luma and chroma components. For example, a 7 x 7 diamond shape (using thirteen samples) may be used for luma components, and a 5 x 5 diamond shape may be used for chroma components.

[0091] With regards to classification within AEF filtering, respective sub-blocks of the luma plane, such a 4 x 4 luma block, can be classified based on its directionality and two- dimensional (2D) Eaplacian activity. The 2D Eaplacian activity uses calculated gradients in multiple directions for the reconstructed luma samples, such as in the horizontal, vertical, 135-degree, and 45-degree directions. Up to 25 classes may be used. For each used class, a filter is signaled from encoder to decoder.

[0092] In addition to this luma sub-block-level filter adaptation, AEF may incorporate superblock (CTB or CTU)-level filter adaptation (CC-AEF). A luma block can use a filter set calculated for the current slice or one of the filter sets calculated for a previously coded slice. The luma block can also use one of multiple (e.g., 16) offline trained filter sets. Within each luma CTB, which filter from the chosen filter set should be applied to each 4 x 4 block is determined by the class calculated for that block. With regards to chroma blocks, AEF may only use CTB-level filter adaptation. Each CTB can select one of a number of filters available to it for its chroma components. For example, up to 8 filters can be used for chroma components in a slice such that each CTB can select one of these filters.

[0093] The filtering described above is by example only. More, fewer, or different filters may be used for filtering in an encoder and decoder, such as the critical sampling filter described below.

[0094] Adaptive range clipping may be combined with the LM process by modifying values for the inverse luma mapping and modifying values for the forward luma sampling using the adaptive range (e.g., the signaled lower and upper bounds for the luma components). In some implementations, the values for the forward luma mapping are stored in a forward lookup table (also called a forward luma mapping table) fwdLut that maps values in the original sample domain to values in the mapped sample domain. Similarly, the values for the inverse luma mapping that maps values in the mapped sample domain to values in the original sample domain may be stored in an inverse lookup table (also called an inverse luma mapping table) invLut. Accordingly, modifying the values for the forward luma mapping can include modifying the forward luma mapping table fwdLut according to:

[0095] fwdLut[y] = fwdLutf yi. ] when y < yi_; or

[0096] fwdLut[y] = fwdLut[yu] when y > yu.

[0097] In the above, yu and yu are signaled lower bound and upper bound, respectively, and y is the value in the original domain.

[0098] Similarly, modifying the values for the inverse luma mapping can include modifying the inverse luma mapping table invLut, where y' is the value in the mapped domain, according to:

[0099] invLut[y'] = yu when y' < fwdLut[yu]; or

[0100] invLut[y'] = yu when y' > fwdLut[yu].

[0101] When processing (e.g., decoding) the portion of the frame (e.g., the whole frame, a slice, etc.) is complete, the values for the forward luma mapping and / or the inverse luma mapping should be restored because, for the next picture / slice / portion, the mapping may or may not be modified. For example, the lookup tables should be reset to the unmodified version.

[0102] Further, when used with LM (i.e., LM is on), the adaptive range clipping described herein may be used in the mapped domain in certain stages. For example, in some implementations when LM is on, clipping with lower bound fwdLut[yL] and upper bound fwdLut[yu] is performed after any prediction in the mapped domain, such as after intra prediction, intra block copy (IBC) prediction, and palette prediction. Clipping may also be performed after adding prediction and residuals to form reconstructed blocks / coding units.Note that this clipping may be combined with existing clipping by simply replacing the bounds with fwdLut[yL] and fwdLut[yu].

[0103] In an example, the final prediction of a block may be a combined prediction from both mapped domain values and original domain values. Accordingly, the predictions (pixels forming one or more prediction blocks) from the mapped domain may be clipped, while the predictions from the original domain may be forward transformed using fwdLut.

[0104] When LM is off, or for in-loop filtering (e.g., deblocking, SAO, ALF), the signaled lower and upper bounds may be directly used in the clipping by using the bounds based on signaled bit depth (e.g., 8-bit, 10-bit, 12-bit, etc.).

[0105] When the clip flag is signaled below the frame / picture / slice level, such as for a region including a group of blocks (CTU, CTB), and the value of the clip flag indicates that adaptive range clipping should be performed, the clip bounds are only applied to prediction and reconstruction of the blocks in that region. The clip bounds are not applied to any filtering process.

[0106] As mentioned initially, critical sample filtering may be used to reduce distortion in a reconstructed frame by modifying the values of certain samples (e.g., pixels) of a reconstructed frame to a critical sample value (also called a defined sample value or fixed sample value). FIG. 8 illustrates an example 800 of an implementation of critical sample filtering where an original block 802 is encoded and subsequently decoded and reconstructed to form a reconstructed block 804. The reconstructed block 804 is subsequently subjected to critical sample filtering to form a filtered reconstructed block 806. The original block 802 may be a portion of an original frame of a video stream, such as the video stream 300, and may include multiple samples 808 (e.g., multiple pixels) that each have values (e.g., values of luma and / or chroma components). One or more of the values of the samples 808 may be identified by an encoder, such as the encoder 400, as a critical sample value based on a cardinality of the one or more of the values in the original block 802. For example, a value may be identified as a critical sample value if a cardinality of the value in the original block 802 is at or above a critical value threshold. In the illustrated example 800, the critical value threshold is 30% of the samples 808 in the original block 802. Accordingly, in the illustrated example 800, the value of ten is identified as a critical sample value because at least 30% of the samples 808 of the original block 802 have a value of ten.

[0107] After decoding and reconstructing the original block 802 to form the reconstructed block 804, values of reconstructed samples 810 of the reconstructed block 804 may have deviated from the values of the samples 808 of the original block 802 due to lossy encodingtechniques (e.g., quantization and / or other compression techniques). Accordingly, critical sample filtering may be used to modify at least some of the values of the reconstructed samples 810 of the reconstructed block 804 to the values of the samples 808 of the original block 802 (e.g., to the critical sample value). To do so, a decoder, such as the decoder 500, or an encoder, such as the encoder 400 in the reconstruction path, may assess the reconstructed samples 810 in a sample area 812 of the reconstructed block 804 and modify the value of a current reconstructed sample 810 in the sample area 812 to the critical sample value based on a critical sample range. The critical sample range may include an upper range bound and a lower range bound relative to the critical sample value identified by the encoder. In the illustrated example 800, the upper range bound of the critical sample range is eleven and the lower range bound of the critical sample range is nine.

[0108] The sample area 812 may also be referred to as a supporting area, and its shape is defined by the filter shape as discussed in more detail below. In this example, a 5-point crossshape filter is used. If a cardinality of the reconstructed samples 810 within the sample area 812 that have values within the critical sample range is at or above a sample area threshold, then the value of the current sample under consideration (i.e., the center sample 814) of the reconstructed samples 810 within the sample area 812 may be modified to the critical sample value to form the filtered reconstructed block 806. As used herein, the terms “within” and “between” are inclusive of the applicable bounding values unless noted otherwise. In the illustrated example 800, the sample area threshold is three of five of the reconstructed samples 810 of the sample area 812, but other sample area thresholds may be used as described in additional detail below. Because three of the reconstructed samples 810 of the sample area 812 include values within the critical sample range (e.g., values between nine and eleven), the center sample 814 of the sample area 812 is modified from the value of eleven to the critical sample value of ten. Otherwise, the center sample 814 would not be modified. By selectively filtering the values of at least some of the reconstructed samples 810 of the reconstructed block 804 to the values of the samples 808 of the original block 802 (e.g., to the critical sample value), distortions may be reduced in the filtered reconstructed block 806.

[0109] FIG. 9 is a flowchart of a technique 900 for identifying and encoding information for critical sample filtering in an encoded bitstream. The technique 900 can be implemented, for example, as a software program that may be executed by computing devices such as the transmitting station 102. The software program can include machine-readable instructions that may be stored in a memory such as the memory 204 or the secondary storage 214, andthat, when executed by a processor, such as the processor 202, may cause the computing device to perform the technique 900. The technique 900 may be implemented at least in part by an encoder, such as the encoder 400. The technique 900 can be implemented using specialized hardware or firmware. Multiple processors, memories, or both, may be used.

[0110] At operation 902, one or more critical sample values are identified in a region of a frame of a video stream, such as the video stream 300. The operation 902 may be implemented at least in part prior to the implementation of any lossy encoding techniques by the encoder (e.g., prior to the quantization stage 406 of the encoder 400). The region of the frame may be the entire frame, a block of the frame, such as the original block 802 in FIG. 8, a group of blocks of the frame (e.g., a CTB or a superblock of the frame), a slice of the frame, a tile of the frame, or some other portion of the frame. Accordingly, the region may comprise multiple samples (e.g., multiple of the samples 808) that each includes values of one or more planes of color data. For example, each of the samples may include a value of a luma component and / or values of one or both chroma components. As another example, each of the samples may include a value of a red component, a value of a green component, and / or a value of a blue component. Filtering is generally performed for each plane separately as described hereinbelow.

[0111] A critical sample value is a value that appears within the samples of the region in a certain quantity. Critical sample values may be identified based on a cardinality of the values of the samples within the region, such that a value may be identified as a critical sample value if the cardinality of the value within the region is at or above a critical value threshold. The critical value threshold may be a fraction (e.g., a percentage) of the samples within the region. For example, where the critical value threshold is 50% of the samples within the region, a value may be identified as a critical sample value if 50% or more of the samples within the region include the value. The critical sample value may be a value for a plane of color data, such as a value for the luma component or a value for one or both of the chroma components. Furthermore, multiple critical sample values may be identified for a plane of color data. For example, three critical sample values may be identified for the luma component, one critical sample value may be identified for one of the chroma components, and zero critical sample values may be identified for the other of the chroma components.

[0112] Furthermore, the same or different critical value thresholds may be used to identify critical sample values for different planes of color data. For example, the critical value threshold for identifying a value as a critical sample value for the luma component may be different than the critical value threshold for identifying a value as a critical sample valuefor one or both of the chroma components. Where multiple critical sample values are identified for a plane of color data, it may be desired that the critical sample values are not close in magnitude. Accordingly, a newly identified critical sample value may be excluded if a difference between an already identified critical sample value and the newly identified critical sample value is at or below a difference threshold.

[0113] At operation 904, a cardinality of each critical sample value in the region is identified. As mentioned previously, multiple critical sample values may be identified for a plane of color data. Therefore, different cardinalities for critical sample values for different planes of color data may be identified. For example, a first cardinality of the critical sample values for the luma component may be identified as three, a second cardinality of the critical sample values for the one of the chroma components may be identified as one, and a third cardinality of the critical sample values for the other of the chroma components may be identified as zero. Where the region is less than the entirety of a plane of color data, the cardinality for the plane may be the sum of the number of critical sample values in each region (e.g., without duplicating values).

[0114] Although the operations 902 and 904 are used by an encoder in this example, the technique 900 may include a different sequence of steps. For example, operation 904 may be performed first. In such a case, the cardinality of critical sample values in the region (e.g., the frame or portion of a frame) may be identified as an encoding parameter, and the encoder can identify the critical sample values using a histogram analysis of the values of the samples in the region to select those values with the highest presence up to the cardinality. In some implementations, the encoder and decoder may share values for the cardinality such that the cardinality does not have to be encoded as described below.

[0115] At operation 906, information relating to a critical sample range is identified. As mentioned previously, the critical sample range may include an upper range bound and a lower range bound. The information relating to the critical sample range may be used to determine (e.g., may define) the upper range bound and the lower range bound relative to respective critical sample values. In some implementations, the information relating to the critical sample range may include one or more integers that define the upper range bound and the lower range bound of the critical sample range for respective critical sample values. In some implementations, the upper range bound and the lower range bound of the critical sample range may be determined using different integers. For example, the upper range bound may be determined by a first integer that is different than a second integer that is used to determine the lower range bound. Furthermore, the upper range bound and / or the lowerrange bound of the critical sample range for different planes of color data may be determined using different integers. For example, the upper range bound for critical sample values for the luma component may be determined by a first integer that is different than a second integer that is used to determine the upper range bound for critical sample values for one of the chroma components.

[0116] The information relating to the critical sample range may apply to each individual critical sample value, to critical sample values for certain planes of color data, or to critical sample values for all planes of color data. Furthermore, the information relating to the critical sample range may apply to critical sample values of the region of the frame or to critical sample values of a level above the region of the frame (e.g., critical sample values of the entire frame). For example, the information relating to the critical sample range may apply to critical sample values for the luma component but not to critical sample values for the chroma components. As another example, the information relating to the critical sample range may apply to the critical sample values for the luma component as well as to the critical sample values for both of the chroma components. As another example, where the region of the frame is a block (e.g., the original block 802), the information relating to the critical sample range may apply to critical sample values of the block as well as to critical sample values of other blocks or to critical sample values of the entire frame.

[0117] The information relating to the critical sample range may be input into the encoder (e.g., by a user) or may be predetermined at the encoder (e.g., as rules, parameters, etc.). In implementations where the information relating to the critical sample range is predetermined at the encoder, the decoder may have corresponding predetermined information (e.g., corresponding rules, parameters, etc.) for implementing the critical sample filter.Furthermore, although the information relating to the critical sample range has primarily been described with respect to one or more integers, other types of information may be used to determine the critical sample range for respective critical sample values.

[0118] At operation 908, the one or more critical sample values of the region, the one or more cardinalities of the critical sample values of the region, and / or the information relating to the critical sample range may be encoded in an encoded bitstream. In some implementations, the one or more critical sample values of the region, the one or more cardinalities of the critical sample values of the region, and / or the information relating to the critical sample range may be signaled in header information in the encoded bitstream to be used by the decoder and / or by the encoder (e.g., in the reconstruction path) to implement the critical sample filter. The header information may include first header information thatsignals the one or more critical sample values of the region, second header information that signals the one or more cardinalities of the critical sample values of the region, and / or third header information that signals the information relating to the critical sample range.

[0119] As mentioned previously with respect to the operation 902, a critical sample value may be a value for a plane of color data, such as the value for a luma component and / or the value for one or both of the chroma components. Accordingly, the first header information and the second header information may apply to only certain planes of color data. Furthermore, as mentioned previously with respect to the operation 906, the information relating to the critical sample range may apply to individual critical sample values, to critical sample values for certain planes of color data, to critical sample values for all planes of color data, to critical sample values of the region of the frame, and / or to critical sample values of a level above the region of the frame (e.g., critical sample values of the entire frame). Thus, the third header information may apply to individual critical sample values, to certain planes of color data, to all planes of color data, to the region of the frame, and / or to a level above the region of the frame.

[0120] For example, the first header information may apply to the luma component and signal the critical sample values of the region for the luma component. The second header information may apply to the luma component and signal the cardinality of the critical sample values of the region for the luma component. The third header information may apply to the luma component and both of the chroma components and signal one or more integers that may be used to determine the critical sample range for the critical sample values of the region for the luma component and for the critical sample values of the region for both of the chroma components.

[0121] It may be desired that the header information be signaled in the encoded bitstream in a certain order. In some implementations, the critical sample values for a certain plane of color data may be signaled in the encoded bitstream after signaling the cardinality of the critical sample values for that plane of color data. For example, the critical sample values of the region for the luma component may be signaled in the first header information after signaling the cardinality of the critical sample values of the region for the luma component in the second header information. Thereafter, the critical sample values of the region for one of the chroma components may be signaled in the first header information after signaling the cardinality of the critical sample values of the region for that chroma component in the second header information, and so on.

[0122] Furthermore, in implementations where the information relating to the critical sample range applies to an individual critical sample value, the information relating to the critical sample range may be signaled in the encoded bitstream after signaling the critical sample value to which the information applies. Furthermore, in implementations where the information relating to the critical sample range applies to critical sample values for a certain plane of color data, the information relating to the critical sample range may be signaled in the encoded bitstream after signaling the critical sample values for that plane of color data. Furthermore, in implementations where the information relating to the critical sample range applies to more than one plane of color data, the information relating to the critical sample range may be signaled in the encoded bitstream after signaling the critical sample values for all of those planes of color data to which the information applies.

[0123] The one or more critical sample values may be signaled in the encoded bitstream (e.g., in the first header information) by any suitable method. In some implementations, the critical sample values for a plane of color data may be signaled in the first header information in order of magnitude. In such an implementation, the critical sample value for the plane of color data having the smallest magnitude is first signaled in the first header information, and then each critical sample value of increasing magnitude is signaled in the first header information thereafter by signaling a difference between the critical sample value and the immediately preceding critical sample value. For example, where the critical sample values for the luma component include ten, fourteen, twenty, and twenty-nine, the critical sample values may be signaled in the first header information by signaling ten, four, six, and nine.

[0124] In some implementations, the critical sample value having the smallest magnitude is first signaled in the first header information by signaling a difference between the critical sample value and a shifted bit-depth of the applicable plane of color data. For example, a critical sample value for a plane of color data having the smallest magnitude may be first signaled in the first header information by signaling a value equaling:

[0125] clowest- (1 « bit_depth » 1)

[0126] In the above, clowestrepresents the critical sample value for the plane of color data having the smallest magnitude, « represents the left-shift operator, bit_depth represents the number of bits used to represent a value for the plane of color data, and » represents the right- shift operator.

[0127] Furthermore, the difference between the critical sample value and the shifted bitdepth of the applicable plane of color data may be signaled using K'1' order signed- l-exponential Golomb coding. The difference between the critical sample value and the immediately preceding critical sample value may be signaled using K'1' order unsigned exponential Golomb coding.

[0128] The one or more cardinalities of the critical sample values of the region may also be signaled in the encoded bitstream (e.g., in the second header information) by any suitable method. In some implementations, the cardinalities of the critical sample values of the region may be signaled in the encoded bitstream using variable length coding. For example, the cardinalities of the critical sample values of the region may be signaled in the second header information using 0thorder unsigned exponential Golomb coding.

[0129] The information relating to the critical sample range may also be signaled in the encoded bitstream (e.g., in the third header information) by any suitable method. For example, in implementations where the information relating to the critical sample range includes one or more integers, the one or more integers may be signaled in the third header information by signaling values equaling:

[0130] int1- 1

[0131] In the above, int1represents the signaled integer.

[0132] In some implementations, rather than encoding the information relating to the critical sample range (e.g., one or more integers) in the encoded bitstream (e.g., in the third header information), the critical sample range (e.g., the upper range bound and the lower range bound) for each of the critical sample values may be encoded. Furthermore, in implementations where the information relating to the critical sample range is predetermined at the encoder, the information relating to the critical sample range may not be signaled in the encoded bitstream, as the decoder will have corresponding predetermined information for determining the critical sample range for the respective critical sample values that are signaled in the encoded bitstream. Accordingly, in such an implementation the third header information may be excluded from the encoded bitstream.

[0133] FIG. 10 is a flowchart of a technique 1000 for implementing the critical sample filter on a reconstructed frame or portion of the reconstructed frame (e.g., the reconstructed block 804). The technique 1000 can be implemented, for example, as a software program that may be executed by computing devices such as the transmitting station 102 or the receiving station 106. The software program can include machine-readable instructions that may be stored in a memory such as the memory 204 or the secondary storage 214, and that, when executed by a processor, such as the processor 202, may cause the computing device toperform the technique 900. The technique 1000 may be implemented at least in part at a loop filtering stage of a reconstruction path of an encoder, such as the encoder 400, at a loop filtering stage of a decoder, such as the decoder 500, and / or at a post filtering stage of the decoder. The technique 1000 can be implemented using specialized hardware or firmware. Multiple processors, memories, or both, may be used.

[0134] The reconstructed frame may be the frame of the video sequence discussed above with respect to FIG. 9 that has been encoded and subsequently decoded and reconstructed. Accordingly, the reconstructed region of the reconstructed frame may correspond to the region described above with respect to FIG. 9 and may comprise multiple reconstructed samples (e.g., multiple of the reconstructed samples 810) that correspond to the samples (e.g., the samples 808) of the region. Like the samples, the reconstructed samples may each include values for one or more planes of color data. As previously mentioned, after decoding and reconstruction, the values of the reconstructed samples may have deviated from the values of the samples.

[0135] At operation 1002, one or more critical sample values of the reconstructed region of the reconstructed frame are determined. The one or more critical sample values of the reconstructed region may be determined from the header information of the encoded bitstream. For example, the one or more critical sample values of the reconstructed region may be signaled in the first header information of the encoded bitstream as described above.

[0136] At operation 1004, the critical sample ranges for each of the one or more critical sample values are determined. Where multiple critical sample values for a plane of color data have been determined, it may be desired that the critical sample range for respective critical sample values do not overlap. Accordingly, where multiple critical sample values for a plane of color data are signaled in the first header information in order of magnitude, the upper range bound and lower range bound for respective critical sample values may be determined according to:

[0137] C^b= Cl+ int1

[0139] In the above, C^band C}brepresent the upper range bound and the lower range bound of the critical sample range for the critical sample value Cl, respectively, int1represents the integer from the information relating to the critical sample range that applies to the critical sample value Cl, and C'-1represents the critical value that is signaled in the first header information immediately preceding the critical sample value Cl. In other words, theupper range bound of the critical sample range relative to a respective critical sample value may be defined as the critical sample value plus the integer, and the lower range bound of the critical sample range relative to the critical sample value is defined as the greater of the critical sample value minus the integer, and an immediately preceding critical sample value plus one.

[0140] Furthermore, in implementations where adaptive range clipping is used, it may be desired that the upper bound and the lower bound used for adaptive range clipping be different than the upper range bound and the lower range bound of the critical sample range used for critical sample filtering. Accordingly, if the lower range bound or the upper range bound of the critical sample range for a respective critical sample value equals either the upper bound or the lower bound used for adaptive range clipping, then the upper range bound or the lower range bound of the critical sample range for the critical sample value may be modified to ensure that the values remain different. In some implementations, an integer may be added to or subtracted from the upper range bound and / or the lower range bound of the critical sample range.

[0141] At operation 1006, values for the reconstructed samples within a sample area (e.g., the sample area 812) of the reconstructed region are determined. The sample area may comprise a primary sample (e.g., the center sample 814) and any number of secondary samples proximate to (e.g., surrounding) the primary sample. Furthermore, the sample area may comprise any configuration of the reconstructed samples. For example, the sample area may comprise the primary sample surrounded by four of the secondary samples such as to form a cross shape configuration in which the primary sample is at the center such as in the example of FIG. 8. As another example, the sample area may comprise the primary sample surrounded by eight of the secondary samples such as to form a 9-point 3x3 square shape filter in which the primary sample is at the center. As another example, the sample area may comprise the primary sample surrounded by twelve of the secondary samples such as to form a 5x5 diamond-shape filter in which the primary sample is at the center.

[0142] The values for a plane of color data of each of the reconstructed samples (e.g., the primary sample and the secondary samples) within the sample area may be assessed to determine whether the values are within the critical sample range for the one or more critical sample values for that plane of color data. Furthermore, more than one sample area may be assessed. For example, sample areas may be structured within the reconstructed region such that all the reconstructed samples are included in at least one of the sample areas. As another example, sample areas may be structured within the reconstructed region such that each ofthe reconstructed samples are included in one of the sample areas as the primary sample. Where multiple sample areas are assessed, the sample areas may be assessed in any consecutive order or may be assessed concurrently.

[0143] At operation 1008, one or more of the reconstructed samples of the sample area may be modified to the critical sample value based on the critical sample range. In some implementations, a reconstructed sample (e.g., the center pixel) within the sample area may be modified to the critical sample value responsive to determining that a cardinality of the reconstructed samples within the sample area that have a value within the applicable critical sample range is above a sample area threshold. The sample area is defined by a filter such as shown by example in FIG. 8, and the pixels of the image (or portion of the image) are processed sequentially using the filter.

[0144] In some implementations, the sample area threshold may be a fraction (e.g., a percentage) of the reconstructed samples within the sample area. For example, in implementations where the sample area comprises the primary sample and four of the secondary samples such as to form a cross shape as shown in FIG. 8, the sample area threshold may be four of the five reconstructed samples of the sample area. Accordingly, the central pixel of the sample area may be modified to the critical sample value responsive to determining that four or more of the reconstructed samples of the sample area have a value within the critical sample range for that critical sample value. In some implementations, however, the sample area threshold may be all of the reconstructed samples within the sample area such that the central (e.g., filter candidate) pixel within the sample area may only be modified to the critical sample value responsive to determining that all of the reconstructed samples of the sample area have values within the critical sample range for that critical sample value.

[0145] FIG. 11 is a block diagram of a decoder 1100 in which critical sample filtering may be implemented. The decoder 1100 may be like the decoder 500. Critical sample filtering may be implemented within the loop filtering stage 512 of the decoder 1100 (e.g., after the reconstruction stage 510). In some implementations, the loop filtering stage 512 may include a deblocking stage 1112, a sample adaptive offset (SAG) stage 1114, a critical sample filtering (CSF) stage 1116, and an adaptive loop filtering (ALF) stage 1118. Although the CSF stage 1116 is depicted as being implemented after the SAO stage 1114 and before the ALF stage 1118, the CSF stage 1116 may be implemented before or after any of the deblocking stage 1112, the SAO stage 1114, the CSF stage 1116, and the ALF stage 1118. For example, the CSF stage 1116 may be implemented after the deblocking stage 1112 andbefore the SAO stage 1114. As another example, the CSF stage 1116 may be implemented after the ALF stage 1118.

[0146] In some implementations, the sequence by which to implement the CSF stage 1116 is implemented in relation to the deblocking stage 1112, the SAO stage 1114, and the ALF stage 1118 may be encoded in the encoded bitstream by the encoder, such as the encoder 400. For example, in some implementations the header information of the encoded bitstream may include fourth header information, and the sequence by which to implement the CSF stage 1116 may be signaled in the fourth header information.

[0147] Furthermore, in some implementations the CSF stage 1116 may not be implemented in the loop filtering stage 512 but, rather, may be implemented in the post filtering stage 514 of the decoder 1100. Furthermore, in some implementations, the CSF stage 1116 may be implemented in a loop filtering stage in a reconstruction path of an encoder.

[0148] For simplicity of explanation, the techniques herein are depicted and described as respective series of steps or operations. However, the steps or operations in accordance with this disclosure can occur in various orders and / or concurrently. Additionally, other steps or operations not presented and described herein may be used. Furthermore, not all illustrated steps or operations may be required to implement a technique in accordance with the disclosed subject matter.

[0149] The aspects of encoding and decoding described above illustrate some examples of encoding and decoding techniques. However, it is to be understood that encoding and decoding, as those terms are used in the claims, could mean compression, decompression, transformation, or any other processing or change of data.

[0150] The word “example” is used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as “example” is not necessarily to be construed as being preferred or advantageous over other aspects or designs. Rather, use of the word “example” is intended to present concepts in a concrete fashion. As used in this application, the term “or” is intended to mean an inclusive “or” rather than an exclusive “or.” That is, unless specified otherwise or clearly indicated otherwise by the context, the statement “X includes A or B” is intended to mean any of the natural inclusive permutations thereof. That is, if X includes A; X includes B; or X includes both A and B, then “X includes A or B” is satisfied under any of the foregoing instances. In addition, the articles “a” and “an” as used in this application and the appended claims should generally be construed to mean “one or more,” unless specified otherwise or clearly indicated by the context to be directed to a singular form. Moreover, use of the term “an implementation” or the term “oneimplementation” throughout this disclosure is not intended to mean the same embodiment or implementation unless described as such.

[0151] Implementations of the transmitting station 102 and / or the receiving station 106 (and the algorithms, methods, instructions, etc., stored thereon and / or executed thereby, including by the encoder 400 and the decoder 500) can be realized in hardware, software, or any combination thereof. The hardware can include, for example, computers, intellectual property (IP) cores, application- specific integrated circuits (ASICs), programmable logic arrays, optical processors, programmable logic controllers, microcode, microcontrollers, servers, microprocessors, digital signal processors, or any other suitable circuit. In the claims, the term “processor” should be understood as encompassing any of the foregoing hardware, either singly or in combination. The terms “signal” and “data” are used interchangeably. Further, portions of the transmitting station 102 and the receiving station 106 do not necessarily have to be implemented in the same manner.

[0152] Further, in one aspect, for example, the transmitting station 102 or the receiving station 106 can be implemented using a general-purpose computer or general-purpose processor with a computer program that, when executed, carries out any of the respective methods, algorithms, and / or instructions described herein. In addition, or alternatively, for example, a special purpose computer / processor can be utilized that can contain other hardware for carrying out any of the methods, algorithms, or instructions described herein.

[0153] The transmitting station 102 and the receiving station 106 can, for example, be implemented on computers in a video conferencing system. Alternatively, the transmitting station 102 can be implemented on a server, and the receiving station 106 can be implemented on a device separate from the server, such as a handheld communications device. In this instance, the transmitting station 102, using an encoder 400, can encode content into an encoded video signal and transmit the encoded video signal to the communications device. In turn, the communications device can then decode the encoded video signal using a decoder 500. Alternatively, the communications device can decode content stored locally on the communications device, for example, content that was not transmitted by the transmitting station 102. Other suitable transmitting and receiving implementation schemes are available. For example, the receiving station 106 can be a generally stationary personal computer rather than a portable communications device, and / or a device including an encoder 400 may also include a decoder 500.

[0154] Further, all or a portion of implementations of the present disclosure can take the form of a computer program product accessible from, for example, a (e.g., non-transitory)computer-usable or computer-readable medium. A computer-usable or computer-readable medium can be any device that can, for example, tangibly contain, store, communicate, or transport the program for use by or in connection with any processor. The medium can be, for example, an electronic, magnetic, optical, electromagnetic, or semiconductor device. Other suitable mediums are also available.

[0155] The above-described embodiments, implementations, and aspects have been described to facilitate easy understanding of this disclosure and do not limit this disclosure. On the contrary, this disclosure is intended to cover various modifications and equivalent arrangements included within the scope of the appended claims, which scope is to be accorded the broadest interpretation as is permitted under the law to encompass all such modifications and equivalent arrangements.

Claims

What is claimed is:

1. A method, comprising: determining a critical sample value for a region of a frame of a video sequence, the region comprising multiple samples, wherein the critical sample value is a value of a sample from the multiple samples having a cardinality within the region that is above a critical value threshold; and after reconstructing the region of the frame to form a reconstructed region, modifying a value of a reconstructed sample within the reconstructed region to the critical sample value.

2. The method of claim 1, comprising: determining a critical sample range for the reconstructed region, the critical sample range comprising an upper range bound and a lower range bound relative to the critical sample value, wherein modifying the value of the reconstructed sample to the critical sample value is based on the critical sample range.

3. The method of claim 2, wherein the critical sample range applies to all planes of color data of the region.

4. The method of claim 2, wherein the reconstructed region comprises multiple reconstructed samples, the method comprising: determining the value for each of the multiple reconstructed samples within a sample area of the reconstructed region, wherein the sample area comprises the reconstructed sample and secondary reconstructed samples proximate to the reconstructed sample; and determining a cardinality of the multiple reconstructed samples within the sample area that have the value between the upper range bound and the lower range bound of the critical sample range, inclusive.

5. The method of claim 4, wherein modifying the value of the reconstructed sample to the critical sample value is responsive to determining that the cardinality of the multiple reconstructed samples within the sample area that have the value between the upper range bound and the lower range bound of the critical sample range, inclusive, is above a sample area threshold.

6. The method of one of claim 4 or claim 5, wherein the critical sample range and the critical sample value are determined at least in part by header information within an encoded bitstream.

7. The method of claim 6, comprising: determining, from the header information, a sequence for modifying the value of the reconstructed sample to the critical sample value and applying at least one of a deblocking filter, a sample offset, or an adaptive loop filter to the region.

8. The method of one of claim 6 or claim 7, wherein the critical sample value is one of multiple critical sample values, and wherein the multiple critical sample values are signaled in the header information in order of magnitude.

9. The method of claim 8, wherein the multiple critical sample values are signaled in the header information by signaling a difference with respect to an immediately preceding critical sample value from the multiple critical sample values.

10. The method of any one of claim 2 or any claim dependent from claim 2, wherein: header information includes an integer; the upper range bound of the critical sample range relative to a respective critical sample value from the multiple critical sample values is defined as the respective critical sample value plus the integer; and the lower range bound of the critical sample range relative the respective critical sample value is defined as a greater of: the respective critical sample value minus the integer, and an immediately preceding critical sample value plus one.

11. The method of any one of claims 1 to 10, wherein the critical sample value corresponds to a plane of color data of the region.

12. The method of one of claim 4 or any claim dependent from claim 4, wherein the sample area comprises at least one of a cross shape comprising five of the multiple reconstructed samples, a square shape comprising nine of the multiple reconstructed samples,or a diamond shape comprising thirteen of the multiple reconstructed samples, and wherein the reconstructed sample is at a center of the at least one of the cross shape, the square shape, or the diamond shape.

13. The method of one of claim 5 or any claim dependent from claim 5, wherein the sample area threshold is equal to a cardinality of the multiple reconstructed samples within the region.

14. A method, comprising: identifying a critical sample value in a region of a frame of a video sequence, the region comprising multiple samples, wherein the critical sample value is a value of a sample from the multiple samples having a cardinality within the region that is above a critical value threshold; and encoding the critical sample value in an encoded bitstream.

15. The method of claim 14, comprising: encoding information relating to a critical sample range in the encoded bitstream, the critical sample range comprising an upper range bound and a lower range bound relative to the critical sample value.

16. The method of claim 15, wherein the information relating to the critical sample range identifies an integer, and wherein at least one of the upper range bound or the lower range bound is based on the integer.

17. The method of any one of claims 14 to 16, wherein the critical sample value is one of multiple critical sample values, the method comprising: identifying a cardinality of the multiple critical sample values in the region; and encoding the multiple critical sample values and cardinality of the multiple critical sample values in the region in the encoded bitstream.

18. The method of claim 17, wherein the multiple critical sample values are encoded in the encoded bitstream in order of magnitude.

19. The method of claim 17, wherein the multiple critical sample values are encoded inthe bitstream by encoding a difference with respect to an immediately preceding critical sample value from the multiple critical sample values.

20. The method of one of claim 15 or any claim dependent from claim 15, wherein the critical sample range applies to all planes of color data of the region.

21. An apparatus, comprising: a processor configured to perform the method according to any one of claims 1 to 20.

22. A non-transitory, computer-readable storage medium storing instructions that, when executed, cause a processor to perform the method according to any one of claims 1 to 20.

23. A non-transitory, computer-readable storage medium storing an encoded bitstream comprising compressed image data corresponding to a region of a frame of a video sequence and first header information comprising a critical sample value for the region, wherein the region comprises multiple samples, and wherein the critical sample value is a value of a sample from the multiple samples having a cardinality within the region that is above a critical value threshold.

24. The non-transitory, computer-readable storage medium of claim 23, wherein the critical sample value is one of multiple critical sample values for the region, and the first header information includes the multiple critical sample values.

25. The non-transitory, computer-readable storage medium of claim 24, wherein the encoded bitstream comprises second header information including a cardinality of the multiple critical sample values for the region.

26. The non-transitory, computer-readable storage medium of claim 25, wherein the encoded bitstream comprises third header information including an integer that defines at least one of an upper range bound or a lower range bound for a critical sample range for the region, and wherein upper range bound and the lower range bound are relative to the multiple critical sample values.

27. The non-transitory, computer-readable storage medium of claim 26, wherein the firstheader information is signaled in the encoded bitstream after the second header information.

28. The non-transitory, computer-readable storage medium of one of claim 26 or claim 27, wherein the third header information is signaled in the encoded bitstream after the first header information.

29. The non-transitory, computer-readable storage medium of one of claim 26 or any claim dependent from claim 26, wherein the encoded bitstream comprises fourth header information that indicates a sequence for modifying the value of a reconstructed sample to the critical sample value and applying at least one of a deblocking filter, a sample offset, or an adaptive loop filter to the region.

30. The non-transitory, computer-readable storage medium of claim 24, wherein the multiple critical sample values are signaled in the first header information in order of magnitude, and wherein the multiple critical sample values are signaled in the first header information by signaling a difference with respect to an immediately preceding critical sample value from the multiple critical sample values.

Images