JP2007006533A - Encoding apparatus, encoding method, filter processing apparatus and filter processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To adaptively use a pre-filter in accordance with features of an input image while performing real-time encoding. <P>SOLUTION: An encoding difficulty computing part 42 in an encoding control section 15 calculates an encoding difficulty representing a difficulty of image encoding. A pre-filter control part 51 controls characteristics of a spatial filter 52 and a temporal filter 53 in a pre-filter part 50 in accordance with a ratio of the encoding difficulty in a bit rate and a ratio between an encoding difficulty of a latest B-picture and an encoding difficulty of a latest I-picture. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an encoding device and an encoding method for encoding image data, and a filter processing device and a filter processing method for filtering image data.

  The transmission side compresses and transmits the image data, and the reception side decompresses the compressed image data. The image data is compressed and recorded. At the time of reproduction, the compressed image data is expanded and output. In a compressed image recording / reproducing apparatus or the like, as a method of compressing image data, for example, there is a bidirectional predictive encoding method adopted in the Moving Picture Experts Group (MPEG) standard. In this bidirectional predictive encoding method, encoding efficiency is improved by using bidirectional prediction. In the bi-directional predictive coding scheme, three types of coding are performed: intra-frame coding, inter-frame forward predictive coding, and bi-directional predictive coding. An image of each coding type is an I picture (intra coded picture), P picture (predictive coded picture) and B picture (bidirectionally predictive coded picture).

  By the way, the image data is compressed and encoded by the image encoding device using the bidirectional predictive encoding method adopted in the above-mentioned MPEG standard, and the compressed and encoded image data is decoded by the image decoding device. In a system that expands and decompresses, the image quality after encoding and further decoding greatly depends on the pattern of the input video material. In other words, if the video contains a lot of high-frequency components, or if it is a fast-moving picture, the redundancy in the spatial and temporal directions is small, that is, the coding difficulty level is large. In the case of further decoding, encoding distortion is likely to occur.

  For the purpose of reducing encoding distortion when a video material having a high encoding difficulty as described above is input to the image encoding device, a low-pass filter in the spatial direction or time direction is provided at the input portion of the image encoding device. In general, a technique is known in which the spatial high frequency component or motion component of the input video material is reduced in advance. A low-pass filter used for such a purpose is called a pre-filter or a pre-filter.

  FIG. 7 shows an example of the configuration of an image encoding device provided with both a pre-filter in the spatial direction and a pre-filter in the temporal direction. This image encoding device includes a spatial filter 252 as a prefilter in the spatial direction and a temporal filter 253 as a prefilter in the temporal direction, and the input image signal S1 is sequentially transmitted through the spatial filter 252 and the temporal filter 253. It has come to pass. The spatial filter 252 is a low-pass filter in the spatial direction, and specifically, can be easily realized by a normal digital filter. The temporal filter 253 is a low-pass filter in the time direction. Specifically, for example, the frame memory 254 that stores the output signal of the spatial filter 252 for one frame, the output signal of the spatial filter 252 and the frame memory 254 are stored. Further, the signal is added by an adder 255 that adds the signal of the previous frame with a predetermined weighting K: (1-K) and outputs the result. Note that 0 ≦ K ≦ 1.

  The image encoding device further receives the output signal of the time filter 253, and rearranges the order of pictures (I picture, P picture, B picture) according to the encoding order, and this image rearrangement circuit. The scan conversion / macroblocking circuit 222 which inputs the output data of the circuit 221, discriminates whether it is a frame structure or a field structure, performs a scan conversion according to the discrimination result and a macroblock of 16 × 16 pixels, and this scan conversion A motion detection circuit 214 that detects a motion vector based on the output data of the macroblocking circuit 222 is provided.

  The image coding apparatus further includes a subtracting circuit 231 that takes a difference between the output data of the scan conversion / macroblocking circuit 222 and the predicted image data, and DCT (discrete cosine transform) on the output data of the subtracting circuit 231. A DCT circuit 232 that performs DCT in units of blocks and outputs DCT coefficients, a quantization circuit 233 that quantizes the output data of the DCT circuit 232, and a variable length that performs variable length encoding of the output data of the quantization circuit 233 The encoding circuit 234, the buffer memory 235 that temporarily holds the output data of the variable length encoding circuit 234 and outputs it as compressed image data S2 consisting of a bit stream of a constant bit rate, and the output data of the quantization circuit 33 Inverse quantization circuit 236 that performs inverse quantization, and inverse DCT that performs inverse DCT on the output data of the inverse quantization circuit 236 A CT circuit 237, an adder circuit 238 that adds and outputs the output data of the inverse DCT circuit 237 and the predicted image data, and holds output data of the adder circuit 238. The motion vector sent from the motion detection circuit 214 is used as a motion vector. Accordingly, a motion compensation circuit 239 that performs motion compensation and outputs the predicted image data to the subtraction circuit 231 and the addition circuit 238 is provided.

  In the image coding apparatus shown in FIG. 7, the spatial filter 252 removes the spatial high frequency component of the input image signal S1, and the temporal filter 253 moves the motion component (or noise between frames) in the output signal of the spatial filter 252. Remove. Since other operations in the image coding apparatus shown in FIG. 7 are the same as those in a general image coding apparatus, description thereof is omitted.

  In addition, the following prior art documents are mentioned as a thing relevant to the above description (patent document 1).

JP-A-3-252284

  By the way, if the pre-filter as described above is always used, even if it is effective for reducing the amount of information of a pattern with a high degree of encoding difficulty, the filter is also applied to a pattern with a low level of encoding difficulty, such as a still image. Therefore, there is a problem that the decoded video often gives a blurred impression as a whole.

  On the other hand, non-compressed image data is preliminarily compressed and encoded (hereinafter referred to as first-pass compression encoding) to estimate the amount of data after compression encoding, and then generated based on the estimated amount of data. In the case of an encoding method that does not require real-time performance, ie, compression encoding (hereinafter referred to as second-pass compression encoding) while controlling the amount of code, it was obtained by the first-pass compression encoding. Based on the data, a technique is also known that uses a pre-filter in the second pass compression encoding only for a portion (temporal portion) where the encoding difficulty is large. However, this method has a problem that the prefilter cannot be adaptively used according to the characteristics of the input image while encoding in real time.

  The present invention has been made in view of such a problem, and a first object thereof is to perform encoding in which a prefilter can be adaptively used according to the characteristics of an input image while encoding in real time. An apparatus and an encoding method are provided. A second object of the present invention is to provide a filter processing device and a filter processing method that can be applied to such an encoding device and an encoding method.

  A first encoding device according to the present invention generates encoded image data by encoding image data, the pre-filter for filtering the image data with a predetermined pass bandwidth, and the upper image data Filter band changing means for changing the pass bandwidth so as to decrease as the parameter representing the motion component increases, and encoding the image data filtered by the prefilter to convert the encoded image data into Encoding means to generate.

  According to a first encoding method of the present invention, encoded image data is generated by encoding image data, and a passband of a prefilter is increased in accordance with an increase in a parameter representing a motion component of the image data. A filter band changing procedure for changing the width so as to be narrowed and an encoding procedure for generating the encoded image data by encoding the image data filtered by the prefilter.

  The first filter processing apparatus of the present invention filters image data, and a pre-filter for filtering the image data with a predetermined pass bandwidth and a parameter representing a motion component of the image data are large. And a filter band changing means for changing the pass band width so as to become narrower.

  According to a first filtering method of the present invention, image data is filtered with a predetermined pass bandwidth, and the pass bandwidth becomes narrower as a parameter representing a motion component of the image data becomes larger. The filter band changing procedure is changed to be changed.

  According to the first encoding device, the first encoding method, the first filter processing device, and the first filter processing method of the present invention, a prefilter is selected in accordance with an increase in a parameter representing a motion component of image data. The pass bandwidth is changed to be narrow.

  A second encoding device according to the present invention generates encoded image data by encoding image data, the pre-filter for filtering the image data with a predetermined pass bandwidth, and the upper image data Filter band changing means for changing the pass bandwidth to increase as the parameter representing the motion component of the image becomes smaller, and encoding the image data filtered by the pre-filter to encode the encoded image data Encoding means to generate.

  According to a second encoding method of the present invention, encoded image data is generated by encoding image data, and the passband of the prefilter is reduced as the parameter representing the motion component of the image data decreases. A filter band changing procedure for changing the width so as to widen and an encoding procedure for generating the encoded image data by encoding the image data filtered by the prefilter.

  The second filter processing apparatus of the present invention filters image data, and a pre-filter for filtering the image data with a predetermined pass bandwidth and a parameter representing a motion component of the image data are small. And a filter band changing means for changing the pass band width so as to increase.

  According to a second filtering method of the present invention, image data is filtered with a predetermined pass bandwidth, and the pass bandwidth becomes wider as the parameter representing the motion component of the image data becomes smaller. The filter band changing procedure is changed to be changed.

  In the second encoding device, the second encoding method, the second filter processing device, and the second filter processing method of the present invention, the pre-filter is selected as the parameter representing the motion component of the image data decreases. Is changed so as to widen the pass bandwidth.

  According to the first encoding device or the first encoding method of the present invention, the pre-filter pass bandwidth is changed to be narrowed as the parameter representing the motion component of the image data increases. Therefore, it is possible to adaptively use the prefilter according to the characteristics of the input image while encoding in real time, and it is possible to improve the impression of visual image quality.

  Further, according to the first filter processing device or the first filter processing method of the present invention, the pre-filter pass bandwidth is changed to be narrowed as the parameter representing the motion component of the image data increases. Since it did in this way, there exists an effect that a pre-filter can be used adaptively according to the characteristic of an input image.

  Further, according to the second encoding device or the second encoding method of the present invention, the pre-filter pass bandwidth is changed so as to increase as the parameter representing the motion component of the image data decreases. As a result, it is possible to adaptively use the pre-filter according to the characteristics of the input image while encoding in real time, and it is possible to improve the impression of visual image quality.

  Further, according to the second filter processing apparatus or the second filter processing method of the present invention, the pre-filter pass bandwidth is changed so as to increase as the parameter representing the motion component of the image data decreases. Since it did in this way, there exists an effect that a pre-filter can be used adaptively according to the characteristic of an input image.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 2 is a block diagram showing a schematic configuration of the image coding apparatus according to the first embodiment of the present invention. This image encoding apparatus receives an input image signal S1, inputs a prefilter unit 50 for reducing the amount of information, and an output signal of the prefilter unit 50, and performs preprocessing for compression encoding, etc. The encoder control unit 11 that performs the above, the FIFO (first-in first-out) memory 12 for outputting the output data of the encoder control unit 11 after being delayed by a predetermined time, and the output data of the FIFO memory 12 are input, A motion vector is detected based on the output data of the encoder 13 as an encoding means for outputting the compressed image data S2 after being compressed and encoded by an encoding method according to the type, and sent to the encoder 13 The detection circuit 14 and the intra AC data S3 output from the encoder control unit 11 and the motion detection circuit 14 A coding control unit 15 for controlling the encoder 13 based on the ME residual data S4 and the generated bit amount data S5 output from the encoder 13, and a prefilter control based on the data from the coding control unit 15. And a prefilter control unit 51 for controlling the characteristics of the prefilter unit 50 according to the signal S6. The image encoding apparatus shown in FIG. 2 is configured to perform feedforward type rate control that controls the amount of generated code based on the characteristics of image data before compression encoding.

  As shown in FIG. 3, the encoding control unit 15 and the pre-filter control unit 51 include a CPU (Central Processing Unit) 16, a ROM (Read Only Memory) 17, and a RAM (Random) connected to each other via a bus 19. Each of the functions in the encoding control unit 15 and the prefilter control unit 51, which will be described later, is configured by a computer having an access memory 18 and the CPU 16 executes a program stored in the ROM 17 using the RAM 18 as a working area. Has come to be realized. The ROM 17 stores an image encoding control program according to the present invention, and may be an IC (integrated circuit), a storage device using a magnetic disk such as a hard disk or a floppy disk as a recording medium, or a CD ( Compact disk) —A storage device using an optical disk such as a ROM as a recording medium, or a storage device using another type of recording medium may be used. In any case, the IC and various recording media correspond to a medium on which the image encoding control program according to the present invention is recorded.

  FIG. 1 is a block diagram showing a detailed configuration of the image encoding device shown in FIG. As shown in this figure, the pre-filter unit 50 includes a spatial filter 52 as a pre-filter in the spatial direction and a temporal filter 53 as a pre-filter in the temporal direction, and the input image signal S1 52 and the time filter 53 are sequentially passed. The spatial filter 52 is a low-pass filter in the spatial direction, and is specifically realized by a digital filter, for example. The temporal filter 53 is composed of a low-pass filter in the time direction. Specifically, for example, the frame memory 54 that stores the output signal of the spatial filter 52 for one frame, and the output signal of the spatial filter 52 and the frame memory 54 are stored. This is realized by an adder 55 that adds and outputs a signal of the previous frame with a predetermined weighting K: (1-K). K is a weighting coefficient, and 0 ≦ K ≦ 1. The spatial filter 52 can change filter characteristics such as a pass bandwidth by changing filter coefficients. Similarly, the time filter 53 can change the characteristics of the filter such as the pass bandwidth by changing the value of the weighting coefficient K.

  The encoder control unit 11 receives the output signal of the prefilter unit 50, and rearranges the order of pictures (I picture, P picture, B picture) according to the order of encoding, and the image rearrangement circuit. 21 output data is input, it is discriminated whether it is a frame structure or a field structure, a scan conversion / macroblocking circuit 22 for performing a scan conversion and a 16 × 16 pixel macroblock according to the discrimination result, and this scan conversion / The output data of the macroblocking circuit 22 is input, the intra AC in the I picture is calculated, the intra AC data S3 is sent to the encoding control unit 15, and the output data of the scan conversion / macroblocking circuit 22 is sent to the FIFO memory 12 And an intra AC arithmetic circuit 23 to be sent to the motion detection circuit 14. Intra AC is defined as the sum of absolute values of the difference between the pixel value of each pixel in the 8 × 8 pixel DCT block and the average value of the pixel values in the DCT block in the I picture. It can be said that it represents.

  The encoder 13 is a subtracting circuit 31 that takes a difference between output data of the FIFO memory 12 and predicted image data, and a DCT circuit that performs DCT on the output data of the subtracting circuit 31 in units of DCT blocks and outputs DCT coefficients. 32, a quantization circuit 33 for quantizing the output data of the DCT circuit 32, a variable length encoding circuit 34 for variable length encoding the output data of the quantization circuit 33, and a variable length encoding circuit 34 A buffer memory 35 that temporarily holds output data and outputs it as compressed image data S2 composed of a bit stream of a constant bit rate, an inverse quantization circuit 36 that inversely quantizes the output data of the quantization circuit 33, and the inverse quantum The inverse DCT circuit 37 that performs inverse DCT on the output data of the conversion circuit 36, the output data of the inverse DCT circuit 37, and the predicted image data Are added and output, and the output data of the addition circuit 38 is held, motion compensation is performed according to the motion vector sent from the motion detection circuit 14, and the predicted image data is subtracted from the subtraction circuit 31 and the addition circuit 38. And a motion compensation circuit 39 for outputting the signal. The buffer memory 35 sends generated bit amount data S5 representing the bit amount generated by the variable length encoding circuit 34 to the encoding control unit 15.

  Based on the output data of the encoder control unit 11, the motion detection circuit 14 calculates the absolute value of the difference in pixel value between the target macroblock of the picture to be compression-encoded and the target macroblock in the referenced picture. A macro block having the smallest sum or sum of squares is searched for, and a motion vector is detected and sent to the motion compensation circuit 39. Further, when obtaining the motion vector, the motion detection circuit 14 sends the absolute value sum or the square sum of the pixel value differences between the macroblocks that are minimized to the encoding control unit 15 as the ME residual data S4. It has become.

  The encoding control unit 15 includes an ME residual calculation unit 41 that calculates an ME residual that is a value obtained by adding the ME residual data S4 from the motion detection circuit 14 for the entire picture, and the ME residual calculation unit 41. Based on the calculated ME residual and the intra AC data S3 from the intra AC arithmetic circuit 23, an encoding difficulty calculating unit 42 for calculating an encoding difficulty indicating the difficulty of encoding a picture, and a buffer memory VBV which is a virtual buffer corresponding to the input buffer in the image decoding apparatus which expands the image data compressed and encoded by the image encoding apparatus according to the present embodiment based on the generated bit amount data S5 from 35 (Video Buffering Verifier) A VBV buffer occupancy calculation unit 43 that calculates the data occupancy of the buffer (hereinafter simply referred to as occupancy) is provided. It can be said that the ME residual represents the speed of motion of the video and the complexity of the pattern.

  The encoding control unit 15 further determines the target code amount based on the encoding difficulty calculated by the encoding difficulty calculating unit 42 and the VBV buffer occupancy calculated by the VBV buffer occupancy calculating unit 43. A target code amount determination unit 44 to determine, and a quantization index corresponding to the quantization characteristic value in the quantization circuit 33 so that the generated code amount in the encoder 13 becomes the target code amount determined by the target code amount determination unit 44. A quantization index determination unit 45 that determines and sends the quantization index to the quantization circuit 33 is provided.

  The prefilter control unit 51 controls the prefilter unit 50 based on the encoding difficulty level calculated by the encoding difficulty level calculation unit 42 in the encoding control unit 15.

  Here, the encoding difficulty level will be described. The encoding difficulty level represents the difficulty level of picture encoding, which can be rephrased as a ratio of the amount of data necessary to maintain the same image quality. There are various methods for quantifying the encoding difficulty level. In this embodiment, the encoding difficulty level is obtained using intra AC for I pictures, and the ME residual is used for P pictures and B pictures. The encoding difficulty level is obtained. As described above, the intra AC represents the complexity of the picture, and the ME residual represents the speed of motion of the video and the complexity of the picture, and these are strongly correlated with the difficulty of encoding. It is possible to calculate the degree of difficulty of encoding from the intra AC or ME residual by a linear function or the like having the ME residual as a variable.

  Next, a method for controlling the prefilter unit 50 based on the encoding difficulty level will be conceptually described. In the present embodiment, a first stage for controlling the characteristics of the prefilter unit 50 based on the ratio of the encoding difficulty level to the bit rate, and a second stage for controlling the characteristics of the prefilter unit 50 based on the motion component, The two-stage control is performed.

  First, the first stage will be described. The amount of information to be reduced by the pre-filter unit 50 is when the pattern is complicated and contains a lot of spatial high-frequency components, or when there is a large amount of motion or random motion, that is, when the coding difficulty level is large. It is. If the pre-filter is used for other patterns, only side effects such as blurring due to the spatial filter and jerkiness (being awkward movement) due to the temporal filter are noticeable and not good.

  On the other hand, even for a video with a high degree of encoding difficulty as described above, when the bit rate in the image encoding apparatus is sufficiently high, that is, when the compression rate is low, the image quality is hardly deteriorated. No need to use. In this way, whether or not it is better to use the pre-filter varies depending not only on the encoding difficulty level but also on the bit rate.

  Therefore, in the present embodiment, control of filter characteristics such as ON / OFF of the pre-filter unit 50 and pass bandwidth is performed based on information indicating how much the input image signal encoding difficulty is relative to the bit rate. Like to do. Furthermore, in order to control the prefilter unit 50 based on the encoding difficulty level, it is preferable to look at the ratio of the encoding difficulty level to the bit rate in a plurality of pictures. This is because the encoding difficulty varies depending on the picture type. Therefore, in the present embodiment, a parameter x defined by the following equation is used as a parameter for controlling the prefilter unit 50 based on the encoding difficulty level.

  x = ΣDk / G

  In the above equation, Dk represents the coding difficulty level of picture k, k represents the coding order of the picture, Σ represents k = 1 to N (N is a picture of 1 GOP (group of pictures)). This means the sum of up to the number of copies. G is defined by the following equation.

  G = ([bit rate] × N) / [picture rate]

  In the above equation, [bit rate] represents the amount of data (bit amount) per second determined based on the transmission capacity of the communication line and the recording capacity of the recording medium, and [picture rate] represents per second. The number of pictures (for example, 30 in the NTSC zone and 25 in the PAL zone). Therefore, G represents the amount of data (bit amount) assigned to the time corresponding to N pictures, and x represents N data for the data amount (bit amount) assigned to the time corresponding to N pictures. This represents the ratio of the sum of the coding difficulty Dk of the minute picture. Note that x does not necessarily have to be obtained using N pictures, and x may be obtained using pictures of a larger or smaller number. However, it should be noted that if the number is much larger than N, the accuracy of the parameter x is lowered, and if the number is smaller than N, the variation of the parameter x is increased.

  In the present embodiment, as the first stage, for example, the characteristics of the prefilter unit 50 are controlled according to the size of the parameter x defined as described above. Specifically, for example, the pass bandwidth of the spatial filter 52 is changed by changing the tap coefficient (filter coefficient) in the digital filter that realizes the spatial filter 52 according to the size of the parameter x. By changing the value of the weighting coefficient K in the time filter 53 according to the magnitude of x, the pass bandwidth of the time filter 53 and the like are changed.

  FIG. 4 shows an example of the relationship between the size of the parameter x and the pass bandwidth of the spatial filter 52. In this figure, the horizontal axis represents the value of x, and the vertical axis represents the pass bandwidth of the spatial filter 52. In this example, five types of pass bandwidths of the spatial filter 52 are prepared: through (pass as it is; filter off), 5.0 MHz, 4.5 MHz, 4.0 MHz, and 3.5 MHz. As shown by the solid line), when 0 ≦ x <δ1, it is through, when δ1 ≦ x <δ2, it is 5.0 MHz, when δ2 ≦ x <δ3, it is 4.5 MHz, and when δ3 ≦ x <δ4 When 4.0 MHz and δ4 ≦ x, 3.5 MHz is selected. Note that Δ1 <Δ2 <Δ3 <Δ4.

  Next, the second stage will be described. When the characteristic of the prefilter unit 50 is determined only from the value of the parameter x as in the first stage, the following problems occur due to human visual characteristics. That is, for example, in a pattern in which an object with a fine pattern moves slowly, the spatial high frequency component is large, so that the degree of difficulty in encoding increases and the value of x also increases. However, if the prefilter is tightly applied (the pass bandwidth is narrowed), the fine pattern is lost. Since the human eye follows the slow movements well, such blur is perceived immediately, which worsens the visual image quality after decoding.

  On the other hand, for images with very fast movement, human visual characteristics are less likely to notice blur. Therefore, if you know that the video is fast moving, you should apply the prefilter tightly. This is because an attempt to express a high-frequency component in a fast moving image causes significant image quality degradation such as block distortion.

  Based on this concept, in the present embodiment, when the motion component representing the magnitude of the motion of the video is large, the prefilter is applied tightly, and when the motion component is small, the prefilter is applied loosely. Consider real-time implementation. The parameters for determining the motion component of the video include, for example, the latest B picture coding difficulty DB calculated based on the ME residual and the latest I picture coding difficulty calculated based on the intra AC. A ratio y = DI / DB with the degree DI is used. In a picture with a large motion, since the motion prediction error increases and the code amount of the motion vector itself also increases, the code amount of the B picture increases, and as a result, y = DI / DB decreases. Therefore, in the present embodiment, as the second stage, it is determined that the smaller the value of y, the larger the motion component, and the prefilter unit 50 is controlled so that the prefilter is tightly applied. Such control can be realized, for example, by changing the filter characteristic determined in the first stage according to the value of y.

  FIG. 4 shows an example in which the pass bandwidth of the spatial filter 52 determined in the first stage is changed to three stages according to the value of y. In this example, when y ≧ 4, the pass bandwidth of the spatial filter 52 determined in the first stage is not changed, and when 3 ≦ y <4, as shown by reference numeral 62, the first stage The passband width of the spatial filter 52 determined in step 1 is narrowed by 0.5 MHz, and when 2 ≦ y <3, the passband width of the spatial filter 52 determined in the first stage is obtained as indicated by reference numeral 63. Is narrowed by 1.0 MHz, and when y <2, the pass band width of the spatial filter 52 determined in the first stage is narrowed by 1.5 MHz as indicated by reference numeral 64.

  Next, the operation of the image coding apparatus according to this embodiment will be described. The input image signal S1 passes through the spatial filter 52 and the time filter 53 of the prefilter unit 50 in order, and then is input to the encoder control unit 11. In the encoder control unit 11, first, the image rearrangement circuit 21 rearranges the order of pictures (I picture, P picture, B picture) in accordance with the encoding order, and then the scan conversion / macroblocking circuit 22 Whether the frame structure or the field structure is determined, scan conversion and macroblock conversion are performed according to the determination result. Next, in the case of an I picture, the intra AC calculation circuit 23 calculates the intra AC to calculate the intra AC data S3. Is sent to the encoding control unit 15. The output data of the scan conversion / macroblocking circuit 22 is sent to the FIFO memory 12 and the motion detection circuit 14 via the intra AC arithmetic circuit 23.

  The FIFO memory 12 delays the input image data by the time required to calculate the encoding difficulty level of the N pictures following the encoded picture in the encoding difficulty level calculation unit 42. To the encoder 13. The motion detection circuit 14 detects the motion vector and sends it to the motion compensation circuit 39, and sends the ME residual data S4 to the ME residual calculation unit 41.

  In the case of an I picture, the encoder 13 performs the DCT by directly inputting the output data of the FIFO memory 12 to the DCT circuit 32 without taking the difference from the predicted image data in the subtracting circuit 31, and the quantizing circuit 33 performs the DCT. The DCT coefficient is quantized, the output data of the quantization circuit 33 is variable-length encoded by the variable-length encoding circuit 34, the output data of the variable-length encoding circuit 34 is temporarily held by the buffer memory 35, and a bit having a constant bit rate is stored. Output as compressed image data S2 consisting of a stream. Further, the output data of the quantization circuit 33 is inversely quantized by the inverse quantization circuit 36, the output data of the inverse quantization circuit 36 is subjected to inverse DCT by the inverse DCT circuit 37, and the output image data of the inverse DCT circuit 37 is converted. The signal is input to and held in the motion compensation circuit 39 via the adder circuit 38.

  In the case of a P picture, in the encoder 13, the motion compensation circuit 39 generates predicted image data based on the stored image data corresponding to the past I picture or P picture and the motion vector from the motion detection circuit 14. The prediction image data is generated and output to the subtraction circuit 31 and the addition circuit 38. Further, the difference between the output data of the FIFO memory 12 and the predicted image data from the motion compensation circuit 39 is obtained by the subtraction circuit 31, DCT is performed by the DCT circuit 32, the DCT coefficient is quantized by the quantization circuit 33, and the variable length is obtained. The output data of the quantization circuit 33 is variable length encoded by the encoding circuit 34, and the output data of the variable length encoding circuit 34 is temporarily held by the buffer memory 35 and output as compressed image data S2. The inverse quantization circuit 36 inversely quantizes the output data of the quantization circuit 33, the inverse DCT circuit 37 performs inverse DCT on the output data of the inverse quantization circuit 36, and the adder circuit 38 performs the inverse DCT circuit 37. The output data and the predicted image data are added and input to the motion compensation circuit 39 to be held.

  In the case of a B picture, in the encoder 13, the motion compensation circuit 39 converts the two past image data corresponding to the past and future I pictures or P pictures and the two motion vectors from the motion detection circuit 14. Based on this, predicted image data is generated, and the predicted image data is output to the subtracting circuit 31 and the adding circuit 38. Further, the difference between the output data of the FIFO memory 12 and the predicted image data from the motion compensation circuit 39 is obtained by the subtraction circuit 31, DCT is performed by the DCT circuit 32, the DCT coefficient is quantized by the quantization circuit 33, and the variable length is obtained. The output data of the quantization circuit 33 is variable length encoded by the encoding circuit 34, and the output data of the variable length encoding circuit 34 is temporarily held by the buffer memory 35 and output as compressed image data S2. Note that the B picture is not held in the motion compensation circuit 39.

  The buffer memory 35 sends generated bit amount data S5 representing the bit amount generated by the variable length encoding circuit 34 to the encoding control unit 15.

  In the encoding control unit 15, the encoding difficulty level calculation unit 42 calculates the encoding difficulty level from the intra AC data S 3 from the intra AC calculation circuit 23 and the ME residual calculated by the ME residual calculation unit 41. The parameter x is obtained, and the VBV buffer occupation amount calculation unit 43 calculates the VBV buffer occupation amount. Then, the target code amount determination unit 44 calculates a target code amount based on the parameter x and the VBV buffer occupancy amount calculated by the VBV buffer occupancy calculation unit 43, and sends the target code amount to the quantization index determination unit 45. The quantization index determination unit 45 determines a quantization index corresponding to the quantization characteristic value in the quantization circuit 33 so that the generated code amount in the encoder 13 becomes the target code amount determined by the target code amount determination unit 44. To the quantization circuit 33.

  Next, operations related to the control of the prefilter unit 50 will be described with reference to the flowchart of FIG. In FIG. 5, only control of the spatial filter 52 is shown for simplicity. In this operation, first, the encoding difficulty level is calculated by the encoding difficulty level calculation unit 42 to obtain the parameter x (step S101). Next, the pre-filter control unit 51 determines the pass bandwidth of the pre-filter (here, the spatial filter 52) as the first stage based on the value of the parameter x, for example, based on the relationship shown in FIG. Step S102). Next, the prefilter control unit 51 calculates the ratio y of the encoding difficulty level between the I picture and the B picture (step S103), determines the magnitude of y (step S104), and according to the magnitude of y. The following processing is performed. That is, in the example shown in FIG. 4, the passband width determined in the first stage is not changed when y ≧ 4, and the passband determined in the first stage is satisfied when 3 ≦ y <4. The width is narrowed by 0.5 MHz (step S105), and when 2 ≦ y <3, the passband width determined in the first stage is narrowed by 1.0 MHz (step S106), and when y <2, The pass band width of the spatial filter 52 determined in the first stage is narrowed by 1.5 MHz (step S107). Next, the prefilter control unit 51 sets the coefficient of the prefilter (in this case, the spatial filter 52) so as to have the pass bandwidth determined as described above (step S108), and returns to the main routine. . The above operation is performed for each picture.

  Up to this point, the control of the spatial filter 52 has been described, but the control of the temporal filter 53 is the same as the control of the spatial filter 52 except that the weighting coefficient K is used instead of the tap coefficient. When representing the characteristics of the time filter, for example, a weighting coefficient K is used on the vertical axis instead of the passband width in FIG. The smaller the value of K, the tighter the filter.

  As described above, according to the image encoding device according to the present embodiment, the characteristics of the prefilter unit 50 are adaptively controlled according to the encoding difficulty level of the input image while encoding the input image signal in real time. Therefore, it is possible to reduce visible distortion such as block distortion by reducing the amount of information when the input image is difficult to encode. When the input image is difficult to encode, the amount of information is reduced. It is possible to prevent the deterioration of the image quality due to the reduction of the image quality. Furthermore, according to the image coding apparatus according to the present embodiment, the pre-filter can be applied tightly when the motion component is large, and the pre-filter can be applied loosely when the motion component is small. While making it difficult, the impression of visual image quality after decoding can be improved.

  FIG. 6 is a block diagram showing a configuration of an image coding apparatus according to the second embodiment of the present invention. The image coding apparatus according to the present embodiment has been subjected to compression coding in the past as represented by TM5 (test model 5; ISO / IEC JTC / SC29 (1993)), which is well known as an MPEG compression algorithm. It is an example of the structure which performs the feedback type rate control which controls the generated code amount based on the obtained generated code amount.

  This image encoding apparatus receives an input image signal S1, inputs a prefilter unit 50 for reducing the amount of information, and an output signal of the prefilter unit 50, and performs preprocessing for compression encoding, etc. An encoder control unit 11 for performing the encoding, and an encoder as an encoding unit that inputs the output data of the encoder control unit 11 and performs compression encoding by an encoding method according to the picture type for each picture, and outputs compressed image data S2 13 and a motion detection circuit 14 that detects a motion vector based on the output data of the encoder control unit 11 and sends the motion vector to the encoder 13, and an encoding that controls the encoder 13 based on the generated bit amount data S7 output from the encoder 13. Based on the control unit 75 and the data from the encoding control unit 75, a prefilter is generated by a prefilter control signal S6. And a pre-filter control unit 51 for controlling the characteristics of 50. In this image encoding device, the configuration other than the encoding control unit 75 is the same as that of the first embodiment.

  The encoding control unit 75 calculates an encoding difficulty level based on the generated bit amount data S7 output from the variable length encoding circuit 34 of the encoder 13, and calculates the encoding difficulty level. Based on the encoding difficulty calculated by the unit 46, the target code amount determination unit 44 that determines the target code amount, and the generated code amount in the encoder 13 becomes the target code amount determined by the target code amount determination unit 44. As described above, the quantization index determining unit 45 that determines the quantization index corresponding to the quantization characteristic value in the quantization circuit 33 and sends the quantization index to the quantization circuit 33 is provided.

  Note that the prefilter control unit 51 controls the prefilter unit 50 based on the encoding difficulty level calculated by the encoding difficulty level calculation unit 46 in the encoding control unit 15.

  In the present embodiment, the global complexity of a picture that has been encoded is used as the encoding difficulty level calculated by the encoding difficulty level calculation unit 46. Global complexity is a parameter indicating the complexity of the screen. Specifically, the generated code amount at the time of picture compression encoding and the average quantization scale code at the time of picture compression encoding (quantization) (See, for example, page 111 of “Television Society Multimedia Selection MPEG” published by Ohm Co.).

  Other operations and effects in the present embodiment are the same as those in the first embodiment.

  Note that the present invention is not limited to the above-described embodiments. For example, when the prefilter unit 50 is controlled based on the encoding difficulty level, the parameters x and ratio y used in the embodiments, or the global In addition to complexity, it is possible to further improve the accuracy of predicting the difficulty of the pattern of the input image in the future by determining whether the coding difficulty level is increasing or decreasing. Accordingly, the prefilter unit 50 can be controlled more appropriately. The information of the change tendency of the encoding difficulty level can be obtained from, for example, a linear approximation of the encoding difficulty level obtained in time series by the least square method or the like and the inclination thereof. In this case, the information on the change tendency of the encoding difficulty level is obtained by, for example, the encoding difficulty level calculation units 42 and 46.

  Further, in the first embodiment, as shown in FIG. 2, the generated code amount is controlled based on the characteristics of the image data before compression encoding without providing another encoder in addition to the encoder 13. Although the feedforward type rate control is performed, the present invention provides a first-pass encoder different from the encoder 13 in order to estimate the amount of data after compression encoding. The present invention can also be applied to a configuration that performs feed-forward type rate control that controls the amount of generated code in the encoder 13 in the second pass based on the amount of data estimated by compression encoding. In this case, it is possible to obtain the encoding difficulty level based on the generated code amount obtained by compression encoding with the first-pass encoder.

  Further, in the configuration that performs feedback type rate control as in the second embodiment, encoding of several pictures in the future by linear approximation or the like based on the encoding difficulty level of a picture that has already been compression-encoded. The difficulty level may be predicted, and the pre-filter unit 50 may be controlled based on the predicted encoding difficulty level.

  In addition, when the feed forward type rate control configuration is used to determine the changing tendency of the encoding difficulty level by linear approximation, etc. When predicting the difficulty level of encoding, the continuity of the encoding difficulty level is lost at the time of a scene change. Therefore, at the time of a scene change, the process for obtaining the change tendency of the encoding difficulty level and the process of predicting the encoding difficulty level are performed before the scene change. If the process is completed and new processing is performed after a scene change, the accuracy will be improved.

  Needless to say, FIG. 4 shows an example of the characteristics of the prefilter, and the characteristics of the prefilter can be set as appropriate.

  In addition, the encoding difficulty level is not limited to the one using the intra AC, ME residual, global complexity, or the like described in the embodiment, but may be any other value that represents the difficulty level of picture encoding. The parameters may be used. Also, the parameter representing the motion component is not limited to the ratio y = DI / DB mentioned in the embodiment, but may be any other parameter as long as it represents the magnitude of motion in the input image data. For example, the ratio DI / DP between the encoding difficulty level DP of the latest P picture and the encoding difficulty level DI of the latest I picture may be used.

  In the embodiment, the prefilter unit 50 includes both the spatial filter 52 and the temporal filter 53. However, the present invention provides an image code having only one of the spatial filter and the temporal filter as the prefilter. The present invention can also be applied to a conversion apparatus.

It is a block diagram which shows the detailed structure of the image coding apparatus which concerns on the 1st Embodiment of this invention. It is a block diagram which shows the schematic structure of the image coding apparatus which concerns on the 1st Embodiment of this invention. It is a block diagram which shows the structure of the computer which implement | achieves the encoding control part and pre-filter control part in FIG. It is a characteristic view for demonstrating the function of the pre filter control part in FIG. 5 is a flowchart showing an operation related to control of a prefilter unit in the image coding apparatus according to the first embodiment of the present invention. It is a block diagram which shows the structure of the image coding apparatus which concerns on the 2nd Embodiment of this invention. It is a block diagram which shows an example of a structure of the image coding apparatus which has a pre filter.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 11 ... Encoder control part, 12 ... FIFO memory, 13 ... Encoder, 14 ... Motion detection circuit, 15 ... Encoding control part, 16 ... CPU, 17 ... ROM, 18 ... RAM, 23 ... Intra AC arithmetic circuit, 31 ... Subtraction Circuit 32... DCT circuit 33 33 quantization circuit 34 variable length coding circuit 35 buffer memory 36 inverse quantization circuit 37 inverse DCT circuit 39 motion compensation circuit 41 ME residual Calculation unit, 42... Encoding difficulty level calculation unit, 43... VBV buffer occupancy calculation unit, 44... Target code amount determination unit, 45 ..Quantization index determination unit, 50. ... spatial filter, 53 ... time filter.

Claims (8)

An encoding device that generates encoded image data by encoding image data,
A pre-filter for filtering the image data with a predetermined pass bandwidth;
Filter band changing means for changing the pass bandwidth so that the parameter representing the motion component of the image data becomes larger;
An encoding apparatus comprising: encoding means for encoding the image data filtered by the prefilter to generate the encoded image data. An encoding method for generating encoded image data by encoding image data,
A filter band changing procedure for changing so that the pass bandwidth of the pre-filter decreases as the parameter representing the motion component of the image data increases;
And a coding procedure for coding the image data filtered by the pre-filter to generate the coded image data. An encoding device that generates encoded image data by encoding image data,
A pre-filter for filtering the image data with a predetermined pass bandwidth;
Filter band changing means for changing the pass bandwidth so that the parameter representing the motion component of the image data becomes smaller;
An encoding apparatus comprising: encoding means for encoding the image data filtered by the prefilter to generate the encoded image data. An encoding method for generating encoded image data by encoding image data,
A filter band changing procedure for changing the pre-filter pass bandwidth so that the parameter representing the motion component of the image data becomes smaller;
And a coding procedure for coding the image data filtered by the pre-filter to generate the coded image data. A filtering device for filtering image data,
A pre-filter for filtering the image data with a predetermined pass bandwidth;
A filter processing apparatus comprising: a filter band changing unit that changes the pass bandwidth so that the parameter representing the motion component of the image data becomes larger. A filtering method for filtering image data with a predetermined pass bandwidth,
A filter processing method comprising: a filter band changing procedure for changing the pass bandwidth so that the parameter representing the motion component of the image data becomes larger. A filtering device for filtering image data,
A pre-filter for filtering the image data with a predetermined pass bandwidth;
A filter processing apparatus comprising: a filter band changing unit that changes the pass bandwidth so that the parameter representing the motion component of the image data becomes small. A filtering method for filtering image data with a predetermined pass bandwidth,
A filter processing method comprising: a filter band changing procedure for changing the pass bandwidth so that the parameter representing the motion component of the image data becomes small.

Images