JP2004104598A - Video signal encoder and decoder, and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a video signal encoder of a reversible compression type which can sufficiently compress even a video signal inputted from a camera or the like and can completely decode the compressed signal into the original signal, a video signal decoder therefor, and a recording medium therefor. <P>SOLUTION: The encoder divides each pixel data (a) of the video signal corresponding to each time series frame and indicative of pixel brightness or color information at a predetermined bit position, into upper bit data (b) corresponding to the upper bits of the pixel data and lower bit data (c) corresponding to the lower bits of the pixel data; encodes the upper bit data on the basis of a prediction error from spatially adjacent pixel data; and encodes the lower bit data with such a minimum bit width that can avoid information missing of the lower bit data. Consequently, the encoder can perform its encoding operation by using the prediction error to the upper bit data as a main component. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
[Industrial applications]
The present invention relates to a video production field such as video production, storage of video data materials, and relaying of location materials, in particular, a field for producing high-definition video of a quality higher than NTSC, and a video recording / reproducing apparatus using a digital recording medium such as a DVD. The present invention relates to a reversible data compression technique suitable for use in the field, transmission of medical moving images in telemedicine, and the like where data modification is not desired.
[0002]
[Prior art]
Conventionally, various techniques have been used for compressing video signals. MPEG-1, MPEG-2, MPEG-4 and the like have been put to practical use as a technique for compressing and encoding a video signal. With such a compression encoding method, a video signal can be treated as small data, which contributes to the efficiency of data recording and transmission.
[0003]
[Patent Document 1]
JP-A-8-322050
[0004]
[Problems to be solved by the invention]
The above-described MPEG-1, MPEG-2, MPEG-4, etc. are all referred to as lossy encoding methods, and can be efficiently compressed. However, decoding involves a considerable deterioration in quality. However, the original signal cannot be completely reproduced. For this reason, in the field of video production such as video production, material storage, and location material relay, these encoding methods cannot be applied, and although inefficient, non-compressed storage and transmission methods are adopted. In particular, recently, the number of productions for handling high-definition HDTV video has increased, and the material capacity has become enormous, which has become a problem in managing work disks.
[0005]
Even for video signals, reversible (lossless) compression technology based on the run-length method is used for computer graphics, animation images, and artificial moving images. Can not. The reason is that a real shot image input by a camera or the like contains an unpredictable quantization noise component and basically has no data run.
[0006]
In view of the above points, the present invention can perform sufficient compression on a video signal input by a camera or the like, and can completely decode the original video signal during decoding. It is an object to provide an encoding device, a decoding device, and a recording medium for a video signal of a reversible compression system.
[0007]
[Means for Solving the Problems]
In order to solve the above problem, the present invention provides a method for reproducing all the pixel data for a video signal composed of pixel data corresponding to luminance or color information of pixels forming each frame of a time series. An encoding device that compresses the information amount is configured by dividing each bit data configuring the pixel data at a predetermined bit position, and configured by upper bit data configured by upper bits of pixel data and lower bits of pixel data. Data separating means for separating the lower-order bit data into upper-order bit data, and upper-order bit encoding for performing coding on the upper-order bit data based on a prediction error from pixel data of a spatially adjacent pixel. Means, and low-order bit encoding means for encoding the low-order bit data with a minimum necessary bit width on condition that information is not lost. And said that the content was.
[0008]
According to the present invention, bit data constituting each pixel data of each frame of a video signal is separated into upper bit data and lower bit data, and upper bit data is separated from pixel data of spatially adjacent pixels. Since the encoding is performed based on the prediction error and the encoding is performed separately from the lower bit data, the thermal noise component of the circuit used when digitizing the video signal is separated as the lower bit. Since encoding using the prediction error can be performed on the higher-order bit data as a component, encoding with high compression efficiency can be performed.
[0009]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
(Device configuration)
FIG. 1 is a configuration diagram showing an embodiment of a video signal encoding device according to the present invention. In FIG. 1, reference numeral 10 denotes an inter-color signal operation unit, 20 denotes a frame difference operation unit, 30 denotes a signal flat part encoding unit, 40 denotes a data separation unit, 50 denotes an upper bit encoding unit, 51 denotes a linear prediction unit, and 52 denotes a linear prediction unit. Prediction error calculating means, 53 is a variable bit length coding means, 54 is a look-up table creating means, 55 is bit data converting means, 56 is bit data coding means, and 60 is lower bit coding means.
[0010]
In FIG. 1, the inter-color-signal calculating means 10 is configured such that when each pixel in one image frame has a luminance value corresponding to a plurality of colors, that is, when one image frame includes a plurality of luminance value planes, It has a function of giving a difference from the reference value to the luminance value of the other two colors using one color as a reference value. The frame difference calculation means 20 has a function of calculating a difference between each pixel data between consecutive frames. The signal flat part encoding means 30 has a function of detecting a flat part in which the value of pixel data is constant for each frame subjected to the difference operation processing, and efficiently encoding the same. The data separating means 40 has a function of separating each pixel data constituting the frame on which the signal flat portion has been processed into upper bit data and lower bit data at a predetermined position. The upper bit encoding unit 50 has a function of efficiently encoding the upper bit data separated by the data separating unit 40. The lower bit encoding means 60 has a function of efficiently encoding the lower bit data separated by the data separating means 40. Each component illustrated in FIG. 1 is actually realized by a computer and a dedicated software program executed by the computer.
[0011]
(Processing operation)
Next, the processing operation of the video signal encoding device shown in FIG. 1 will be described. First, an image is captured and digitized by a TV camera or the like. Thereby, a digital video signal is obtained. Such a digital video signal has a plurality of frames (still images) per second, and each frame has luminance values of three colors of R, G, and B. In the present embodiment, description will be made assuming that 30 frames are allocated per second and 8 bits are allocated to each color of each pixel.
[0012]
The digital video signal digitized in this way is subjected to a difference operation between color signals by the inter-color-signal operation means 10. Here, the state of the difference calculation between the color signals will be described with reference to FIG. FIG. 2A schematically shows a digital video signal. As shown in the figure, one frame of the video signal is composed of three planes of RGB. The inter-color-signal calculating means 10 performs a difference calculation between planes for each frame of the video signal as shown in FIG. Specifically, a difference operation is performed on the pixel data at the same coordinates of the R plane and the G plane with reference to the G plane, and the value is set as a new value for each pixel of the R plane, and the same coordinates of the B plane and the G plane are calculated. Is performed, and the value is used as a new value for each pixel on the B plane. Here, the G plane is used as a reference because the luminance signal is mainly composed of G, and the spatial resolution of the RG and BG color difference signals is about 1/2 of that of the G signal. That is, when the two planes of R and B are used as the difference signal, the luminance difference between adjacent pixels is reduced, and the prediction error at the time of encoding described later is reduced. As a result, each frame is in a state as shown in FIG. When the input digital video signal is input in the form of a Y signal, which is a luminance signal, and Pb, Pr, which are color difference signals, such as Y, Pb, Pr, the processing by the inter-color signal calculating means 10 is performed. And the processing by the inter-frame difference calculating means 20 is performed directly.
[0013]
Next, the inter-frame difference calculation means 20 performs a difference calculation between adjacent frames on the digital video signal calculated between the color signals. More specifically, a difference operation is performed between pixels having the same coordinates in each plane between frames, and a process of setting the value as a new value for each pixel in a subsequent frame is performed. In this case, only the start frame that is temporally the first is not subjected to the difference calculation. As a result, the digital video signal shown in FIG. 2B is in a state as shown in FIG. That is, frame F ₁ Is not changed by the processing of the inter-frame difference calculating means 20, but each plane of the frame F2 has the frame F ₁ And frame F ₂ Of the frame F _n Each plane has a frame F _n-1 And frame F _n Are given. However, when the scene changes in the video, the illumination, natural light, and the like also change, and the luminance value also changes greatly. In such a case, even if the difference calculation is performed, the value after the difference becomes rather large, and does not contribute to the compression of the data amount. Therefore, in the present embodiment, when the value of each pixel in the updated frame increases to a predetermined level or more compared to the value of each pixel in the frame before update, a process of returning the frame to the frame before update is performed. I have. As a result, a plurality of start frames exist in the video signal.
[0014]
Subsequently, the signal flat part encoding means 30 performs a signal flat part process on each plane of each frame. The signal flat portion refers to a portion where the signal level has a uniform value. However, except for computer graphics images and the like, the signal level has a uniform minimum value of “0” or a maximum signal level value (for example, 255). ) Is usually a candidate. In actual shooting images, such locations are generally rare, but the illumination is totally dark and the input of the A / D converter is below the zero calibration level. This may occur when the image is backlit and the A / D converter is saturated. Regardless of whether the signal level is “0” or the signal level is the maximum value, the same signal level is continuously recorded in a predetermined area (a predetermined number of pixels) in the signal flat portion. For this reason, this part is data that can be easily compressed. More specifically, pixel coordinates (x, y) and pixel data values (luminance value, RG plane, BG plane in the case of a G plane) of an upper left point and a lower right point of a rectangular area serving as a signal flat portion Is recorded as the flat portion information separately from each plane of each frame. The signal flat portion is deleted from each plane of each frame. This is schematically shown in FIG. In FIG. 3A, a shaded portion indicates a signal flat portion. By the processing of the signal flat part encoding means 30, the signal flat part is deleted from each plane of the frame. However, the deleted signal flat part is recorded in a format as shown in FIG. The signal flat portion information is recorded as the pixel coordinates (x, y) of the upper left point, the pixel coordinates (x, y) of the lower right point, and the value of pixel data for each signal flat portion. In this embodiment, since 8 bits are assigned to each pixel of each plane, a signal flat portion is easily formed when pixel data is 0 or 255. However, the signal flat part encoding means 30 does not unconditionally process the signal flat part. This is because the purpose of the present invention is to compress data, and it is meaningless if the flat portion information is larger than the amount reduced from each plane. Therefore, only when the flat portion has a size equal to or larger than a predetermined value, flat portion information is created and deleted from each plane.
[0015]
As described above, after the difference calculation processing between the respective planes (between the color signals), the difference calculation processing between the respective frames, and the encoding processing of the signal flat part are performed, next, the data separating unit 40 Separates upper and lower bits of data. In the present embodiment, when digitizing a captured image, each color is quantized with a quantization bit number of 8. Therefore, the digital data is assigned 8 bits for each pixel of each plane. In this case, in the present embodiment, the upper bits are separated into 6 bits and the lower bits are separated into 2 bits. This separation is basically caused by quantization noise (fractional fluctuation component when converting from analog to digital numerical value) due to thermal noise of a circuit used for digitizing a video signal such as an A / D converter. Perform to separate. Therefore, lower bits considered as quantization noise are separated. The extent to which the lower bits are separated depends on the shooting environment and the characteristics of the circuit used, but it is usually desirable to set the number of quantization bits to about 1/4. Therefore, in this case, 2 bits, which are 1/4 of 8 bits, are separated as lower bits.
[0016]
Here, the state of data separation by the data separation means 40 is schematically shown in FIG. In FIG. 4, H indicates upper bit data, and L indicates lower bit data. FIG. 4A shows pixel data before separation. The sample data is separated by the data separating means 40 into upper bit data shown in FIG. 4B and lower bit data shown in FIG. 4C. The pixel data separated in this way will be separately processed thereafter.
[0017]
(Encoding of upper bits)
The upper bit data is processed by the upper bit encoding means 50. In the upper bit encoding unit 50, after the linear prediction unit 51 and the prediction error calculation unit 52 calculate the prediction value and the prediction error of each pixel based on the two neighboring pixels, each bit replaced with the prediction error is calculated. The variable length bit encoding means 53 converts the data to a variable length. Here, a calculation method of the prediction error will be described with reference to FIG. Here, assuming that prediction is performed in the scanning line direction (x-axis direction), consider the case where the upper bit value is in a state as shown in FIG. 5A. In FIG. 5A, the horizontal axis is the x-axis coordinate on the plane, and the vertical axis is the upper bit value G (x). The line segment at each coordinate value indicates the value of the upper bit value G (x) of the pixel at each coordinate value. In such a state, when calculating the prediction error E (x) at the pixel at the coordinates (x, y), the higher-order bit value G (x-1) at the immediately preceding pixel x-1 and the pixel x−2 at the immediately preceding pixel x−1 2 using the upper bit value G (x−2) in accordance with the following [Equation 1].
[0018]
[Formula 1]
E (x) = G (x) −2 × G (x−1) + G (x−2) −E (x−1) / 2
[0019]
In the above [Equation 1], “2 × G (x−1) −G (x−2)” is a linear prediction component based on the upper bit values of the two immediately preceding pixels, and is calculated by the linear prediction unit 51. . The prediction error calculation unit 52 uses the calculated linear prediction component and the prediction error “E (x−1) / 2” (error feedback component) calculated for the immediately preceding pixel to calculate a prediction error E ( x) is calculated. The prediction error is calculated for all the pixels, and the prediction error is recorded instead of the upper bit value of the pixel data.
[0020]
This will be described based on the upper bit value of the pixel data shown in FIG. First, each prediction error Eo (x) is calculated without adding an error feedback component. As shown in FIG. 5B, when calculating the prediction error Eo (x) of the pixel x, the upper bit value G (x−1) of the immediately preceding pixel x−1 and the pixel G−2 of the immediately preceding pixel x−2 are calculated. The prediction error Eo (x) is based on the difference between the value taken by the prediction line connecting the upper bit values G (x−2) at the coordinate x and the upper bit value G (x) at the coordinate x (indicated by the thick dotted line in the figure). Is calculated. The prediction error Eo (x + 1) is calculated in the same manner for the pixel x + 1 and thereafter. The calculated prediction error Eo (x) is as shown in FIG. As can be seen from a comparison between FIG. 5A and FIG. 5C, the range in which the value fluctuates is greatly narrowed, and data compression becomes more convenient. Subsequently, based on [Equation 1], the prediction error Eo (x) is subtracted by 50% of the prediction error E (x-1) obtained by adding the correction in the immediately preceding pixel x-1, and the error feedback processing is performed. FIG. 5D shows the added result. Compared with FIG. 5C, the reduction of the prediction error in the pixels x + 1 and x + 2 is remarkable. Conversely, although the prediction error is increased in the pixels x + 3 and x + 4, the prediction error is reduced on average, and the range in which the value fluctuates is further narrowed as compared with FIG. 5A, and the data compression effect is improved.
[0021]
Although the prediction error is calculated as described above, in the example of FIG. 5, the prediction is performed according to the x-axis direction of the image, that is, the scanning line direction. In the present invention, in order to further increase the prediction accuracy, a process of selecting a direction most suitable for calculating a prediction error from among a plurality of directions is performed. Here, which of the three directions of the scanning line direction, the sub-scanning line direction, and the diagonal direction as shown in FIG. Specifically, when determining the direction for obtaining the prediction error of the pixel (x, y), first, the prediction error Eh by the two pixels immediately before the immediately preceding pixel (x-1, y) in the scanning line direction. (X-1, y), a prediction error Ev (x, y-1) by two pixels immediately before the immediately preceding pixel (x, y-1) in the sub-scanning line direction, and a immediately preceding pixel (x-y A prediction error Ed (x-1, y-1) by two pixels immediately before (1, y-1) is obtained. The prediction error Eh (x, y) in the scanning line direction, the prediction error Ev (x, y) in the sub-scanning line direction, and the prediction error Ed (x, y) in the diagonal direction are expressed by the following [Equation 2] to [Equation 2], respectively. Equation 4].
[0022]
[Formula 2]
Eh (x, y) = G (x, y) −2 × G (x−1, y) + G (x−2, y) −E (x−1, y) / 2
[0023]
[Equation 3]
Ev (x, y) = G (x, y) −2 × G (x, y−1) + G (x, y−2) −E (x, y−1) / 2
[0024]
[Equation 4]
Ed (x, y) = G (x, y) -2 × G (x-1, y-1) + G (x-2, y-2) -E (x-1, y-1) / 2
[0025]
Then, the direction of the pixel that minimizes the prediction error is set as the prediction direction of the pixel (x, y). For example, when Ev (x, y-1) is the smallest among Eh (x-1, y), Ev (x, y-1), and Ed (x-1, y-1), sub-scanning is performed. A prediction error is determined for the line direction. That is, the prediction error of the pixel (x, y) is calculated as Ev (x, y) using the values of the pixel (x, y-1) and the pixel (x, y-2) according to the above [Equation 3]. Is calculated. Similarly, the same process is performed for all pixels in all frames and all planes, and the value of each pixel is replaced with a prediction error value.
[0026]
Some of the pixel data on the plane is collectively processed as a signal flat portion area by the signal flat portion coding means 30 and is not coded as each pixel. For such a pixel, the prediction error is not calculated, but is used for calculating the prediction error of a pixel that does not belong to the signal flat part. For example, in a plane as shown in FIG. 6B, it is assumed that the diagonal direction is determined to be optimal in obtaining the prediction error of the pixel A which does not belong to the signal flat portion. In this case, pixel data (upper bits) of pixel B and pixel C belonging to the signal flat portion and a prediction error are required. In such a case, the prediction error of the pixel A is calculated using the original data of the pixels B and C.
[0027]
Next, the variable bit length coding unit 53 performs a conversion of the bit configuration in order to represent the higher-order bit data recorded with the prediction error value with a smaller data amount. First, a look-up table used for converting the bit configuration is created. Specifically, a histogram of each pixel data is calculated for all frames and all planes. The pixel data has positive and negative values due to the replacement with the prediction error. Here, a histogram is calculated for each absolute value. As a result, a bit pattern with a smaller number of bits is assigned in order from the pixel value with the highest appearance frequency. At this time, there is a rule on the bit pattern to be assigned, the most significant bit is always set to “1”, and the bit pattern including the bit pattern of “01” is prohibited. Therefore, the minimum value of the bit pattern is 1 bit “1”. FIG. 7C shows an example of the created lookup table.
[0028]
Using the look-up table created as described above, 6-bit fixed-length continuous high-order bit data is converted into a variable-length bit pattern. Since the length is variable, it is necessary to distinguish the breaks of each data after conversion. Therefore, in the present embodiment, a one-bit separator bit “0” is inserted between each data. In the look-up table, the bit pattern including “01” is prohibited by separating the data with a “01” pattern consisting of a separator bit “0” and the most significant bit “1” of the bit data that follows. This is because if there is a “01” pattern in the bit data, it will be erroneously determined to be a break. In the look-up table, since the conversion is performed using the absolute value of the value taken by the upper fixed-length bit data, it is necessary to record positive and negative values. Therefore, inverted positive / negative data having a value that does not overlap with the bit data is inserted. For example, if the variable-length bits before the positive / negative inverted data are positive, the variable-length bits thereafter are recognized as negative. FIGS. 7A and 7B schematically show how data is converted by the variable bit length coding means 53. FIG. 7 (a) and 7 (b) correspond to the upper bit portion of the pixel data, and FIG. 7 (a) shows a state in which fixed-length upper bit data is continuously recorded. The upper bit data as shown in FIG. 7A is converted as shown in FIG. 7B using the look-up table shown in FIG. 7C.
[0029]
(Encoding of lower bits)
On the other hand, the lower bit data is processed by the lower bit encoding means 60. Specifically, the lower two bits of data separated by the data separating means 40 are continuously arranged.
[0030]
(Recording of code data)
The code data obtained as described above is as shown in FIG. That is, upper variable length bit data, lower fixed length bit data, look-up table, frame structure information, signal flat portion information, and color signal information. This data is recorded in a format suitable for the recording medium to be recorded.
[0031]
(Decryption)
Next, decoding of the encoded data encoded by the encoding device will be described. FIG. 9 is a functional block diagram showing a configuration of a video signal decoding device according to the present invention. 9, reference numeral 91 denotes data reading means, 92 denotes upper bit conversion means, 93 denotes lower bit conversion means, 94 denotes data integration means, 95 denotes signal flat section insertion means, 96 denotes frame restoration means, and 97 denotes color signal restoration means. It is. The configuration illustrated in FIG. 9 is realized by a computer and a dedicated software program mounted on the computer.
[0032]
Subsequently, the processing operation of the decoding device shown in FIG. 9 will be described. First, the data reading means 91 reads a recording medium on which code data as shown in FIG. 8 is recorded. The data reading means 91 transfers the upper variable length bit data and the look-up table of the read data to the upper bit conversion means 92 and also transfers the lower fixed length bit data to the lower bit conversion means 93. The upper bit conversion means 92 restores the fixed-length upper fixed-length bit data, that is, the linear prediction error E (x, y), from the upper variable-length bit data by referring to the look-up table. The 6-bit fixed-length high-order bit data G (x, y) is sequentially restored based on an equation in which the left-hand term and the right-hand first term are exchanged. In addition, the lower-order bit conversion means 93 restores lower-order 2-bit fixed-length bit data from lower-order fixed-length bit data. Subsequently, the data integrating means 94 integrates the upper fixed-length bit data and the lower bit data. Specifically, a process of extracting 6 bits from the upper bit data and extracting 2 bits from the lower bit data and sequentially integrating them is performed.
[0033]
At this point, the pixel data of each plane has a format in which each pixel is represented by 8 bits. The signal flat portion insertion means 95 inserts a signal flat portion at a predetermined position using signal flat portion information as shown in FIG. 3B, and is restored to a plane in which all pixels have values. At this point, each frame and plane are in a state as shown in FIG. The frame restoring means 96 uses the frame structure information to recognize which frame is the first frame and which frame has been subjected to the difference operation with respect to each frame, and restores the original frame. As a result, a frame as shown in FIG. 2B is restored. Finally, the color signal restoring means 97 recognizes which color signal remains the original and which color signal is difference information from which color signal by using the color signal To restore. As a result, a digital video signal in which an analog signal is digitized is restored without data loss.
[0034]
【The invention's effect】
As described above, according to the present invention, all the pixel data can be reproduced with respect to a video signal composed of pixel data corresponding to luminance or color information of pixels forming each frame of a time series. An encoding device for compressing the amount of information is divided into upper bit data composed of upper bits of pixel data and lower bits of pixel data by dividing each bit data constituting pixel data at predetermined bit positions. Data separating means for separating the data into lower-order bit data, and upper-bit encoding means for performing coding on the upper-order bit data based on a prediction error from pixel data of a spatially adjacent pixel And low-order bit encoding means for encoding the low-order bit data with a minimum necessary bit width on condition that information is not lost. With the quantization error component, which is difficult to predict when digitizing the video signal, separated as the lower bits, high-order bit data, which is the main component, can be coded using the prediction error, resulting in compression efficiency. Is achieved.
[Brief description of the drawings]
FIG. 1 is a functional block diagram showing an embodiment of a video signal encoding device according to the present invention.
FIG. 2 is a diagram showing a state of processing by an inter-color-signal operation unit 10 and a frame difference operation unit 20;
FIG. 3 is a diagram illustrating a state of processing by a signal flat part encoding unit 30;
FIG. 4 is a diagram showing a state of processing by a data separating unit 40;
FIG. 5 is a diagram illustrating a state of a prediction error calculation process performed by a higher-order bit encoding unit 50;
FIG. 6 is a diagram illustrating a positional relationship between pixels when performing a prediction error calculation process.
FIG. 7 is a diagram showing a state of data conversion by a variable bit length encoding means 53.
FIG. 8 is a diagram showing code data obtained by the video signal coding apparatus according to the present invention.
FIG. 9 is a functional block diagram showing an embodiment of a video signal decoding apparatus according to the present invention.
[Explanation of symbols]
10... Arithmetic means between color signals
20 ... Frame difference calculation means
30 ... Signal flat part coding means
40 ... Data separation means
50: Upper bit encoding means
51 ··· Linear prediction means
52... Prediction error calculation means
53 ... variable bit length coding means
54 ... Lookup table creation means
55 ··· Bit data conversion means
56... Bit data encoding means
60 ... Lower bit encoding means
91 Data reading means
92 ··· Upper bit conversion means
93 ... Lower bit conversion means
94 ・ ・ ・ Data integration means
95 ... Signal flat part insertion means
96 ・ ・ ・ Frame restoration means
97 ... Color signal restoration means

Claims (17)

For a video signal composed of pixel data corresponding to luminance or color information of pixels forming each frame of a time series, an encoding device for compressing an information amount so as to reproduce all the pixel data, ,
Each bit data constituting the pixel data is divided at a predetermined bit position to separate upper bit data composed of upper bits of the pixel data and lower bit data composed of lower bits of the pixel data. Data separation means;
For the upper bit data, upper bit encoding means that performs encoding based on a prediction error from pixel data of a spatially adjacent pixel,
For the lower bit data, a lower bit encoding means that performs encoding with a minimum necessary bit width under a condition that information is not lost,
A video signal encoding device comprising: In claim 1,
The higher-order bit encoding unit performs linear prediction from upper-order bit data corresponding to two neighboring pixels to obtain a prediction value;
Prediction error calculating means for calculating an error value between the predicted value and the target upper bit data,
Variable bit length encoding means for encoding the bit data of the error value with different bit lengths,
A video signal encoding device comprising: In claim 2,
The video signal encoding apparatus according to claim 1, wherein said prediction error calculating means has an error feedback function of subtracting a value obtained by multiplying an error value obtained for a previous pixel by a predetermined ratio. In claim 2,
The linear prediction means may select two neighboring pixels as candidates for neighboring pixels located in three directions of a scanning line direction, a sub-scanning line direction, and a diagonal direction with respect to a target pixel for which a prediction error is to be calculated. A video signal encoding apparatus which performs the calculation by selecting a direction in which an error value calculated by the means is the smallest. In claim 4,
The selection of the direction in which the error value becomes the smallest is to select the direction in which the prediction error of the pixel located immediately before by one pixel in three directions with respect to the target pixel for which the prediction error is calculated is the smallest. A video signal encoding device characterized by the above-mentioned. In claim 2,
The variable bit length encoding means,
Lookup table creation means for creating a lookup table described with a minimum bit length such that the most significant bit of the converted bit data is 1 based on the histogram of the target bit data;
Bit data conversion means for performing conversion using the lookup table for the target bit data,
Bit data encoding means for encoding bit data so as to insert a predetermined number of bits of bit data between the converted bit data;
A video signal encoding device comprising: In claim 6,
The look-up table performs a conversion so that the higher the frequency of appearance of the target bit data, the smaller the bit length. If the division bit data is 1 bit 0, it does not include a 2-bit pattern of 01. A video signal encoding device characterized by being converted into such bit data. In claim 6 or claim 7,
When the target bit data has a positive or negative sign, the look-up table is converted such that the higher the frequency of appearance of the absolute value of the target bit data, the smaller the bit length, and the absolute value of the target bit data. Is defined as positive / negative inverted data, and the absolute value of the bit data is encoded by inserting the positive / negative inverted data immediately before the bit data of the encoded bit data whose polarity is inverted. A video signal encoding apparatus characterized in that: In claim 1,
In a set of pixel data forming the frame, a rectangular area in which the value of the pixel data is continuously the same in the scanning line direction or the sub-scanning line direction is extracted, and two diagonal points of the extracted rectangle are extracted. Two pairs of two-dimensional pixel coordinate values, and signal flat portion encoding means for encoding three values of pixel data,
A video signal encoding apparatus, wherein the data separation unit performs processing on pixel data excluding a rectangular area processed by the signal flat unit encoding unit. In claim 1,
When the pixel data has luminance values corresponding to three kinds of colors at the same coordinate value of one frame, the luminance value of one color is used as a reference value, and the luminance value of the other two colors is used as the reference value. Having a color signal calculation means for updating the pixel data so as to give a difference value with
A video signal encoding apparatus, wherein the data separation means performs processing on pixel data processed by the color signal calculation means. In claim 1,
The pixel data is subjected to an operation between frames to be replaced with a difference value from a luminance value of a corresponding color of a pixel located at the same coordinate value of the immediately preceding frame, and all frames subsequent to the start frame are processed. Having frame difference calculating means for updating the pixel data of
A video signal encoding apparatus, wherein the data separation means performs processing on the pixel data processed by the frame difference calculation means. In claim 11,
The frame difference calculation means, when the value of each pixel of the updated frame is greater than or equal to a predetermined level compared to the value of each pixel of the frame before the update, returns the frame to the frame before the update, A video signal encoding device, wherein one video signal is provided with a plurality of start frames. 13. A recording medium for a video signal in which code data output by a video signal encoding device having a configuration according to any one of claims 1 to 12 is recorded for a given video signal. Data reading means for reading code data recorded on a recording medium,
Upper bit conversion means for restoring upper fixed-length bit data based on the read code data,
Lower bit conversion means for restoring lower fixed length bit data based on the read code data,
The upper fixed-length bit data, data integration means for integrating the lower fixed-length bit data to restore pixel data,
Using signal flat portion information included in the read code data, signal flat portion inserting means for inserting a signal flat portion into pixel data corresponding to each luminance or color information of each frame composed of the pixel data; ,
Frame restoration means for restoring each luminance or chrominance signal information in which the signal flat part is inserted, according to the frame structure information included in the read code data, into luminance or chrominance signal information for each frame unit;
Using color signal information included in the read code data, color signal restoration means for restoring the original color signal information,
A video signal decoding device comprising: Data reading means for reading code data recorded on a recording medium,
Upper bit conversion means for restoring upper fixed-length bit data using a prediction error from pixel data of spatially adjacent pixels based on the read code data,
Lower bit conversion means for restoring lower fixed length bit data based on the read code data,
The upper fixed-length bit data, data integration means for integrating the lower fixed-length bit data to restore pixel data,
A video signal decoding device comprising: A program for causing a computer to function as a video signal encoding device having a configuration according to any one of claims 1 to 12. A program for causing a computer to function as a video signal decoding device having the configuration according to claim 14 or 15.

Images