JP2005175891A - Recorder - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a recorder capable of encoding a moving picture signal with higher image quality by configuring it so that the allocation of a code amount to the detected area of an object is reduced as compared to the other area in a present frame. <P>SOLUTION: The recorder is provided with an encoding means for encoding the dynamic image signal by inter-frame predictive encoding and a recording means for recording the encoded dynamic image signals on a recording medium. The encoding means detects the area of the object which is not to be present any more in the next and succeeding frames of the object present in the present frame in the dynamic image signals and reduces the allocation of the code amount to the detected area of the object in the present frame compared to the other area. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a data recording apparatus, and more particularly to a recording apparatus that performs digital data recording.

In the conventional technique, in the inter-frame difference encoding, an activity calculated for each block is used from the luminance value, and the allocation of the information amount to the complicated block is reduced (for example, refer to Patent Document 1).
JP-A-5-103315

  Even if the amount of code is reduced, if a higher quality image is encoded, the above conventional technique performs the same encoding as other portions even if it is not necessary, and therefore the amount of code is wasted. There was a problem of using it.

  In order to solve the conventional problems and achieve the above object, the present invention provides an encoding means for encoding an image signal by inter-frame prediction encoding, and records the encoded moving image signal on a recording medium. Recording means for detecting, in the moving picture signal, a region of a subject that does not exist after the next frame among subjects existing in the current frame, and the detected subject in the current frame The configuration is such that the allocation of the code amount to one area is reduced with respect to the other areas.

  The present invention can solve the above problems by providing the following configuration.

(1) encoding means for encoding a moving image signal by inter-frame prediction encoding;
Recording means for recording the encoded moving image signal on a recording medium,
The encoding means detects a region of a subject that does not exist after the next frame among subjects existing in the current frame in the video signal, and assigns a code amount to the detected region of the subject in the current frame A recording apparatus characterized in that the number is reduced with respect to other areas.

  (2) An imaging unit that outputs the moving image signal and the angular velocity sensor are provided, and the encoding unit determines a region of a subject that does not exist in the next frame based on output information of the angular velocity sensor. The recording apparatus according to (1), wherein:

  (3) The recording apparatus according to (1), wherein the encoding unit determines a region of a subject that does not exist in the next frame based on zoom information of the imaging unit.

  (4) The encoding means encodes the moving image signal by motion compensated predictive coding, and generates the next frame based on a motion vector between the current frame and the next frame at the time of the motion compensated predictive coding. The recording apparatus according to (1), wherein an area of a subject that does not exist is determined.

  As is clear from the above description, the present invention has an effect of encoding with higher image quality.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a block diagram that best represents the first embodiment of the present invention. In the figure, 101 is a lens, 102 is an image sensor, 103 is a camera signal processing circuit, 104 is a screen rearrangement circuit, 105 is a switch 1, 106 is a subtractor, 107 is A DCT circuit, 108 is a quantization circuit, 109 is a variable length coding circuit, 110 is an inverse quantization circuit, 111 is an IDCT circuit, 112 is an adder, and 113 is motion compensation. A prediction circuit, 114 is a switch 2, 115 is a buffer, 116 is a rate control circuit, 117 is a recording medium, and 118 is a button switch.

  There are two types of coding: intra coding and inter coding. Intra coding is coding using only data in a frame, and inter coding is coding including inter-frame prediction. In frame coding, there are an I picture obtained by intra-coding all data in a frame, a P picture including previous interframe prediction, and a B picture including previous and subsequent interframe prediction. A group from the I picture to a picture before the next I picture is called one group of pictures (GOP).

  The operation of this embodiment will be described with reference to FIG. An image input from the lens 101 is input to the camera signal processing circuit 103 as a digital image signal for each frame by the image element 102. In the camera signal processing circuit 103, a correction value is determined using the image signal, and the image data obtained by correcting the image signal is divided into a color difference signal and a luminance signal and input to the screen rearrangement circuit 104 in units of one frame. Further, the camera signal processing circuit 103 compares the previous frame with the current frame, compares the characteristic portions with each other, predicts how the screen moves, and calculates a screen movement vector. FIG. 4 is an image of n and n + k frames. (N and k may be integers and k may be a variable) A characteristic building is selected in n frames, and a search is made for where it has moved in n + k frames.

  Then, the output screen movement vector is output to the quantization circuit 108. The screen movement vector is calculated by n frames and n + k frames, and n + k frames are output in the form of moving by the screen movement vector with respect to n frames.

  The search is performed by comparing luminance values and color difference values. Search multiple data for accuracy.

  In addition, a characteristic item is selected based on a luminance value or a color difference value.

  Returning to FIG. The screen rearrangement circuit 104 has a memory capable of storing a plurality of frames, and outputs the images by changing the order of the input frames.

  FIG. 2 shows an example of the order of rearrangement by the screen rearrangement circuit 104.

  The screen rearrangement will be described with reference to FIG.

  The first frame, the second frame, the third frame,... Are input to the screen rearrangement circuit 104, the screen is rearranged, and the third frame, the first frame, the second frame,.

  FIG. 3 shows the generation order of the I picture, P picture, and B picture.

  Returning to FIG. 1, the normal operation will be described first.

  In the case of an I picture, the switch 1105 falls to the A side. The image data output from the screen rearrangement circuit 104 is input to the DCT circuit 107 via the switch 1105 and is orthogonally transformed. In the case of P and B pictures, the switch 1105 falls to the B side. The image data output from the screen rearrangement circuit 104 is subtracted by the subtractor 106 from the predicted image from the motion prediction circuit 113 via the switch 1105. This is because the subtractor 106 reduces the redundancy in the time axis direction. The image data whose redundancy in the time axis direction is reduced by the subtractor 106 is input to the DCT circuit 107 and orthogonally transformed.

  For all of the I, P, and B pictures, image data that has been orthogonally transformed by the DCT circuit 107 is quantized by the quantization circuit 108. The quantization circuit 108 makes a decision based on the screen movement vector output from the camera signal processing circuit 103, and coarsely quantizes the portion not displayed in the n + k frame. The quantized image data is input to the inverse quantization circuit 110 and the variable length coding circuit 109. The image data output from the screen rearrangement circuit 104 is input to the motion compensation prediction circuit 113.

  The quantized data is inversely quantized by the inverse quantization circuit 110, IDCT is performed by the IDCT circuit 111, and the IDCT image data is input to the motion compensation circuit 113 with the switch 2114 turned OFF. The motion compensation prediction circuit 113 outputs a predicted image for the next inter coding.

  The quantized data is input to the variable length encoding circuit 109, variable length encoded, and input to the buffer 115. The image data in the buffer 115 is recorded on the recording medium 117. Also, the amount of encoded data is calculated from the image data in the buffer 115, and the amount of encoded data is input to the rate control circuit 116.

  The rate control circuit 116 adjusts the generated bit amount by determining the allocated bit amount of the next picture or calculating the quantization coefficient of the next macroblock according to the determined bit rate.

  Comparing the n frame and the n + k frame and roughening the quantization of the portion not displayed in the n + k frame means that the code amount is not allocated to the shaded portion in FIG. 5 and the code amount is allocated to the other portion. . This improves the overall image quality. Further, regarding the n frame and the n + k frame, the n frame is I, P-picture, and the n + 1 frame is P, B-picture.

  In FIG. 6, n frames are I-pictures, and n + 1 frames are P, B-pictures. A portion not displayed in B-picture which is the next frame with respect to I-picture is a shaded portion, and a portion not displayed in P-picture is a hatched portion. The coarse quantization of the shaded portion is the same as in FIG. The rough quantization of the hatched portion is not displayed in the P-picture, and thus is a portion that is not necessary in the P-picture encoding. A code amount is not allocated to a portion used only in the second B-picture encoding, and a code amount is allocated to other portions to improve the overall image quality.

  A second embodiment of the present invention will be described with reference to FIG.

  The lens 101 and the image element 102 are the same as those in the first embodiment. In the camera signal processing circuit 103, a correction value is determined using the image signal, and the image data obtained by correcting the image signal is divided into a color difference signal and a luminance signal and input to the screen rearrangement circuit 104 in units of one frame. Further, the camera signal processing circuit 103 converts the angular velocity sensor signal, zoom information, and the like into vectors. This vector is used as the screen movement vector. The following is the same as in Example 1.

  FIG. 7 is a block diagram that best represents the third embodiment of the present invention.

  In the figure, 701 is a lens, 702 is an image sensor, 703 is a camera signal processing circuit, 704 is a screen rearrangement circuit, 705 is a switch 1, 706 is a subtractor, 707 is A DCT circuit, 708 is a quantization circuit, 709 is a variable length coding circuit, 710 is an inverse quantization circuit, 711 is an IDCT circuit, 712 is an adder, and 713 is motion compensation. A prediction circuit, 714 is a switch 2, 715 is a buffer, 716 is a rate control circuit, 717 is a recording medium, and 718 is a button switch.

  The operation of the second embodiment will be described with reference to FIG.

  The lens 701 and the image element 702 are the same as those in the first embodiment. In the camera signal processing circuit 703, a correction value is determined using the image signal, and the image data obtained by correcting the image signal is divided into a color difference signal and a luminance signal and input to the screen rearrangement circuit 704 in units of one frame. The screen rearrangement circuit 704, the switch 1705, the subtractor 706, and the DCT circuit 707 are the same as those in the first embodiment.

  All of the I, P, and B pictures are quantized by the quantization circuit 708.

  The quantized data is inversely quantized by the inverse quantization circuit 710, IDCT is performed by the IDCT circuit 711, and the image data subjected to IDCT is input to the motion compensation circuit 713 with the switch 2714 turned OFF. The motion compensation prediction circuit 713 outputs a prediction image for the next inter coding. In addition, the motion compensation prediction circuit 713 derives a screen movement vector from a dominant vector, an average value of vectors, and the like by performing motion search. The quantization circuit 708 makes a decision based on the screen movement vector output from the motion compensation prediction circuit 713, and coarsely quantizes the portion not displayed in the corresponding picture. The quantized image data is input to the inverse quantization circuit 710 and the variable length coding circuit 709. The image data output from the screen rearrangement circuit 704 is input to the motion compensation prediction circuit 713. The variable length coding circuit 109 and the subsequent steps are the same as those in the first embodiment.

  The screen movement vector is derived from the motion compensation prediction circuit 713, and as in the first embodiment, the code amount is not assigned to the portion that is not displayed, and the code amount is assigned to other portions to improve the overall image quality.

Example 1 and Example 2 block diagram best representing the present invention Diagram showing screen reordering Diagram showing encoding Frame diagram showing comparison of features Frame diagram showing the difference between the display and the previous screen A frame diagram showing a difference in display section of I-picture Example 3 block diagram best representing the present invention

Explanation of symbols

101 Lens 102 Image Sensor 103 Camera Signal Processing 104 Screen Rearrangement 105 Switch 1
106 Subtractor 107 DCT circuit 108 Quantization circuit 109 Variable length encoding circuit 110 Inverse quantization circuit 111 IDCT circuit 112 Adder 113 Motion compensation prediction circuit 114 Switch 2
115 Buffer 116 Rate Control Circuit 117 Recording Medium

Claims (4)

Encoding means for encoding a moving image signal by inter-frame prediction encoding;
Recording means for recording the encoded moving image signal on a recording medium,
The encoding means detects a region of a subject that does not exist after the next frame among subjects existing in the current frame in the video signal, and assigns a code amount to the detected region of the subject in the current frame A recording apparatus characterized in that the number is reduced with respect to other areas.   An imaging unit that outputs the moving image signal and the angular velocity sensor, wherein the encoding unit determines a region of a subject that does not exist in the next frame based on output information of the angular velocity sensor. The recording apparatus according to claim 1.   The recording apparatus according to claim 1, wherein the encoding unit determines a region of a subject that does not exist in the next frame based on zoom information of the imaging unit.   The encoding means encodes the moving image signal by motion compensation prediction encoding, and no longer exists in the next frame based on a motion vector between the current frame and the next frame at the time of motion compensation prediction encoding. The recording apparatus according to claim 1, wherein an area of the subject is determined.

Images