JP2005151485A - Video signal coding device and video signal recording device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device which makes constant a total recording time for a recording medium by suppressing variation of a coding volume when images of a plurality of channels are input, and recorded in the medium by compression encoding. <P>SOLUTION: Inputted video signals of the plurality of channels are transformed into one video signal, rearranged in a unit of predetermined frame number for each channel, and are compressed and coded in a unit of the predetermined frame number. The coding parameter of encoding is acquired in a unit of GOP, and used for compression encoding of the next picture in a unit of GOP of the channel. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a video signal encoding apparatus that suitably compresses and encodes image signals from a plurality of cameras and the like, and a video signal recording apparatus that records the compressed and encoded video signals on a recording medium.

  In a crime prevention / security monitoring system using a video camera, a plurality of cameras are often connected to a single recording device so that multipoint monitoring can be performed simultaneously. At that time, if the recording apparatus tries to record a plurality of videos simultaneously, the recording capacity increases in proportion to the number of cameras, and the recording apparatus becomes expensive. Therefore, a method is known in which a plurality of camera images are sequentially switched for each frame and recorded in time series (for example, Patent Document 1). According to this, although the recording rate per camera is lowered, the capacity of the recording medium is not increased, and the recording can be performed for a long time.

JP 2001-103471 A

  The above-described conventional technology can suppress the influence of other camera images having no frame correlation, and is effective in preventing undesired image quality degradation and fluctuation of the generated code amount. However, even when compression at a fixed bit rate is used, due to the characteristics of MPEG compression, some variation is unavoidable due to the fineness of the picture and the amount of motion of the video, and the target code amount may not always be obtained. . In other words, in an actual video, there may be cases where it cannot be absorbed by only rate control that is completed within one group of pictures (hereinafter referred to as GOP). Fluctuation is large when the value is set low. In addition, in surveillance applications that are used at the same angle of view for a long time, a specific camera (channel) may be constantly transmitted at a higher bit rate, and an error in the total recording time is unavoidable.

  An object of the present invention is to solve the above-described problems and to make the total recording time for the recording medium constant by suppressing the fluctuation of the encoding amount.

  In order to solve the above-described problem, an encoding parameter related to a rate or a generated code amount is acquired for each compression encoding for 1 GOP, and is then used when the channel is subjected to compression encoding.

  In order to solve the above problem, an average value of past generated code amounts is obtained for each channel, and a generated code amount target value for each GOP or picture is set and controlled.

  According to this configuration, when a plurality of image signals are input, the encoding amount is stabilized, and thus the total recording time for the recording medium is made constant.

  According to the present invention, it is possible to provide an easy-to-use video signal compression apparatus and recording apparatus with little error in recording time on a medium.

  The purpose of making the total recording time for the recording medium constant when a plurality of image signals are input is realized by controlling so as to suppress fluctuations in the encoding amount.

A first embodiment of the present invention will be described with reference to FIGS.
FIG. 1 is an overall block diagram of a video signal encoding apparatus and recording apparatus according to the present invention. FIG. 2 is a block diagram showing a detailed configuration of the main part of FIG. 3 and 4 schematically show signal processing in each block. Reference numerals 101a to 101d denote video input cameras (in this case, four channels are shown), reference numeral 102 denotes multi-input selection / sequentialization means, reference numeral 103 denotes channel distribution / reordering means, reference numeral 104 denotes image compression / encoding means, and reference numeral 105 Is a system control CPU, and 106 is a recording means.

  The cameras 101a to 101d are installed so as to be able to monitor images at different points. Video signals are output from each camera at a predetermined frame rate (for example, 30 fps in the NTSC standard) and input to the multi-input selection / sequentialization means 102. The multi-input selection / sequencing unit 102 selects a signal of any channel at a frame period as shown in FIG. 2 according to an instruction from the control CPU 105, converts the signal into one system, and outputs the signal.

  FIG. 3 shows an input / output example of the multi-input selection / sequencing means 102 at this time. The four-channel frame images (a) to (d) input from the cameras 101a to 101d are sequentially switched and output as a sequentially-converted one-channel image (e). In the drawing, an input frame with a “−” sign is a frame that is thinned out by the multi-input selection / sequentialization means. In FIG. 3, an example is shown in which Ch-A to Ch-D are sequentially switched at equal intervals, but the switching interval may be appropriately changed for each channel.

  The video signal data of a plurality of channels serialized by the multi-input selection / sequencing unit 102 is input to the channel distribution / rearrangement unit 103. The channel distribution / rearrangement unit 103 includes a memory, a write control unit, and a read control unit.

  FIG. 4 shows the signal processing of the channel distribution / reordering means 103. The serialized video signal data of a plurality of channels is stored for each channel in each address area assigned to each channel (FIG. 4 (a) → (b)). Here, the control CPU 105 performs control such as recognition of a video channel to be written and selection of a storage destination area in synchronization with the channel selection control of the serialization means 102.

  Further, the control CPU 105 reads out the video signal (data) stored in the memory by the number of frames of one encoding unit for each channel and inputs it to the image compression / encoding means 104 (FIG. 4 (b) → (C)). In FIG. 4, for the sake of simplicity, one coding unit (GOP) is shown as a six-frame configuration, but the number of frames is appropriately set according to the application and system configuration. The image compression / encoding unit 104 compresses / encodes input data by a technique such as MPEG, and generates and outputs an encoded data string (bit stream). The bit stream is recorded on the recording medium by the recording means 106. Alternatively, the bit stream can be sent to decompression / decoding means or an external device.

  The compression / encoding means 104 performs so-called bit rate control for adjusting the generated code amount so as to satisfy the target code amount designated in advance by the control CPU 105 during compression / encoding. As for the following bit rate control, it is assumed that control conforming to the MPEG2 / TM5 standard and a configuration for realizing this are performed unless otherwise specified. Here, the compression / encoding means 104 is configured so that bit rate control is not inherited between adjacent GOPs of different channels in order to handle sequential images of a plurality of channels.

  FIG. 5 shows the correlation of coding parameters between GOPs of the same channel in compression / coding. The control CPU 105 obtains internal state parameters (for example, quantization scale Q, generated code amount S, image complexity X, virtual buffer occupation amount d, etc.) at that time every time 1 GOP encoding is completed, and holds it for each channel. Then, when the compression / encoding of the channel is started next, it is supplied to the compression / encoding means so that it becomes an encoding state equivalent to continuous video in a pseudo manner for each channel. Thereby, the correlation between non-adjacent identical channel GOPs becomes effective.

FIG. 6 schematically shows the effect of the correlation, where the horizontal axis represents time (number of frames) and the vertical axis represents the code amount. Further, for simplification of description, only one channel (A-ch in the figure) among a plurality of serialized channels is shown. The compression / encoding means controls the bit amount for each picture so that the average value (S _AV ) of the generated code amount (S) becomes equal to the designated target code amount (S _T ) indicated by the dotted line by the above bit rate control. Adjust.

FIG. 6B shows a conventional operation example in which simple GOP completion bit rate control is performed. As described above, when the image has many fine patterns or the intensity of motion is particularly strong, the average generated code amount (S _AV ) cannot be controlled to the target value (S _T ) within the GOP. Show. In addition, when performing bit rate control, a predetermined time constant is provided in order to prevent abrupt deterioration of image quality and overcorrection / oscillation, and the slope of the code amount (S) corresponds to this. In this way, regardless of the preceding GOP result, the generated code amount is determined only by the feature of the image, so that the result is the same as each GOP, and the average value (S _AV ) of the generated code amount is constantly high. Become.

On the other hand, in this embodiment, as shown in FIG. 6C, the average value of the generated code amount in the initial stage is high, but the state of the GOP or picture that has been encoded is the next encoding parameter. Taken over. Second and subsequent GOP is the average value of the generated code amount _{(S AV)} decreases compared to gradually before GOP, convergence average generated code quantity _{(S AV)} is the target code amount _{(S T)} To do.

  As described above, according to the present embodiment, in an apparatus for processing video inputs of a plurality of channels with a single compression / encoding means, the influence of other adjacent channels is maintained while maintaining the correlation within the same channel. Therefore, efficient coding can be achieved without affecting the image quality of other channels. This stabilizes the amount of generated code for each channel, thereby improving the recording time prediction accuracy.

  In the above-described embodiment, the channel distribution / rearrangement means is configured to allocate storage areas for each channel. However, the storage areas are continuously stored in the order of input, and the storage destination address value is stored at this time. When reading, it is possible to realize sorting / reordering at the stage of reading by reading with reference to the stored storage address value.

Next, a second embodiment will be described with reference to FIG. Although the overall configuration is the same as that of the first embodiment shown in FIG. 1, the contents of data transmitted and received between the image compression / encoding unit 104 and the system control CPU 105 and the processing by the control CPU The contents are different. In this embodiment, every time GOP encoding is completed, the GOP generated code amount is read from the image compression / encoding unit, and a predetermined period, for example, nGOP is integrated to obtain an average code amount (S _AV ). The code amount corresponding to the target bit rate (S _Tar : overall target value) is compared. In accordance with the comparison result, if target code amount> average code amount, a larger GOP or a locally generated code amount target value (T _ref ) for each picture is set in the image compression / encoding unit, and target code amount <average If the code amount is small, the smaller GOP or picture unit local generated code amount target value (T _ref ) is set in the image compression / encoding unit. In other words, the overall target value becomes a desired value by respecifying the correction target code amount to the image compression / encoding unit. At this time, a difference obtained by multiplying the difference between the two code amounts by the sensitivity coefficient (K) is set as a correction amount.

_{T ref = S Tar -K × (} S AV -S Tar) ····· ( Equation 1)
The timing for specifying the correction target code amount T _ref is performed in synchronization with the start of the encoding process for the channel.

  Also in the second embodiment, in the same way as in the first embodiment, a long-cycle bit rate control can be realized for each channel in an apparatus that processes a plurality of channels of video input by a single compression / encoding unit. Therefore, efficient encoding and improvement in recording time prediction accuracy can be achieved without affecting the image quality of another channel. Further, according to the present embodiment, management of characteristics such as time constant and sensitivity can be unified in the control CPU, so that the various characteristics can be easily changed.

  Although the case where inter pictures (inter-frame prediction coding, P or B picture) are used has been described above, the present invention can also be applied to a case where only an intra picture (intra-frame prediction coding, I picture) is used. In this case, a configuration in which 103 channel sorting / reordering means is not used is possible.

  The present invention can be applied to a monitoring apparatus that is not easily affected by the monitoring location, the angle of view, and the state of the monitoring target (the fineness of the pattern and the magnitude of the movement) and that has little fluctuation in the total recording time. Further, the present invention can be applied to an image distribution apparatus with little band fluctuation. At that time, there is little error even when applied to the display of the expected recording time.

1 is an overall block diagram showing a first embodiment of a video signal encoding apparatus and recording apparatus according to the present invention. The block diagram which shows the detailed structure of the principal part of the Example shown in FIG. The schematic diagram which shows the signal processing of the Example shown in FIG. The schematic diagram which shows the signal processing of the Example shown in FIG. The schematic diagram which shows the signal processing of the Example shown in FIG. The schematic diagram which shows the signal processing of the Example shown in FIG. The schematic diagram which shows the signal processing in 2nd Example of the video signal encoding apparatus by this invention.

Explanation of symbols

101a, 101b, 101c, 101d ... camera 102 ... multi-input selection / sequential means (frame switch / multiplexer)
103 ... Channel distribution / reordering means (frame memory)
104 ... Image compression / encoding means 105 ... System control CPU
106: Recording means

Claims (9)

A selection means for selecting a channel in synchronization with a frame of input video signals of a plurality of channels, and converting the video signal into a single video signal;
Rearrangement means for rearranging the one-line video signal in units of a predetermined number of frames for each channel;
Encoding means for compressing and encoding the rearranged video signals in units of a predetermined number of frames;
Control means for controlling the encoding means,
The video signal encoding apparatus characterized in that the control means controls the code amount of the video signal compressed and encoded by the encoding means to be a predetermined value.   The control unit obtains the encoding parameter used in the encoding unit in units of 1 GOP (group of pictures), and uses the encoding parameter for compression encoding of video in the next 1 GOP unit of the channel. The video signal encoding apparatus according to claim 1, wherein:   The encoding parameter is information regarding at least one of image complexity, generated code amount, quantization scale, quantization scale ratio for each picture type, and virtual buffer occupancy. Video signal encoding device.   The control means acquires the generated code amount of the video signal compressed and encoded by the encoding means in units of 1 GOP (group of pictures), and from the past average value of the generated code amounts in GOP or picture units. 2. The video signal encoding apparatus according to claim 1, wherein the generated code amount target value is set and controlled.   The video signal encoding apparatus according to claim 1 or 2, wherein the selection means can appropriately change the frame interval at which each channel is selected.   The number of frames as a unit for rearrangement by the rearrangement unit and the number of frames as a unit for compression encoding by the encoding unit are the same and can be appropriately changed. The video signal encoding device according to 1 or 2. A selection means for selecting a channel in synchronization with a frame of input video signals of a plurality of channels, and converting the video signal into a single video signal;
Rearrangement means for rearranging the one-line video signal in units of a predetermined number of frames for each channel;
Encoding means for compressing and encoding the rearranged video signals in units of a predetermined number of frames;
Control means for controlling the encoding means;
Comprising recording means for recording the compression-encoded video signal on a medium;
The video signal recording apparatus characterized in that the control means controls the code amount of the video signal compressed and encoded by the encoding means to be a predetermined value.   The control unit obtains the encoding parameter used in the encoding unit in units of 1 GOP (group of pictures), and uses the encoding parameter for compression encoding of video in the next 1 GOP unit of the channel. The video signal recording apparatus according to claim 7, wherein:   9. The coding parameter according to claim 8, wherein the coding parameter is information relating to at least one of image complexity, generated code amount, quantization scale, quantization scale ratio for each picture type, and virtual buffer occupation amount. Video signal recording device.

Images