JP2007019817A - Image coding apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image coding apparatus which suppresses the required memory capacity and reduces the data transfer quantity to a picture recorder. <P>SOLUTION: The image coding apparatus comprises a temporary memory 105 and a controller 107. High-quality image data are recorded in a recorder 106, and specified high frequency coefficients are deleted from quantized DCT coefficients before storing them in the memory 105. When no T1 time event happens, the controller 107 controls so that VLC is applied to the coefficients stored in the memory 105, the result is recorded in the recorder, and corresponding high-quality image data are deleted from the recorder. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an image encoding device of a video monitoring system that records a video to be monitored, and more particularly to an image encoding device that records high-quality video when an event occurs.

  In recent years, various video surveillance systems for crime prevention have been proposed (see, for example, Patent Document 1).

  FIG. 10 is a block diagram showing a configuration of an image encoding device in a conventional video surveillance system. The image encoding device 401 includes an image sensor 402, an event detector 403 that detects an abnormal event that has occurred in the monitoring target, a clock unit 404 that outputs time information, and high-quality image data from the image sensor 402. A video recording unit 405 that records low-quality image data from a video conversion unit, which will be described later, and a video conversion unit 406 that converts high-quality image data read from the video recording unit 405 into low-quality image data. And a write control unit 407 that controls recording of video information to the video recording unit 405 based on information from the event detection unit 403 and the clock unit 404.

The operation of the conventional image coding apparatus having the above configuration will be described. The video recording unit 405 records high-quality image data from the image sensor 402 and also records low-quality image data from the video conversion unit 406. Then, when normal, that is, when an abnormal event does not occur, the writing control unit 407 performs control so that high-quality image data before the time obtained by subtracting a predetermined set time from the current time is sequentially deleted from the video recording unit 405. ing.
JP 2004-356955 A (page 4-6, FIG. 2)

  However, since the conventional image encoding apparatus performs three types of data transfer, that is, writing and reading high-quality image data and writing low-quality image data with respect to the video recording unit 405, high-quality image data. There has been a problem that the amount of data transfer is much larger than when only data is written.

  In a monitoring system for roads, stores, warehouses, etc., usually, images from a plurality of monitoring cameras (imaging devices) are often recorded in one image recording unit 405. Therefore, in order to process a plurality of images simultaneously. It is necessary to reduce the amount of data transferred to the video recording unit 405.

  The present invention has been made in order to solve the conventional problem, and by temporarily storing only predetermined conversion coefficients of image data encoded until halfway in a temporary storage unit, a required memory capacity can be reduced. An object of the present invention is to provide an image encoding device that can suppress the data transfer amount to the video recording unit.

An image encoding apparatus according to the present invention includes an image sensor that converts an optical signal into an electrical signal, an A / D converter that converts an analog signal into a digital signal, and a code that generates required encoded image data from the digital signal. An encoding unit, a temporary storage unit that retains image data encoded until halfway for a predetermined time, a recording unit that records completely encoded image data, and processing related to encoding and selection of image data A control unit for controlling, and an event detection unit for detecting an abnormal event near the imaging, the control unit is based on the picture type of encoding and the condition of the presence or absence of event detection by the event detection unit, It is characterized in that it is determined whether to store or delete the image data encoded up to a halfway stage stored in the temporary storage unit and the completely encoded image data recorded in the storage unit.
With this configuration, the required memory capacity and data transfer amount to the recording unit can be suppressed, and high-quality image data before and after the occurrence of the event can be recorded.

Further, in the image encoding device according to the present invention, the control unit performs the encoding with high image quality when a predetermined time elapses after the image data is recorded in the recording unit and no event is detected by the event detecting unit. The image data is thinned based on the picture type, processed into low-quality image data, and then intermittently recorded in the recording unit, and the encoded high-quality image data part is deleted, and the event by the event detection unit When there is detection, the encoded high-quality image data is held in the recording unit for a preset time before and after the occurrence of the event.
With this configuration, the required memory capacity and data transfer amount to the recording unit can be suppressed, and high-quality image data before and after the occurrence of the event can be recorded.
Further, in the normal state, that is, in a state where no abnormal event occurs, since low-quality intermittent image data is recorded in the recording unit, it is possible to save the storage capacity of the recording unit, and to record for a longer time or an image encoding device. The manufacturing cost can be reduced.

In the image encoding device of the present invention, the control unit stores only a predetermined transform coefficient of image data encoded up to an intermediate stage in the temporary storage unit.
With this configuration, it is possible to adjust the image quality of low-quality image data to be recorded when no abnormal event occurs.

In the image encoding device of the present invention, when the user determines that the video before and after the occurrence of the event is unnecessary after the occurrence of the event, the control unit reads the high-quality image data before and after the occurrence of the event from the recording unit. It is characterized in that low-quality image data that has been extracted and decoded halfway through and encoded again using only predetermined transform coefficients is intermittently recorded in the recording unit.
With this configuration, even with high-quality image data related to the event, images that the user determines to be unnecessary are sequentially converted into low-quality intermittent image data, so the storage capacity of the recording unit can be saved, and longer recording time In addition, the manufacturing cost of the image encoding device can be reduced.

The image encoding apparatus of the present invention includes an input unit for inputting, as parameters, a time when the user wants to record image data before and after the occurrence of an event, and an intermittent recording time for recording the image data when the event has not occurred. In addition, the control unit may determine which conversion coefficient is to be stored in the temporary storage unit from the parameter and a usable memory capacity.
With this configuration, it is possible to record low-quality intermittent image data that satisfies the recording time and intermittent recording time before and after the occurrence of an event input by the user, using a predetermined memory capacity.

The image coding apparatus according to the present invention further includes a display unit for confirming an image quality of image data stored according to the parameter set by a user.
With this configuration, the user can confirm the image quality of the low-quality intermittent image data recorded according to the parameters set by the user.

  The present invention can suppress the required memory capacity and the amount of data transferred to the video recording unit, and can record high-quality image data before and after the occurrence of the event. Further, in the normal state, that is, in a state where no abnormal event occurs, since low-quality intermittent image data is recorded in the recording unit, it is possible to save the storage capacity of the recording unit, and to record for a longer time or an image encoding device. The manufacturing cost can be reduced.

  Hereinafter, an image coding apparatus according to an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram showing an internal configuration of an image coding apparatus according to Embodiment 1 of the present invention. The image encoding device 101 includes a solid-state imaging device 102 that converts an optical signal into an electrical signal, an A / D converter 103 that converts an analog signal that is an electrical signal into a digital signal, and required encoded image data from the digital signal. A temporary storage unit 105 that retains image data encoded until halfway for a predetermined time, a recording unit 106 that records completely encoded image data, and encoding until halfway A control unit 107 that controls a process of encoding and selecting the encoded image data or the completely encoded image data, and an event detection unit 108 that detects an abnormal event near the imaging.

  When encoding the digital signal from the A / D converter 103, the encoding unit 104 generates an I picture by performing encoding using only one frame. As inter-frame coding, P picture generation by forward prediction using the relationship between the current and past frames and B picture generation by bidirectional prediction using past and future frames are performed. A set of one I picture followed by a plurality of P or B picture is called GOP (Group Of Piccture).

  In the encoding unit 104, the coefficient converted into the frequency domain by a DCT (Discrete Cosine Transform) 109 is quantized by a quantization 110. Further, the high-frequency deletion 117 deletes a predetermined high-frequency coefficient of the coefficient after quantization (ie, sets the level to 0) under the control of the control unit 107. As a result, high-quality image data is converted into low-quality image data with a small data amount.

  FIG. 2 is a flowchart showing an operation procedure of the image coding apparatus according to Embodiment 1 of the present invention. First, the control unit 107 sets variables based on the recording time [T1] before the event occurrence, the recording time [T2] after the event occurrence, and the intermittent time [td] (step S101), and initializes the temporary storage unit 105 as an initial value. (Step S102).

  The digital signal from the image sensor is encoded into high-quality image data for one frame by the encoding unit 104 (step S103). At this time, one I picture and a plurality of P pictures are generated every intermittent time [td]. As a result, 1 GOP data is generated every intermittent time [td]. FIG. 3 is a diagram showing the state of the image data in the storage unit and the temporary storage unit of the image coding apparatus according to Embodiment 1 of the present invention. As shown in the figure, in the present embodiment, the number of frames acquired in the intermittent time [td] is four, and B picture is not generated, and inter-picture encoding is performed only for P picture. In this embodiment, the recording time T1 before the occurrence of the event is recorded as a time corresponding to 3 GOP, and the recording time T2 after the occurrence of the event is recorded as a time corresponding to 2 GOP.

  Next, the encoded high-quality image data (I) is recorded in the recording unit 106 (step S104). At the same time, the control unit 107 determines whether the recorded image data is an I picture (step S105) and determines whether an event is detected (step S106). Since the image data recorded at time t0 is an I picture and no event is detected, the process proceeds to the determination in step S110. In this case, since X2> 0, the process further proceeds to step S112, and the DCT transform coefficient obtained by quantizing the I picture in the temporary storage unit 105 is deleted from the predetermined high-frequency coefficient by the high-frequency deletion 117, and then the low-quality image data (I ′) is stored (step S112), and the variable is updated (step S113).

  Next, the control unit 107 determines whether data before time T1 is stored in the temporary storage unit 105 (step S114). At this stage, since the previous data is not stored in the temporary storage unit 105, the process returns to step S103, and the high-quality image data of the next frame is encoded.

  Since the encoded image data is a P picture before time t1, the encoding of the image data and the recording of the high-quality image data in the recording unit 106 from step S103 to step S105 is the number of P pictures. Repeated. Since processing similar to that at time t0 is performed at subsequent times t1 and t2, description thereof will be omitted.

  In FIG. 3, high-quality GOPs (consisting of one I and three Ps, each of which is variable-length encoded by a variable length coding (VLC) 111) recorded in the recording unit 106 are denoted by H1, H2, The low-quality GOP recorded in the temporary recording unit 105 (consisting of one I ′) is represented as L1, L2, and L3. In addition, L1, L2, and L3, which are variable length encoded by the VLC 111, are represented as L1 ′, L2 ′, and L3 ′, respectively.

  At time t3, high-quality image data H1, H2, and H3 are recorded in the storage unit 106, and low-quality image data DCT conversion coefficients L1, L2, and L3 are stored in the temporary storage unit 105. Has been. At the time t3, the control unit 107 determines that the data before the time T1 is stored in the temporary storage unit 105 (step S114). The VLC 111 of the encoding unit 104 performs variable length encoding on the encoded image data L1 (step S115), and the completely encoded low-quality image data L1 ′ is recorded on the recording unit 106 (step S116).

  Further, the control unit 107 deletes the corresponding high-quality image data H1 from the recording unit 106 (step S117), and at the same time, temporarily stores L1 that is a quantized DCT transform coefficient for low-quality image data. It deletes from 105 (step S118).

  Next, between time t3 and time t4, the event detection unit 116 detects an event, and the control unit 107 sets an event detection flag.

  At time t4, L1 ′, which is low-quality image data, and H2, H3, H4, which are high-quality image data, are recorded in the recording unit 106, and a predetermined high-frequency coefficient is deleted for the low-quality image data. The quantized DCT transform coefficients L2, L3, and L4 are stored in the temporary storage unit 105. At time t4, the control unit 107 determines that an event has been detected in determining whether or not an event has been detected (step S106). Therefore, the control unit 107 has quantized the low-quality image data stored in the temporary storage unit 105. All the DCT conversion coefficients (L2, L3, L4) are deleted (step S107), and the recording time after the occurrence of the event is set to X2 (step S108). Thereafter, the event detection flag is canceled (step S109).

  Next, at time t5, the control unit 107 determines that image data is being recorded in the determination of whether or not image data is being recorded after the occurrence of the event (step S110), and therefore updates the variable (step S111). That is, at this time, L5 and L6 which are DCT transform coefficients after quantization in which a predetermined high-frequency coefficient is deleted for low-quality image data are not stored in the temporary storage unit 105.

  At the time t6, the control unit 107 determines that the data recording is finished in the determination of whether or not the image data is being recorded after the occurrence of the event (step S110). Therefore, the control unit 107 proceeds to step S112 and performs the processing described at the time t0. Move to the same processing procedure.

  The procedure for deleting a predetermined high-frequency coefficient by the high-frequency deletion 117 is as follows. First, the memory capacity [SIZE MB] assigned to one macroblock is changed from the capacity [SIZE] of the temporary storage unit 105 and the recording time before the occurrence of the event [ T1] and the intermittent time [td] are calculated by the following formula, and the DCT transform coefficients after quantization in the zigzag scan order are temporarily stored until the memory capacity allocated to one macroblock is exceeded. This is realized by storing in the unit 106.

  FIG. 4 is a diagram illustrating an example of deleting a high frequency coefficient. In the figure, an example of deletion is shown in which up to 15 DCT transform coefficients after quantization can be stored in the memory capacity allocated to one macroblock. In addition, by leaving the conversion coefficients after quantization in the zigzag scan order in this way, it is possible to preferentially retain low frequency coefficients that are important in visual characteristics, and image quality when converting to low-quality image data. Easy to maintain.

  As described above, according to the image coding apparatus in Embodiment 1 of the present invention, a temporary storage unit that deletes and stores a predetermined high-frequency coefficient with respect to a DCT transform coefficient after quantization of an I picture is provided, If the T1 time event does not occur, it is necessary to perform VLC on the quantized DCT transform coefficient in the temporary storage unit, record it in the recording unit, and delete the corresponding high-quality image data from the recording unit. The memory capacity and the amount of data transferred to the video recording unit can be suppressed, and high-quality image data before and after the occurrence of the event can be recorded. Further, in a normal state, that is, in a state where an abnormal event does not occur, low-quality intermittent image data is recorded in the recording unit, so that the storage capacity of the recording unit can be saved, longer recording time, and an image encoding device The manufacturing cost can be reduced.

  FIG. 5 is a block diagram showing an internal configuration of the image coding apparatus according to Embodiment 2 of the present invention. Compared to Embodiment 1, an input unit 209 that accepts user input, and a VLD (Variable Length Decoding) that performs variable-length decoding of high-quality image data read from the storage unit 206 in the encoding unit 204 The difference is that 219 is added. Since other configurations are the same as those of the first embodiment, description thereof is omitted.

  FIG. 6 is a flowchart showing an operation procedure of the image coding apparatus according to Embodiment 2 of the present invention. First, the same operation as in the first embodiment is performed, and the previous and subsequent images are recorded as high-quality image data in the storage unit 206 when the event occurs, and the low-quality image data is recorded in the recording unit 206 except before and after the event occurrence. To record. Next, when the user confirms high-quality image data after the occurrence of the event (step S201) and determines that the data is necessary / unnecessary (step S202), similarly to the image data other than before and after the occurrence of the event. The image data is converted into low-quality image data and recorded in the recording unit 206.

  Therefore, first, the input unit 209 determines that the user does not need it, the control unit 207 determines the designated high-quality image data, and reads out the corresponding high-quality image data from the recording unit 206 (step S203). Variable length decoding is performed by the VLD 219 (step S204). Next, a predetermined high frequency coefficient is deleted by the high frequency deletion 228 (step S205), and the decoded DCT transform coefficient is subjected to variable length encoding again by the VLC 212 to generate low-quality image data. (Step S206). The low-quality image data is recorded in the recording unit 206 (step S207), and the high-quality image data read out earlier is deleted from the recording unit 206 (step S208). Note that the conversion to low-quality image data is performed only for the I picture of the high-quality image data, and not for the P and B pictures. Further, high-quality image data is deleted in GOP units.

  As described above, according to the image coding apparatus in the second embodiment of the present invention, when the user confirms high-quality image data before and after the occurrence of the event and determines that it is unnecessary, unnecessary high-quality image data is displayed. By providing an input unit that can be specified and a VLD that performs variable-length decoding of high-quality image data in the encoding unit, high-quality image data related to an event that is determined to be unnecessary becomes a low-quality intermittent image. Since the data is sequentially converted, it is possible to save the storage capacity of the recording unit, and it is possible to record for a longer time and to reduce the manufacturing cost of the image encoding device.

  FIG. 7 is a block diagram showing an internal configuration of the image coding apparatus according to Embodiment 3 of the present invention. Compared to Embodiment 1, the input unit 309 for the user to input the recording time [T1] before the event occurrence, the recording time [T2] after the event occurrence, and the intermittent time [td], and the user A D / A converter 310 for converting a digital signal, which is low-quality decoded image data from which the high-frequency coefficient determined by the input parameter is deleted, into an analog signal, and a display unit 311 for displaying the analog signal as an image are added. The difference was made. Since other configurations are the same as those of the first embodiment, description thereof is omitted.

  FIG. 8 is a diagram schematically illustrating an example of a user interface of the display unit 311. The user interface 401 includes an input unit 402 for input by the user and a display unit 403 for displaying low-quality decoded image data from which the high frequency coefficient determined by the parameter input by the input unit 402 is deleted. Here, the input unit 402 includes an input form 404 for the recording time [T1] before the occurrence of the event, an input form 405 for the recording time [T2] after the occurrence of the event, and an input form 406 for the intermittent time [td]. An execution button 407 for executing display of low-quality decoded image data from which the high frequency coefficient determined by the values input in the three input forms is deleted, and a stop button 408 for shifting from the execution state to the stop state. Have.

  FIG. 9 is a flowchart showing an operation procedure of the image coding apparatus according to Embodiment 3 of the present invention. First, the user inputs values in the input form 404 for the recording time [T1] before the event occurrence, the input form 405 for the recording time [T2] after the event occurrence, and the input form 406 for the intermittent time [td]. When the execution button 407 is pressed (step S302), the control unit 307 writes the input parameters (T1, T2, td) (step S303).

  The encoding unit 304 first performs encoding for one frame (step S304). Then, it is determined whether or not the encoded frame is an I picture (step S305), whether or not the stop button is pressed (step S306), and if the encoded frame is a P picture or a B picture, or If the stop button is pressed, the process proceeds to encoding of the next frame without executing the subsequent processing.

  On the other hand, when the encoded frame is an I picture and the stop button is not pressed, the input parameters (T1, T2, td) are read (step S307), and the control unit 307 is based on the set parameters. Then, the high frequency coefficient to be deleted is determined by the above-described determination formula for the high frequency coefficient to be deleted (step S308). In the above formula, the capacity [SIZE] of the temporary storage unit 305 is determined at the time of hardware design, and is not usually an adjustment item for the user. On the other hand, since the number of macroblocks in one frame is determined by the resolution of the image to be encoded, in a system in which the resolution can be changed, the resolution setting may be included in the user interface 401 so that the user can set it.

  Thereafter, the encoding unit 304 deletes the determined high-frequency coefficient (step S309), the quantized DCT coefficient after dequantization is dequantized by the inverse quantization 315 (step S310), and the inverse DCT 316 performs inverse DCT. The low-quality image data is decoded (step S311). Next, the decoded low-quality image data is displayed on the display unit 403 (step S312). At this time, as shown in the first embodiment, low-quality image data is created for each I picture, so that the image on the display unit 403 is updated every time specified by the input form 406 for the intermittent time. .

  As described above, according to the image coding apparatus in Embodiment 3 of the present invention, the time for which the user wants to record image data before and after the occurrence of an event and the intermittent recording time for recording image data by thinning out when no event occurs are recorded. By providing an input unit for input as a parameter and a display unit for confirming the image quality of image data stored according to the parameters set by the user, the user can record intermittent image data with low image quality recorded according to the parameters set by the user. The parameters can be determined while checking the image quality.

  In the above description, the example in which the encoding unit 104 is configured by MPEG has been described. However, it goes without saying that other compression schemes, such as wavelet transform, can be similarly implemented.

  As described above, the image coding apparatus according to the present invention can suppress the required memory capacity and the data transfer amount to the video recording unit, and can record high-quality image data before and after the occurrence of the event. It has an effect and is useful in an image encoding device of a video surveillance system that records video to be monitored.

The block diagram which shows the internal structure of the image coding apparatus in Embodiment 1 of this invention. The flowchart which shows the operation | movement procedure of the image coding apparatus in Embodiment 1 of this invention. The figure which shows the state of the image data in a memory | storage part and a temporary memory part of the image coding apparatus in Embodiment 1 of this invention. Diagram showing examples of deleting high-frequency coefficients The block diagram which shows the internal structure of the image coding apparatus in Embodiment 2 of this invention. The flowchart which shows the operation | movement procedure of the image coding apparatus in Embodiment 2 of this invention. FIG. 9 is a block diagram showing an internal configuration of an image coding apparatus according to Embodiment 3 of the present invention. The figure which shows the user interface example of a display part typically The flowchart which shows the operation | movement procedure of the image coding apparatus in Embodiment 3 of this invention. The block diagram which shows the structure of the image coding apparatus in the conventional video surveillance system

Explanation of symbols

101, 201, 301 Image encoding device 102, 202, 302 Solid-state imaging device 103, 203, 303 A / D converter 104, 204, 304 Encoding unit 105, 205, 305 Temporary storage unit 106, 206, 306 Recording unit 107, 207, 307 Control unit 108, 208, 308 Event detection unit 109, 210, 312 DCT
110, 211, 313 Quantization 111, 212, 314 VLC
112, 213, 315 Inverse quantization 113, 214, 316 Inverse DCT
114, 215, 317 Frame memory 115, 216, 318 Motion compensation 116, 217, 319 Motion vector detection 117, 218, 320 High frequency deletion 209, 309 Input unit 219 VLD
310 D / A Converter 311 Display Unit

Claims (6)

An image sensor that converts an optical signal into an electrical signal, an A / D converter that converts an analog signal into a digital signal, an encoding unit that generates required encoded image data from the digital signal, and encoding up to an intermediate stage A temporary storage unit that retains the image data for a predetermined time, a recording unit that records the completely encoded image data, a control unit that controls processing related to encoding and selection of the image data, An event detection unit for detecting abnormal events;
The control unit, based on the picture type of encoding and the condition of presence / absence of event detection by the event detection unit, encoded image data stored in the temporary storage unit up to a halfway stage and the complete recorded in the storage unit An image encoding apparatus characterized by determining whether to store or delete image data encoded in (1). The control unit thins out the encoded high-quality image data based on the picture type when a predetermined time elapses after the image data is recorded in the recording unit and there is no event detection by the event detection unit, After being processed into low-quality image data and intermittently recorded in the recording unit, the encoded high-quality image data portion is deleted,
2. The image encoding according to claim 1, wherein when there is event detection by the event detection unit, the encoded high-quality image data is held in the recording unit for a preset time before and after the occurrence of the event. apparatus.   3. The image encoding device according to claim 1, wherein the control unit stores in the temporary storage unit only predetermined transform coefficients of image data encoded up to an intermediate stage. 4.   When the user determines that the video before and after the event occurrence is unnecessary after the event occurs, the control unit takes out the high-quality image data before and after the event occurrence from the recording unit, decodes it until an intermediate stage, 4. The image encoding apparatus according to claim 1, wherein low-quality image data encoded again using only the transform coefficient is intermittently recorded in the recording unit. 5.   An input unit for inputting, as a parameter, a time for which the user wants to record image data before and after the occurrence of an event, and an intermittent recording time for recording image data by thinning out when no event occurs, and the control unit includes the parameter and 5. The image encoding apparatus according to claim 1, wherein which transform coefficient is stored in the temporary storage unit is determined based on an available memory capacity.   6. The image coding apparatus according to claim 5, further comprising a display unit for confirming an image quality of image data stored according to the parameter set by a user.

Images