JP4541758B2 - Image transmission device - Google Patents

Description

  The present invention relates to an image transmission method and a transmission apparatus, and more particularly to a moving image transmission method and a transmission apparatus for compressing a moving image and transmitting the moving image to a network.

  In remote image monitoring systems or image distribution systems, as represented by public lines and the Internet, needs for moving image transmission apparatuses using an IP (Internet Protocol) network as a conduction path are rapidly expanding. For example, conventionally, in the delivery of image stream data (composed of compressed data) in MPEG-4 (Moving Picture Experts Group Phase 4), the image transmission unit encodes the image data to be sent by MPEG-4. The converted image data is temporarily stored in the storage unit of the image transmission unit as stream data. The image data includes images such as still images, moving images, CG (Computer Graphics), and animation, and also includes voice, audio, synthetic music, and the like. These image data are distributed from the storage unit in response to a request from the network.

  In order to distribute such image data, in particular, moving images, it is necessary to digitize and transmit, but when digitized, the amount of information becomes enormous, so that the transmission capacity of the information is reduced. In addition, video compression technology is required. Here, as a well-known moving image compression method, a compression world standard method such as MPEG-2 or MPEG-4 is used.

Here, an MPEG image compression technique will be described. MPEG-2 and MPEG-4 compressed image data, that is, stream data, includes Intra Picture (hereinafter referred to as I picture), Predictive Picture (hereinafter referred to as P picture), and Biderectionally Predictive Picture (hereinafter referred to as B picture). ), And each picture is compressed in three different encoding modes. An I picture is data obtained by coding and converting all image data of one frame of analog video within the frame. Therefore, when receiving this picture, the image receiving unit can reproduce the image with only one I picture. A P picture is obtained by performing inter-frame prediction in one direction from previous image data (I picture or P picture) and encoding only difference data. Therefore, the image receiving unit cannot reproduce the image only with the received P picture, and cannot reproduce the image without the original I picture. Further, if there is no P picture in the middle, the image becomes an erroneous image, for example, an image in which block distortion or the like occurs. A B picture is obtained by performing inter-frame prediction in two directions from two image data of the previous image data and the next image data, and encoding only difference data. As with the P picture, the original picture cannot be reproduced with only the B picture. Since the P picture and B picture have reduced redundancy in the time axis direction with respect to the preceding and succeeding pictures, the amount of compressed data can be reduced, but the original picture cannot be reproduced by itself. An example of a general MPEG-2 picture combination is shown below.
(I) (B) (B) (P) (B) (B) (P) (B) (B) (P) (B) (B) (P) (B) (B) (I) (B ) (B) (P) ...
In this way, an I picture is present once in 15 pictures, and this is generally repeated.

  Next, a system for distributing the compressed moving image as described above to the network will be described. FIG. 8 shows a network type moving image distribution system described in the image transmission method and the image transmission apparatus (JP 2003-309847, filing date: November 27, 2002) previously proposed by the present inventors. . In FIG. 8, a monitoring image captured by the camera 120 is encoded by an image transmission unit 111 such as an encoder, and image reception units 112-1, 112-2, and 112- such as decoders are respectively connected via a network 122. 3 is decoded and is displayed as a monitoring image on the image monitors 124-1, 124-2, and 124-3.

  The image transmission unit 111 that compresses a moving image is compressed at a predetermined bit rate (compression rate) by a compression processing unit inside the image transmission unit, and image compression data (stream) generated therefrom is converted into an image reception unit 112-1. , 112-2 and 112-3, each image receiving unit expands the stream into the original image data and outputs it to the monitor. In FIG. 8, the output stream from the image transmission unit 111 is directly transmitted to the networks 122-1, 122-2, and 122-3. Such a transmission method is called a unicast configuration.

  For example, the operation of this system requests stream data from the image transmission unit 111 from the image reception unit 112-1 via the network 122-1. The image transmission unit 111 distributes the stream data to the image reception unit 112-1 that has requested stream data.

  The image receiving unit 112-1 receives the stream data, expands the compressed stream data, displays the stream data on the monitor 124-1, and records it in a recording unit (not shown) as necessary. Next, the image reception unit 112-1 requests the next stream data from the image transmission unit 111 via the network 122-1.

  The image transmission unit 111 transmits the next stream data to the image reception unit 112-1 that has requested the stream data. The image receiving unit 112-1 receives the next stream data, expands the compressed stream data, displays it on the monitor 124-1, and records it in the recording unit as necessary. The same applies to the subsequent processes, and the other image receiving units 112-2 and 112-3 continuously perform stream data transmission requests, reception, and decompression.

  Next, the image transmission unit 111 will be described in detail with reference to FIG. In FIG. 9, the video signal from the camera 120 is input to the image transmission unit 111 via the input terminal 130. The image transmission unit 111 includes an encoding processing unit 131 and a protocol control unit 132. In this example, the encoding processing unit 131 is described as an MPEG-4 encoding processing unit. However, the encoding processing unit 131 is not limited to this, and is configured by an encoding processing unit of another system such as MPEG-2. It is also possible. The protocol control unit 132 includes an I-VOP (Video Object Plane) periodic buffer 133, an RTP (real time transport protocol) packet processing unit 134-1, 134-2, 134-3, and a TCP (transmission control protocol) _UDP (user datagram protocol) processing unit 135. Note that the output of the TCP_UDP processing unit 135 is sent from the output terminal 136 to each network 122. Here, three RTP packet processing units are shown, but in the case of this example, there are three types of transmission paths 122 having different transmission rates, and the number is not limited to three.

  Thus, the transmission rate adaptive packet transmission is realized by configuring the protocol control unit 132 as described above. This configuration will be described in detail below. The I-VOP cycle buffer 133 is a buffer having a capacity capable of storing at least encoded data from I-VOP (corresponding to the I picture described above) to immediately before the next I-VOP. is there.

  The RTP packet processing unit 134 generates a packet suitable for transmitting MPEG-4 encoded data or the like over the network. That is, according to the basic specification of RTP, the encoded data is divided into one to several packets for each VOP and is output to the next TCP_UDP processing unit 135.

  The TCP_UDP processing unit 135 transmits the RTP packet to the network 122 using either a connection type TCP protocol or a connectionless type UDP protocol. This selection can also be configured so that the user can remotely set it with a personal computer or the like.

  The protocol control unit 132 is mainly software processing by a processor, and the RTP packet processing unit 134 performs processing of the three types of transmission paths 122 of the image reception unit 112 connected to the transmission path for simultaneous delivery by unicast. .

  The MPEG-4 specification encoding processing unit 131 inputs a video signal, outputs MPEG-4 encoded data, and writes the encoded data in the I-VOP cycle buffer 133. The RTP packet processing unit 134 reads the encoded data from the I-VOP cycle buffer 133 using a ready signal (a signal indicated by a dotted line in FIG. 9) corresponding to the transmission rate from the TCP_UDP processing unit 135. That is, each RTP packet processing unit 134-1, 134-2, 134-3 has a data amount according to the transmission path rate (transmission speed) to each image receiving unit 112-1, 112-2, 112-3. Are read from the I-VOP cycle buffer 133. That is, the image data is inevitably discarded in the I-VOP cycle buffer 133 from the high bit rate transmission line to the low rate transmission line. In this way, the image data is automatically transmitted according to the transmission speed of the transmission path.

  Note that the method of generating a ready signal according to the transmission rate in the TCP_UDP processing unit 135 differs depending on the selected protocol. In the case of the TCP protocol, since it is a connection type, a ready signal corresponding to the transmission rate can be automatically generated by a response to the transmission packet from the encoding processing unit 131.

  On the other hand, since the UDP protocol is a connectionless type, a ready signal cannot be automatically generated. Therefore, the TCP_UDP processing unit 135 collects packet discard rate information that is periodically transmitted from the image receiving unit 112. From this periodic information, the TCP_UDP processing unit 135 controls the packet transmission rate so that the packet discard rate becomes zero, and generates a ready signal corresponding thereto. As a result, it is possible to generate a ready signal according to the transmission rate.

Here, the packet discard rate can be obtained from the expected number of received RTP packets and the actual number of received packets. The expected number of received packets is the number of packets received from the transmission side, and includes delayed packets and duplicate packets. The period is from the previous RTCP packet reception to the current RTCP packet reception. The calculation is obtained from the maximum sequence number and the minimum sequence number of the received packet. The sequence number represents the order of packets included in the RTP header. Note that details are described in RFC (Request For Comment) 1889, and thus description thereof is omitted here.
(Expected number of received packets) = (maximum sequence number)-(minimum sequence number) + 1
The packet discard rate is obtained as follows.
(Number of discarded packets) = (expected number of received packets)-(actual number of received packets)
(Packet discard rate) = ((Packet discard count) / (Expected received packet count)) × 255
Regarding the discard rate transmission interval, RFC 1889, A.I. As described in detail in the calculation of the 7RRCP transmission interval, the discard rate is information included in the header of the RTCP RR packet, and is transmitted at intervals of about 5 seconds.

  As described above, discarding of image data when image data is transmitted from a high bit rate transmission line to a low bit rate transmission line will be described in more detail.

  First, an example of a configuration in which image data is discarded will be described with reference to FIG. In FIG. 3, a case where image data is transmitted from the image transmission unit 111 to the image reception unit 112 will be described. The image data referred to here is encoded image data. When image data is transmitted from the image transmission unit 111 to the image reception unit 112, the image data is sent via a high bit rate transmission path (including a network, the same applies hereinafter) 125 and a low bit rate transmission path 126. At this time, since the bit rate is lowered from the high bit rate transmission path 125 to the low bit rate transmission path 126, the image data is discarded. Further, in detail, in the image data transmission method described above, since the image encoded data is managed in units of pictures as described above, a high bit rate transmission path 125 (for example, transmission) is used as shown in FIG. The rate is 1 Mbps), and is transmitted in bursty image encoded data indicated by P1 and P2. On the other hand, in the low bit rate transmission path 126 (for example, the transmission rate is 384 Kbps), since it is transmitted as the low bit rate image encoded data P3, the image encoded data is discarded here. Although not a high bit rate transmission path as described above, for example, it may occur when data is sent from a network having a transmission rate of 320 Kbps to a general public line having a transmission rate of 38.8 Kbps.

  Thus, since retransmission processing is performed when using the TCP protocol, there is no problem even if the image encoded data is discarded, but there is no guarantee that the image encoded data is sent to the other party when using the UDP protocol. For bursty data, image coded data is frequently discarded. Therefore, image encoded data may not be transmitted correctly. In order to cope with this, it is necessary to control generation of a ready signal for requesting picture data and to transmit average encoded image data.

  Further, as a method that does not cause the discard of the encoded image data, there is a system and method for image transmission (see, for example, Patent Document 1). This is because each packet data is opened for a predetermined time or more. A method for delaying transmission timing has been proposed.

Japanese Patent Laid-Open No. 2002-77260

  As described above, when the UDP protocol is used, the image encoded data is frequently discarded, and there is no guarantee that the image encoded data is reliably transmitted to the other side. The problem to say occurs. Further, the technique disclosed in Japanese Patent Application Laid-Open No. 2002-77260 is a method for calculating a delay time from the size of packet data, a preset network bandwidth, and a minimum delay time. There is a problem of getting bigger.

  An object of the present invention is to provide an image transmission method and an image transmission apparatus that enable image encoded data requested on the receiving side to be transmitted correctly.

  Another object of the present invention is to provide an image transmission method and an image transmission apparatus that can transmit average image encoded data in time units, not bursty data.

  Another object of the present invention is to provide an image transmission method and an image transmission apparatus that reduce packet discard in a low bit rate transmission path and enable transmission of image encoded data.

  Still another object of the present invention is to provide an image transmission method and an image transmission apparatus that enable transmission of image encoded data even against a temporary transmission path failure such as concentration of transmission load.

An image transmission apparatus according to the present invention includes an image input unit that inputs an image signal, an image transmission unit that receives and encodes the image signal, and a plurality of units that transmit encoded data from the image transmission unit. Transmission channels having different transmission speeds and image receiving units coupled to the transmission channels, the image transmission unit converting the image signal into stream data in units of pictures, and the code A protocol control unit that packetizes and transmits the stream data encoded by the conversion processing unit, and the protocol control unit is coupled to the buffer for storing the stream data, and is coupled to the buffer. a plurality of packet processing section for reading the stored packet data to the buffer, a plurality of data requesting packet data according to the request interval set Mer processing unit and a memory for storing a delay time table showing the relationship between the delay time to be added to the discard ratio information and the request interval of the transmission path, each of said plurality of timer processing unit to each of the transmission path When receiving the information that the bit rate of the transmission path or the bit rate is unknown from the image receiving unit at the time of connection, the initial value of the request interval is set, and the packet data from the respective image receiving units When there is a transmission request, the delay time corresponding to the discard rate information sent from the image receiving unit defined in the delay time table is added to the request interval, and after the added request interval has passed, reading packet data from the buffer via the packet processing unit transmits the packet data to each of said transmission lines for which the above request, the pro Coll control unit further includes a plurality of trigger generators corresponding to a plurality of the transmission path, the each of the trigger generating unit, the request interval set by the respective timer processing unit at predetermined intervals The timer processing unit is configured to shorten the request interval set by the timer processing unit by a predetermined time when receiving the trigger pulse. The

  According to the present invention, it is possible to correctly transmit image encoded data requested on the receiving side, and it is possible to transmit average image encoded data in time units, not bursty data. . Furthermore, packet discard on the low bit rate transmission path is reduced, image encoded data can be transmitted, and image encoded data can be transmitted even for temporary transmission path failures such as concentration of transmission load. An image transmission method and an image transmission apparatus can be provided.

  The principle of the present invention will be described with reference to FIG. In FIG. 5, when image encoded data is sent from the high bit rate transmission path 125 to the low bit rate transmission path 126, transmission of the image encoded data from the high bit rate transmission path 125 is a low bit as image encoded data in which the transmission is distributed. It is configured to transmit to the rate transmission path 126. This will be described in detail below.

  FIG. 1 is a block diagram showing an embodiment of the present invention. Reference numerals 137-1, 137-2, and 137-3 denote RTP packet processing units. In particular, when there is no need to distinguish between them, the RTP packet processing unit 137 is collectively referred to. Reference numeral 138 denotes a UDP processing unit, 139-1, 139-2, and 139-3, a timer processing unit (when collectively referred to as a timer processing unit 139). In addition, the same code | symbol is attached | subjected to the same thing as FIG. The image transmission unit 111 is also used, for example, in the image distribution system shown in FIG. 8, and the details thereof have already been described.

  In FIG. 1, the video signal from the camera 120 is input to the image transmission unit 111 via the input terminal 130. The image transmission unit 111 includes an encoding processing unit 131 and a protocol control unit 132. The encoding processing unit 131 can be configured by an encoding processing unit of a scheme such as MPEG-4 or MPEG-2. The protocol control unit 132 includes an I-VOP cycle buffer 133, RTP packet processing units 137-1, 137-2, and 137-3, and a UDP processing unit 138. The UDP processing unit 138 includes timer processing units 139-1, 139-2, and 139-3. The output of the UDP processing unit 138 is sent from the output terminal 136 to each network 122. Here, three RTP packet processing units and three timer processing units are shown, but the number is not limited to three.

  Next, this operation will be described with reference to FIG. FIG. 6 is a diagram for explaining the operation of the timer processing unit 139-1 of the I-VOP cycle buffer 133, the RTP packet processing unit 137-1, and the UDP processing unit 138, for example, with the configuration in FIG. Note that encoded image data encoded from the encoding processing unit 131 is applied to the input terminal 145. Further, the operations of the other RTP packet processing units 137-2 and 3 and timer processing units 139-2 and 3 are also the same, and thus description thereof is omitted.

  First, when the bit rate of the network 122 is known in advance, the bit rate of the network 122 is notified from the image receiving unit 112 to the image transmission unit 111 via the terminal 136 when each network is connected. The bit rate setting of the image receiving unit 112 can be set by a PC or the like. For example, if the image reception unit 112-1 is connected to the image transmission unit 111 via a network, the bit rate information is sent to the UDP processing unit 138. The UDP processing unit 138 sends bit rate information to the timer processing unit 139-1, and the timer processing unit 139-1 sets a waiting time setting in accordance with the bit rate information, that is, a so-called bit rate waiting time. When the ready signal is sent from the UDP processing unit 138, the timer processing unit 139-1 pauses the processing for the bit rate waiting time and then sends the ready signal to the RTP packet processing unit 137-1. For example, when the bit rate waiting time is set to 10 msec, the timer processing unit 139-1 is configured to perform packet (packetization for transmitting image encoded data) at intervals of 10 msec during system operation (at the time of initial value setting). To send a request to the RTP packet processing unit 137-1. Note that a delay time is appropriately added to the request interval (the initial value is the bit rate waiting time) based on the discard rate information sent from the transmission path 122. For example, Table 1 shows the relationship between the discard rate and the delay time, and such a table is stored in a memory (not shown) of the UDP processing unit 138.

表１において、破棄率情報は、０（０％に相当）〜２５５（１００％に相当）で表され、それらに対応して遅延時間が設定されている。即ち、破棄率０で遅延時間０ｍsec（正常時、遅延時間なし。）、破棄率１〜１０では、１０ｍsec、破棄率１１〜２０では、２０ｍsec、・・・のように破棄率が大きくなるに従ってパケットの要求間隔に加算される遅延時間が増加するように設定されている。また、表１のように破棄率に応じて遅延時間を異ならせるのではなく、破棄率が０でない場合は、要求間隔に一定の遅延時間を加算する、例えば、１０ｍsecずつ加算するようにすることも可能である。このように破棄率に対する遅延時間の設定は、破棄率が伝送路の伝送帯域、使用頻度等の種々の要因で変化するため、予め実験等により定めておくことが望ましい。

In Table 1, the discard rate information is represented by 0 (corresponding to 0%) to 255 (corresponding to 100%), and a delay time is set corresponding to them. That is, when the discard rate increases, the discard rate becomes 0 msec (normal, no delay time), the discard rate 1 to 10 is 10 msec, the discard rate 11 to 20 is 20 msec, and so on. The delay time added to the request interval is set to increase. Also, instead of changing the delay time according to the discard rate as shown in Table 1, when the discard rate is not 0, a certain delay time is added to the request interval, for example, 10 msec is added. Is also possible. As described above, the setting of the delay time with respect to the discard rate is preferably determined in advance by experiments or the like because the discard rate changes due to various factors such as the transmission band of the transmission path and the use frequency.

  In FIG. 6, first, the UDP processing unit 138 issues a request to send the packet D with the ready signal R1 to the RTP packet processing unit 137-1. When the RTP packet processing unit 137-1 receives the ready signal R1, the RTP packet processing unit 137-1 starts the operation of sending the packet D. When there is no packet to be sent, the RTP packet processing unit 137-1 receives the I-VOP by the ready signal R2. The encoded image data S1 is requested to the cycle buffer 133. When the encoded image data S1 is sent, the RTP packet processing unit 137-1 packetizes and stores the encoded image data S1, and sends the packet D1 to the UDP processing unit 138. The UDP processing unit 138 transmits the transmitted packet D 1 to the transmission path 122 via the output terminal 136. In the RTP packet processing unit 137-1, the case where the encoded image data S1 is packetized into three packets D1, D2, and D3 is shown, but the present invention is not particularly limited thereto. On the other hand, the principle explanatory diagram of FIG. 5 illustrates the case where P4, P5, P6, and P7 are each packetized as four packets as image encoded data.

  Next, when the UDP processing unit 138 receives the packet D1, the UDP processing unit 138 requests transmission of the next packet D2 by the next ready signal R3. Here, when transmitting the ready signal R3 requesting the packet D2, the ready signal R3 is transmitted according to the request interval set in the timer processing unit 139-1. For example, if the request interval set in the timer processing unit 139-1 is 10 msec, the packet D1 is received, and after 10 msec, the ready signal R3 is sent to the RTP packet processing unit 137-1. The packet D2 is sent to the UDP processing unit 138. Similarly, the ready signal R4 is further sent to the RTP packet processing unit 137-1 after 10 msec, and the packet D3 is sent to the UDP processing unit 138.

  As described above, the UDP processing unit 138 is configured to control the request timing of the ready signal R for requesting the packet D based on the request interval set in the timer processing unit 139-1. Further, the request interval set in the timer processing unit 139-1 is appropriately increased in accordance with the relationship between the discard rate and the delay time shown in Table 1 described above. In FIG. 6, the case where the packet D does not exist when the ready signal R1 is requested has been described. Needless to say, when the packet D exists, the packet D1 is transmitted in response to the arrival of the ready signal R1. In the above description, the operations of the timer processing unit 139-1 and the RTP packet processing unit 137-1 of the UDP processing unit 138 have been described, but the same applies to the other timer processing units 139 and RTP packet processing unit 137. Since there is, explanation is omitted. In this way, as shown in FIG. 5, the encoded image data P4 and P5 distributed from the high bit rate transmission path 125 can be transmitted to the low bit rate transmission path 126. Thus, it is possible to obtain the image encoded data P6 and P7 with reduced packet discard in

  The above is a case where the bit rate of the network 122 is known in advance. However, when the bit rate of the network 122 is unknown, the image receiving unit 122 indicates that the bit rate is unknown at the time of connection. Specifically, it notifies the UDP processing unit 138. In this case, the timer processing unit 139 sets the request interval of the ready signal R to the minimum waiting time (initial value) of the system set in advance. For example, it is 10 msec here. Thus, the image distribution system transmits the encoded image data from the image transmission unit 111 to the network 122 at a minimum interval. For this reason, the encoded image data becomes bursty, and data is discarded when passing through the low bit rate transmission path, and the image receiving unit 112 cannot receive correct data. As a result, the image reception unit 112 transmits the packet discard rate information to the UDP processing unit 138 of the image transmission unit 111. In the UDP processing unit 138, an appropriate request interval for the ready signal R is set by the timer processing unit 139 from the table shown in Table 1 based on the discard rate information. That is, the request interval is increased by adding the delay time based on the discard rate information. With this configuration, it is possible to transmit encoded image data with reduced packet discard on the low bit rate transmission path, as in the case where the bit rate of the network 122 is known in advance.

  Thus, in the case of the embodiment described with reference to FIG. 1, temporary network restrictions occur due to an increase in the amount of calls on the network 122, so-called traffic increase or network failure. In this case, the discard rate increases, and the bit rate transmitted by the network decreases for a while. When the bit rate is lowered, the packet is discarded, and the operation of the embodiment as described above is performed, and the delivery of the encoded image data becomes slow. As a result, the UDP processing unit 138 recognizes the network 122 as a low bit rate transmission path even though the bit rate of the network is temporarily lowered, and the packet processing interval (waiting time) of the timer processing unit. ). As a result, the elimination of traffic and network failures is eliminated, and the extended request interval is used even though the network transmission capacity is sufficient (for example, even if the packet discard rate is 0). As a result, a small amount of encoded image data is transmitted, resulting in excess line capacity. Therefore, the techniques described in the above embodiments may not be sufficient for temporary failures such as an increase in traffic and network failures.

  Next, another embodiment of the present invention for improving this will be described with reference to FIG. In FIG. 2, reference numeral 140 denotes a UDP processing unit, which is newly provided with trigger generation units 141-1, 141-2, and 141-3. Note that the trigger generation unit is collectively referred to as the trigger generation unit 141. In addition, the same components as those in FIG. In FIG. 2, the video signal from the camera 120 is input to the image transmission unit 111 via the input terminal 130. The image transmission unit 111 includes an encoding processing unit 131 and a protocol control unit 132.

  Note that the encoding processing unit 131 can be configured by an encoding processing unit of a scheme such as MPEG-4 or MPEG-2. The protocol control unit 132 includes an I-VOP cycle buffer 133, RTP packet processing units 137-1, 137-2, and 137-3, and a UDP processing unit 140. In the UDP processing unit 140, timer processing units 139-1, 139-2, and 139-3 and trigger generation units 141-1, 141-2, and 141-3 are provided. The output of the UDP processing unit 140 is sent from the output terminal 136 to each network 122. Here, three RTP packet processing units, three timer processing units, and three trigger generation units are shown, but the number is not limited to three.

  Next, the operation of the embodiment shown in FIG. 2 will be described. Since the basic operation of the image transmission unit 111 is the same as that of the above embodiment, the control of the transmission waiting time by the trigger generation unit 141 will be described. As described in the previous embodiment, for example, it has been described that the time interval of packets sent from the RTP packet processing unit 137-1 to the UDP processing unit 140 is determined by the ready signal request interval. The request interval of the ready signal is determined by an interval set by the timer processing unit 139, for example, 10 msec, but the request interval of the ready signal packet increases due to the discard rate from the network. The trigger generation unit 141 functions to periodically reduce the request interval of the increased ready signal packet and return it to the initial value (the above-described bit rate waiting time or the minimum waiting time).

  For example, the trigger generation unit 141 works to return the request interval of the timer processing unit 139 by 10 msec, for example, at intervals of 1 minute, 10 minutes, 1 hour, or 24 hours, for example. For example, when the trigger generation unit 141-1 is set to reduce the request interval of the ready signal of the timer processing unit 139-1 by 10 msec every hour, the request interval of the ready signal of the timer processing unit 139-1 is initially set. Is set to 10 msec (initial value), and the ready signal request interval is increased to 60 msec due to the discard rate from the network one hour later. At this time, the trigger pulse from the trigger generation unit 141-1 is applied to the timer processing unit 139-1, whereby the request interval of the ready signal of the timer processing unit 139-1 is reduced by 10 msec and changed (reset) to 50 msec. Is done. Similarly, after another hour, a trigger pulse is applied from the trigger generation unit 141-1 to the timer processing unit 139-1, and the ready signal request interval of the timer processing unit 139-1 is reduced by 10 msec and changed to 40 msec. The When such an operation is repeated, after 3 hours, the ready signal request interval of the timer processing unit 139-1 returns to the initial value of 10 msec. It should be noted that the setting value is not necessarily returned to the initial value, and the setting value is changed due to various factors such as the transmission band of the transmission path and the usage frequency. In practice, the request interval for the ready signal gradually increases during this period due to the discard rate from the network.

  Thus, the trigger generation unit 141 serves to shorten the ready signal request interval of the timer processing unit 139. As described above, the trigger generation interval can be set to various values depending on the actual system status, such as 1 minute, 10 minutes, 1 hour, or 24 hours as described above. (Not shown). In the above description, the request interval of the timer processing unit 139 is reduced by 10 msec. However, various values can be used for the value of how many msec the interval is reduced depending on the actual system situation. And can be registered in advance in a memory (not shown).

  Details of the operation will be described with reference to FIG. In step 151, the trigger generation unit 141 measures an elapsed time from the previous trigger generation. In step 152, it is checked whether the elapsed time from the previous trigger generation has reached the trigger generation interval. If the trigger generation interval has been reached, the timer processing unit 139 generates a trigger (step 153). If the trigger generation interval has not been reached, the system waits for a time (step 154).

  In step 155, the timer processing unit 139 that has received the trigger detects whether the bit rate is notified from the image receiving unit 112. When the bit rate is notified, the current waiting time (request interval) is compared with the bit rate waiting time (step 156), and when the current waiting time is larger than the bit rate waiting time, the waiting time is reduced (step 157). ) Perform processing. In the opposite case, there is no need to reduce the waiting time. When the bit rate is not notified, the process proceeds to step 158, and when the current waiting time is longer than the minimum waiting time, for example, 10 msec, the waiting time is reduced (step 157). In the opposite case, there is no need to reduce the waiting time. By repeating the above operation, it is possible to restore the waiting time of the timer processing unit 139 (to return to the initial value) and cope with a temporary change in the bit rate of the transmission path.

  In the above-described embodiment, every time a trigger occurs, the timer processing unit 139 reduces the current request interval, for example, by 10 msec. However, as another example, every time a trigger occurs. The timer processing unit 139 may forcibly return the request interval to an initial value (bit rate waiting time or minimum waiting time). At that time, the request interval immediately before returning to the initial value is stored in a memory (not shown) of the UDP processing unit 140, and for example, the discard rate immediately after returning the request interval to the initial value is predetermined. If it is greater than or equal to the value (for example, greater than or equal to the discard rate immediately before returning to the initial value), the request interval may be returned to the immediately preceding request interval stored in the memory.

次に、表２に言及して本発明の更に他の実施例を説明する。なお、画像伝送部１１１の構成は図１に示したものと同様である。上述した実施例では、タイマー処理部１３９は、表１に示したテーブルを用いて破棄率に基づいて遅延時間を加算して要求間隔を増大させる処理を行う一方で、その処理とは別に、所定間隔で発生するトリガーに基づいて図７で示したように要求間隔を減らす処理を行っていた。本実施例では、表２で示したテーブルを用いて、伝送路から送られてくる破棄率情報に基づいて適宜要求間隔を再設定するようにしたため、上述した実施例とは異なり、破棄率の増減に応じて要求間隔は適宜長い間隔となったり、短い間隔となったりする。表２は、破棄率と要求間隔との関係を示すもので、このようなテーブルがＵＤＰ処理部１３８のメモリに記憶されている。表２において、破棄率情報は、０（０％に相当）〜２５５（１００％に相当）で表され、それらに対応して要求間隔が設定されている。即ち、一例として、破棄率０（初期値）で要求間隔１０ｍsec（正常時）、破棄率１〜１０では、２０ｍsec、破棄率１１〜２０では、６０ｍsec、・・・のように破棄率が大きくなるに従ってパケットの要求間隔が増加するように設定されている。

Next, still another embodiment of the present invention will be described with reference to Table 2. The configuration of the image transmission unit 111 is the same as that shown in FIG. In the above-described embodiment, the timer processing unit 139 performs a process of increasing the request interval by adding a delay time based on the discard rate using the table shown in Table 1, but separately from the process, Based on the trigger generated at the interval, the processing for reducing the request interval is performed as shown in FIG. In the present embodiment, since the request interval is reset as appropriate based on the discard rate information sent from the transmission line using the table shown in Table 2, unlike the above-described embodiment, the discard rate Depending on the increase / decrease, the request interval is appropriately long or short. Table 2 shows the relationship between the discard rate and the request interval, and such a table is stored in the memory of the UDP processing unit 138. In Table 2, the discard rate information is represented by 0 (corresponding to 0%) to 255 (corresponding to 100%), and the request interval is set corresponding to them. That is, as an example, when the discard rate is 0 (initial value), the request interval is 10 msec (normal), when the discard rate is 1 to 10, the discard rate is 20 msec, when the discard rate is 11 to 20, the discard rate is 60 msec. Is set so that the packet request interval increases.

  For example, when the bit rate of the network 122 is unknown, when the ready signal is sent from the UDP processing unit 138, the timer processing unit 139-1 regards the system as operating (initial value setting) as normal, For example, the RTP packet processing unit 137-1 is requested to send a packet at a request interval corresponding to the discard rate 0 shown in Table 2, that is, an interval of 10 msec. This request interval (waiting time) is appropriately changed based on the discard rate information sent from the transmission path.

  Therefore, for example, when the amount of calls on the network 122 increases and packet discard occurs, a request interval corresponding to the discard rate is set according to the table shown in Table 2. On the other hand, when the increase in the traffic volume of the network 122 is resolved, the packet discard does not occur. For example, when the packet discard rate becomes 0, the request interval is set to the discard rate 0 as shown in Table 2. The corresponding request interval, that is, the initial value is returned.

  When the bit rate of the network 122 is known, the waiting time setting according to the bit rate information, that is, the so-called bit rate waiting time may be set as the initial request interval when the discard rate is 0.

  According to the embodiment of the present invention, the line speed of each transmission path is individually controlled to enable the image encoded data requested on the receiving side to be transmitted correctly and from the high bit rate transmission path to the low bit rate. It is possible to provide an image transmission method and an image transmission apparatus capable of reducing the packet discard rate with respect to transmission of encoded data to a transmission path and transmitting image encoded data. In addition, it is possible to provide an image transmission method and an image transmission apparatus that can transmit encoded image data in an optimum state even for a temporary transmission path failure.

  Although the embodiments of the present invention have been described in detail above, the present invention is not limited to the embodiments of the image transmission method and the image transmission apparatus described herein, and other image transmission methods and image transmissions than those described above. Needless to say, it can be widely applied to the apparatus. In the embodiments described so far, the data to be processed has been described by taking MPEG-4, MPEG-2, etc. as examples. However, the data to be processed is merely an example and the data to be processed is not limited to MPEG.

It is a block diagram for demonstrating one Example of this invention. It is a block diagram for demonstrating other one Example of this invention. It is principle explanatory drawing for demonstrating discard of the conventional image data. It is principle explanatory drawing for demonstrating discard of the conventional image data. It is a figure for demonstrating the principle of this invention. It is a block diagram for demonstrating operation | movement of this invention. It is a flowchart for demonstrating operation | movement of one Example of this invention shown in FIG. It is a block diagram which shows an example of a network type moving image delivery system. It is a block diagram for demonstrating an example of the conventional image transmission part.

  111: Image transmission unit, 112: Image reception unit, 120: Monitoring camera, 122: Network, 124: Monitor, 125: High bit rate transmission path, 126: Low bit rate transmission path, 130, 145: Input terminal, 131: Encoding processing unit, 132: protocol control unit, 133: I-VOP period buffer, 136: input / output terminal, 137: RTP packet processing unit, 138, 140: UDP processing unit, 139: timer processing unit, 141: trigger generation R1,... R4: Ready signal, D1,... D3: Packet, S1: Image encoded data.

Claims (1)

Image input means for inputting an image signal, an image transmission unit for inputting, encoding, and transmitting the image signal, and a plurality of transmission paths having different transmission speeds for transmitting encoded data from the image transmission unit And an image receiving unit coupled to each of the transmission paths,
The image transmission unit includes an encoding processing unit that converts the image signal into stream data in units of pictures, and a protocol control unit that packetizes and transmits the stream data encoded by the encoding processing unit,
The protocol control unit includes a buffer that stores the stream data, a plurality of packet processing units that are coupled to the buffer and that read the packet data stored in the buffer in each of the transmission paths, according to a set request interval. A plurality of timer processing units for requesting packet data and a memory storing a delay time table indicating a relationship between the discard rate information of the transmission path and a delay time to be added to the request interval ;
Each of the plurality of timer processing units receives the information indicating that the bit rate of the transmission channel or the bit rate is unknown from the image receiving unit when connected to the transmission channel, and sets an initial value of the request interval. and sets, when there is a transmission request of the packet data from the respective image receiving unit, a delay time corresponding to the discard rate information sent from the image receiving section defined by the delay time table the Add to the request interval, read the packet data from the buffer via the packet processing unit after the added request interval, and transmit the packet data to each of the requested transmission path ,
The protocol control unit further includes a plurality of trigger generators corresponding to a plurality of the transmission paths, each of the trigger generating unit described above, the request at predetermined intervals are set by the respective timer processing unit A trigger pulse for shortening the interval by a predetermined time is generated , and the timer processing unit shortens the request interval set by the timer processing unit by a predetermined time when receiving the trigger pulse. An image transmission device.

Images