📄 rfc809.txt
字号:
facsimile machine on the end of a telephone line. Thus the local facsimile machine could transmit data to the computer quite happily, believing that it was actually talking to a remote facsimile machine on the other end of a telephone wire. Because of the property of the DACOM 6450 used in the experiment [16], the interface could be identical to one developed for connecting to an X25 network. The binary synchronous mode of the chip used (SMC COM5025) was appropriate to drive the DACOM machine. At the other side of the computer network there was a similar computer with an identical facsimile machine. The problem of transmitting a facsimile picture now appeared simple: data was taken from the facsimile machine into the computer, transmitted over the network as if it was normal computer data, and then sent from the computer to the facsimile machine at the remote end. The data being sent over the network appears - 4 -UCL FACSIMILE SYSTEM INDRA Note 1185 exactly as any other computer data; there is nothing special about it to signify that it came from a facsimile machine. The schematic of such facsimile transfer system is shown in Fig. 2. facsimile machine +---+ interface ! ! +--+ +-----+ ! ! == ! ! == ! ! computer +---+ +--+ +-----+ | - - - - - - computer / \ network \ / facsimile - - - - - - machine | interface +---+ +-----+ +--+ ! ! computer ! ! == ! ! == ! ! +-----+ +--+ +---+ Fig. 2 Facsimile transfer system The experimental system was used to perform a joint experiment between UCL and two groups in the United States. Pictures were exchanged via the ARPANET/SATNET [21], [22] between UCL in London, ISI in Los Angeles, and COMSAT in Washington D.C. (Fig. 3). This environment was chosen because no equivalent group was available in the UK. One problem concerned with such image data transmission is the quantity of data. Even with data compression, a single page of facsimile data can produce as much computer data as would normally be sufficient for sending over 20,000 alphabetic characters - or over a dozen typed pages. Thus for a given number of pages put into the system, an immense amount of computer data is produced. This means that the transmission will be slower than for sending text, and that far more storage will be required to hold the data. Another problem was encountered which became only too apparent when we implemented this system. The network we were using was often unable to keep up with the speed of the facsimile machine. When this happened the - 5 -UCL FACSIMILE SYSTEM INDRA Note 1185 US UK satellite COMSAT __ +---+ +--+ / \ ! ! -- ! ! / \ +---+ +--+ / \ | \ / \ +---+ \ / \ UCL !fax! \+--+/ \+--+ +---+ +---+ ARPANET ! ! SATNET ! ! -- ! ! /+--+ +--+ +---+ / | ISI / +---+ +---+ +--+ !fax! ! ! -- ! ! +---+ +---+ +--+ | +---+ !fax! +---+ Fig. 3. The three participants of the facsimile experiments computer tried to slow down the facsimile machine. The facsimile machine would detect this 'slowness' as a communication problem (as a telephone line would never act in this manner), and would abandon the transfer mid-way through the page. This is because the the facsimile machine we were using was never intended for use on a computer; it was designed and built for use on telephone lines. Indeed, being unaware that it was connected to a computer, the facsimile machine transmitted data at a constant rate, which exceeded the limit that the network could accept. In other words, the computer network we were using was not designed for the transfer rate that we were trying to use over it. Both these problems are surmountable. Facsimile machines are coming on the market that are designed for direct communication with a computer. These machines do not mind the delays on the computer interface and are tolerant of the stops and re-starts. On the other hand, if there were a serious use of facsimile machines on a computer network, the network could be designed for the high data rate required. Our problem was aggravated by - 6 -UCL FACSIMILE SYSTEM INDRA Note 1185 using a network that was never designed for the data rates required in our mode of usage. Despite the problems we encountered being a result of the experimental equipment we were working with, we still had to improve the situation to permit more extensive communications to take place. The easiest way to do this was to introduce a local storage area in our computer where the data could be held prior to transmission. The transfer of a page is now done in three stages. First, the facsimile data is read from the facsimile machine and stored on a local disk. This takes place at high speed as this is just a local operation. When this is complete, the data is sent over the network to a disk on the remote computer. Finally, the data from that disk is output to the remote facsimile machine. This improved system is shown in Fig. 4. computer network fax computer - - - - computer fax +---+ +-----+ / \ +-----+ +---+ ! ! = ! ! = ==> = ! ! = ! ! +---+ +-----+ \ / +-----+ +---+ - - - + | - - - - | + - - > | | + - - - - - - - - - + | | | | | | | | V | | V | | +---+ +---+ ! ! ! ! ! ! ! ! +---+ +---+ disk disk Fig. 4. The improved facsimile transfer system The idea behind this method is to decouple the facsimile machine from the network communications. The data is read from the facsimile machine at full speed, without the delays caused by the computer network. This also has the effect of being more acceptable to the human operators: each page is now read in less than a minute. The transmission over the network then takes place at whatever speed the network can sustain. This does not affect the facsimile machines at all; they are not involved in the sending or receiving. Only when all the data has been received at the remote disk is the remote facsimile machine told that the data is ready. - 7 -UCL FACSIMILE SYSTEM INDRA Note 1185 The facsimile machine is then given the data as fast as it will accept it. The disadvantage of such a system is that the person sending the pages does not know how long it will be before they are actually printed at the other side. If several pages are input in quick succession by the operator, they will be stored on disk; it may then be some time before the last page is actually delivered to the destination. This is not always a disadvantage; where many operators are sending data to the same destination, it is a definite advantage to be able to input the pages and have the system deliver them when the destination becomes free. Such a system is preferable to use of the current telephone system where the operator has to keep re-dialing the remote facsimile machine until the call is answered. 2.2 Interworking with Other Equipment 2.2.1 Facsimile machines As was mentioned earlier, facsimile machines produce a large amount of data per page due to the way in which the pages are encoded. To reduce the data that has to be transmitted, various compression techniques are employed. The manufacturers of facsimile machines have developed proprietary ways in which the data is compressed and encoded. Unfortunately this has meant that interworking of different facsimile machines has been impossible. In the system described in the last section, exchange of pictures was only possible between sites that had identical facsimile machines. The new set of CCITT recommendations will reduce the extent to which differences in equipment persist. Having the data on a computer gives us the opportunity to manipulate data in any way we wish. In particular we could convert the data from the form used in one facsimile machine to that required by another. This means that interworking between different types of facsimile machines can be achieved. The development of this system took place in two stages: the decompression of the facsimile data from the coded form used in our machine into an internal data form and the recompression of the data in the internal form into the encoded form required for the destination machine. Two programs were developed to perform these two operations. - 8 -UCL FACSIMILE SYSTEM INDRA Note 1185 At the same time we were developing compression and decompression programs for machines that use other techniques. In particular, we developed programs to handle the recently approved CCITT recommendation for facsimile compression [15]. The CCITT came up with two varieties of compression, depending upon the resolution being used. Unfortunately there were no facsimile machines on the network that use the CCITT compression technique. However, the programming of the new methods achieved two goals: it proved that the data could be converted inside a small computer, so that machines of different types could be supported on the network, and it enabled us to compare the compression results. These are described in more detail in [13]. Essentially, these show that the DACOM technique used by our facsimile machine is comparatively poor, and that considerably less data need be transmitted if some other method is used. This brings up another possibility: we could change the compression of the data to reduce the volume for transmission and then change the data back again at the destination. This may save considerable transmission time, especially if fast computers or special hardware was easily available. This has not been tried yet in our system, as none of the other users on the network have the capability of changing the data format back into that required by their machines. There are many other more efficient compression schemes, e.g. block compression [7] and predictive compression [8], but we have not yet incorporated them into our system. 2.2.2 Output Devices One area that we have explored is the use of devices other than facsimile machines for outputting the data. Facsimile machines are both expensive to buy and relatively slow to operate. We have investigated the use of a TV-like screen to display the data, just as character VDUs are commonly used to display text. This activity requires bit-map displays, with an address in memory for each postion on the screen. Full colour and multiple shades can be used with appropriately large bit-map storage. Although simple in principle, the implementation of the relevant techniques took considerable effort. - 9 -UCL FACSIMILE SYSTEM INDRA Note 1185 The problems arise in the way that the facsimile image is encoded. Raw facsimile images consist of rows of small dots, each dot recorded as a black or white space. When these dots are arranged together they build up a picture in a similar manner to the way in which a newspaper picture is made up. Unfortunately the number of dots used in a facsimile page is not the same as the number used on most screens. For instance, the DACOM facsimile machine uses 1726 dots across each page, but across a screen there are usually just 512 dots. Thus to show the picture on the screen the 1726 dots must be 'squeezed' into just 512 dots; stated another way, 1214 dots must be thrown away without losing the picture! It is in reducing the number of picture elements that the problem arises. We could just every third dot or so from the facsimile page and just display those. Alternatively, we could take three or more at a time and try to convert the group of them into a single black or white dot. Unfortunately, in both these cases, data can get lost that is necessary to the picture. For instance, a facsimile encoding of an architect drawing could easily end up with a complete line removed, radically changing the presentation of the image. After much experimentation, we developed a method of reducing the number of dots without destroying the picture. This is a thinning technique, whereby key elements of the picture are thinned, but not removed. Occasionally, when the detail gets too fine, some elements are merged, but under these circumstances the eye would not have been able to see the detail anyway.
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -