JP2000217113A - Image communication unit and compression coding method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain an image communication unit, where a data length of coded data can be approached to a desired data length with less number of times of quantization and coding processing. SOLUTION: This image communication unit, that converts image data into spatial frequency data, quantizes the converted data, encodes the quantized data and transmits the coded data, is provided with a scaling factor generating means 15 that generates a scaling factor, a quantization table generating means 16 that multiplies the scaling factor with each table element for generating a quantization table, and a data storage means 17. Every time coded data are transmitted, the scaling factor generating means 15 calculates ratio data which denotes lower compression rate resulting from many high-frequency components of the image data and writes the ratio data to the storage means. When new image data are compressed, a prescribed number of the ratio data written in the past is read from the storage means, and the generating means 15 generates a scaling factor by multiplying a prescribed value with the means value of the ratio data. A data length of the coded data is close to a prescribed data length as time elapses by continuing a series of the processing above.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image communication apparatus for compressing arbitrary image data and transmitting the compressed image data to a communication partner, and a compression encoding method therefor. In particular, the data length after compression approaches a desired data length. This is what made it possible.

[0002]

2. Description of the Related Art In recent years, image communication apparatuses for transmitting image data to a communication partner have become remarkably widespread. In image communication, compressed and encoded still images are continuously transmitted in order to realize moving image communication. At this time, in order to keep the number of transmitted images per unit time constant, compressed image data is transmitted. Needs to be adjusted to a desired data length.

As shown in FIG. 3, a conventional image communication apparatus includes a spatial frequency converting means 1 for converting image data 101 having a predetermined number of pixels into spatial frequency components, and a spatial frequency data 102.
, A coding means 3 for coding the quantized data 104 so as to reduce the data length, a communication means 4 for transmitting the coded data 105, and a quantizer used for the quantization. A quantization table generating means 6 for generating a quantization table and a scaling factor generating means 5 for generating a scaling factor 107 for multiplying the quantization table 103 are provided.

[0004] In this device, the spatial frequency conversion means 1 comprises:
The image data 101 having a predetermined number of pixels is converted into a spatial frequency component, and is output as spatial frequency data 102 which is array data indicating the spatial frequency component. The quantization means 2 quantizes the spatial frequency data 102 by using the quantization table 103 output from the quantization table generation means 6, and outputs it as the quantized data 104.

[0005] The encoding means 3 encodes the quantized data 104 using a predetermined encoding table so that the data length becomes short. Then, the data length 106 of the encoded data 105 is output to the scaling factor generating means 5, and the data length
When 106 is shorter than the desired data length, the coded data 105 is output to the communication means 4. The communication means 4 includes the encoded data 105
To the communication partner.

When the data length 106 of the coded data 105 is larger than a predetermined data length, the scaling factor generating means 5 increases the scaling factor 107 which is a value to be multiplied by the quantization table 103, and Is smaller than the predetermined data length, the scaling factor 107 is reduced. The quantization table generating means 6 multiplies the predetermined table by the scaling factor 107 and outputs the result to the quantization means 2 as the quantization table 103.

With this configuration, when the data length 106 of the encoded data 105 is longer than the predetermined data length, the scaling factor 107 generated by the scaling factor generating means 5 becomes large.
Each value of 03 increases, and the compression ratio of image data increases.
Further, when the data length 106 of the encoded data 105 is smaller than the predetermined data length, the scaling factor 107 generated by the scaling factor generation unit 5 becomes smaller, so that each value of the quantization table 103 becomes smaller,
The compression ratio of image data decreases.

The quantizing means 2 again quantizes the spatial frequency data 102 using the updated quantization table 103, updates the quantized data 104, and outputs it. The encoding unit 3 encodes the updated quantized data 104 using a predetermined encoding table. If the data length 106 of the encoded data 105 is not shorter than the desired data length, the encoding unit 3 outputs the data length 106 to the scaling factor generation unit 5 again.

The above series of processing is repeated until the data length 106 becomes equal to or less than the desired data length. The encoding unit 3 outputs the encoded data 105 to the communication unit 4 when the data length 106 of the encoded data 105 becomes equal to or less than the desired data length, and the communication unit 4 transmits the encoded data 105 to the communication partner. .

[0010]

However, in the configuration of the conventional image communication apparatus, quantization and encoding are performed for one image until the data length of the encoded data reaches a desired data length. Since the processing is repeated, a larger number of processings than the number of images to be transmitted per unit time is required. Therefore, it is difficult for an apparatus having a low quantization and encoding processing speed to reduce the data length of encoded data to a desired data length or less.

In this case, if the scaling factor is set to be large from the beginning and the compression ratio of the image data is set to be high, the data length of the encoded data can be reduced to a desired data length or less by one quantization and encoding. However, the quality of the transmitted image is degraded due to higher compression.

The present invention solves such a conventional problem. The data length of encoded data can be made closer to a desired data length by a small number of processes of quantization and encoding. It is an object of the present invention to provide an image communication apparatus capable of making the data length of encoded data close to a desired data length while maintaining the image quality of the image data, and a compression encoding method thereof.

[0013]

Therefore, in the present invention, every time encoded data is transmitted, data representing a low compression rate of image data (ie, representing a large number of high frequency components of the image data). Is calculated as a ratio between a value obtained by dividing a predetermined fixed data length by a scaling factor and a data length of encoded data, and a scaling factor of a quantization table used for quantization of new image data is calculated in the past. Is obtained by multiplying the average value of the plurality of ratio data by a predetermined value.

Further, a value obtained by multiplying a difference between the data length of the encoded data and the desired data length by a predetermined number is added to a scaling factor used for quantizing the encoded data, and the added value is added to a new value. It is used as a scaling factor of a quantization table used for quantization of image data.

Therefore, every time the image data is compressed and transmitted, that is, as time elapses, the data length of the coded data approaches the desired data length, and one image corresponds to one image.
It is sufficient to perform quantization and encoding only once.

In the present invention, image data is demodulated by performing inverse quantization and inverse spatial frequency conversion on the quantized data, and a difference value between the image data before compression and the demodulated image data is compared with a predetermined comparison value. Then, when the difference value is larger than the predetermined comparison value, the scaling factor of the quantization table used for quantization of the new image data is reduced by a predetermined number from the previous scaling factor, and the difference value is smaller than the predetermined comparison value. In this case, the scaling factor of a quantization table used for quantization of new image data is increased by a predetermined number from the previous scaling factor.

Therefore, when the image quality of the demodulated image data is relatively good, the difference value becomes smaller than a predetermined comparison value,
At this time, the compression ratio is further increased. Conversely, when the image quality of the demodulated image data is relatively poor, the difference value becomes larger than a predetermined comparison value, and at this time, the compression ratio is reduced. Therefore, the data length of the encoded data can be made closer to the desired data length while maintaining the image quality of the predetermined level.

[0018]

DESCRIPTION OF THE PREFERRED EMBODIMENTS According to the first aspect of the present invention, image data is converted into spatial frequency data, the spatial frequency data is quantized based on a quantization table, and the quantized data is converted into a predetermined code. In an image communication apparatus that encodes based on an encoding table and transmits the encoded data to a communication partner, a scaling factor generating unit that generates a scaling factor, and a quantization table that multiplies the generated scaling factor by a predetermined table. And a storage means for storing numerical data used by the scaling factor generation means to generate the scaling factor. Each time the scaling factor generation means transmits encoded data, Calculates ratio data indicating the low compression ratio due to the large number of high frequency components in the data. When the image data to be compressed is updated in the storage unit, a predetermined number of previously written ratio data is read out from the storage unit, and the average value of the ratio data is multiplied by a predetermined value to generate a calling factor. Each time a new image is input, a calling factor is generated from the average value of the ratio data stored in the storage unit, and the ratio data is calculated from the encoded data and stored in the storage unit. It is memorized. By continuing this series of processing, the data length of the encoded data approaches a predetermined data length as time passes.

According to a second aspect of the present invention, the scaling factor generating means obtains ratio data from a ratio between a value obtained by dividing a predetermined fixed data length by a scaling factor and a data length of encoded data. Yes, the calculation result can provide data indicating a low compression ratio due to the large number of high frequency components of the image data.

According to a third aspect of the present invention, the image data is converted into spatial frequency data, the spatial frequency data is quantized based on a quantization table, and the quantized data is encoded based on a predetermined encoding table. In an image communication apparatus for transmitting encoded data to a communication partner, a scaling factor generating means for generating a scaling factor, and a quantization table generating means for generating a quantization table by multiplying a predetermined table by the generated scaling factor Means and storage means for storing numerical data used by the scaling factor generation means for generating the scaling factor, wherein the scaling factor generation means sets a predetermined number to a difference between the data length of the encoded data and the desired data length. The multiplied value is used as the scaling factor used to quantize the encoded data. Then, when the image data to be compressed is updated, the added value written last is read out from the storage means and used as a scaling factor. If the data length of the encoded data is larger than the desired data length, a difference obtained by subtracting the desired data length from the data length of the encoded data is multiplied by a predetermined number, and the multiplied value is added to the scaling factor. If the data length of the encoded data is smaller than the desired data length, a difference obtained by subtracting the data length of the encoded data from the desired data length is multiplied by a predetermined number, and the multiplied value is subtracted from the scaling factor. You. By performing this series of processes each time a new image is input, the data length of the encoded data approaches a predetermined data length as time passes.

According to a fourth aspect of the present invention, the image data is converted into spatial frequency data, the spatial frequency data is quantized based on a quantization table, and the quantized data is encoded based on a predetermined encoding table. In an image communication apparatus for transmitting encoded data to a communication partner, a scaling factor generating means for generating a scaling factor, and a quantization table generating means for generating a quantization table by multiplying a predetermined table by the generated scaling factor Means, inverse quantization means for inversely quantizing the quantized data based on the quantization table and outputting it as demodulated spatial frequency data, and converting the demodulated spatial frequency data to image data,
A spatial frequency inverse transforming means for outputting as demodulated image data; comparing a difference value between the image data before compression and the demodulated image data with a predetermined comparison value; if the difference value is larger than the predetermined comparison value, a scaling factor decrease instruction signal And a comparing means for outputting a scaling factor increase instruction signal to the scaling factor generating means when the difference value is smaller than the predetermined comparison value, wherein the scaling factor generating means , A scaling factor reduced or increased by a predetermined number from the previous scaling factor is generated, and when the image quality of the demodulated image data is relatively good, the difference value becomes smaller than a predetermined comparison value. Sometimes the compression ratio is further increased. Conversely, when the image quality of the demodulated image data is relatively poor, the difference value becomes larger than a predetermined comparison value, and at this time, the compression ratio is reduced. Therefore, the data length of the encoded data can be made closer to the desired data length while maintaining the image quality of the predetermined level.

According to a fifth aspect of the present invention, the image data is converted into spatial frequency data, the spatial frequency data is quantized based on a quantization table, and the quantized data is encoded based on a predetermined encoding table. In the method for compressing and encoding image data in an image communication apparatus for transmitting the encoded data to a communication partner, each time encoded data is transmitted, a value obtained by dividing a predetermined fixed data length by a scaling factor, Ratio data with the data length of data is obtained, and a value obtained by multiplying a predetermined value by an average value of a plurality of past ratio data is used as a scaling factor of a quantization table used for quantization of new image data. And a quantization table based on ratio data indicating a large number of high frequency components of the image data, that is, a low compression ratio of the image data. Scaling factor is set.

According to a sixth aspect of the present invention, a value obtained by multiplying a difference between a data length of encoded data and a desired data length by a predetermined number is added to a scaling factor used for quantization of encoded data. The added value is used as a scaling factor of a quantization table used for quantization of new image data, and by using this scaling factor, the data length of encoded data is set to a desired data length. You can get closer.

According to a seventh aspect of the present invention, image data is demodulated by performing inverse quantization and inverse spatial frequency conversion on the quantized data, and a difference value between the image data before compression and the demodulated image data is compared with a predetermined value. If the difference value is larger than the predetermined comparison value, the scaling factor of the quantization table used for quantization of new image data is reduced by a predetermined number from the previous scaling factor, and the difference value is compared with the predetermined value. If the value is smaller than the value, the scaling factor of the quantization table used for quantization of the new image data is increased by a predetermined number from the previous scaling factor, and the encoding is performed while maintaining the image quality of the predetermined level. The data length of the data can be made closer to the desired data length.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First Embodiment) As shown in FIG. 1, an image communication apparatus according to a first embodiment comprises: a spatial frequency conversion unit 11 for converting image data 111 having a predetermined number of pixels into spatial frequency data 112; Quantizing means 12 that quantizes the spatial frequency data 112 based on the quantization table and outputs it as quantized data 114, and outputs the quantized data 114 as coded data 115 based on a predetermined coding table. Encoding means 13 for outputting the data length 116 of the encoded data 115, communication means 14 for transmitting the encoded data 115 to a communication partner, and storage means 17 for storing input numerical data
A scaling factor generating means 15 for generating a scaling factor 117 for multiplying the quantization table by using the data stored in the storage means 17,
And a quantization table generating means 16 for multiplying a predetermined table by 117 and outputting the result to the quantization means 12 as a quantization table 113.

The spatial frequency conversion means 11 is, for example, a DC
A DSP for performing a T operation;
Is composed of, for example, a DSP for performing a division process, and the encoding means 13 is composed of a ROM in which Huffman code data is recorded, and a control CPU for extracting data from the ROM. The communication means 14 is composed of, for example, a PHS unit. Further, the quantization table generating means 16 is composed of, for example, a control CPU having a multiplication function, and the storage means 17
AM. Further, the scaling factor generating means 15 is constituted by, for example, a control CPU capable of performing a comparison operation process.

The operation of this image communication apparatus is the same as that of the conventional apparatus (FIG. 3) except that the operation of the scaling factor generating means 15 for generating the scaling factor is different.

That is, the spatial frequency converting means 11 converts the image data 111 of a predetermined number of pixels into spatial frequency components and outputs spatial frequency data 112, and the quantizing means 12 outputs from the quantization table generating means 12. The spatial frequency data 112 is quantized using a quantization table 113, and the quantized data 114 is output. The encoding means 13 outputs the quantized data 114
Is encoded by a predetermined encoding table, and encoded data is
The data length 116 of 115 is output to the scaling factor generation means 15, and when the data length 116 is equal to or less than the desired data length, the encoded data 115 is output to the communication means 14. Communication means 1
4 transmits the encoded data 115 to the communication partner. Further, the quantization table generation means 16 multiplies a predetermined table by a scaling factor 117 output from the scaling factor generation means 15 and outputs the result to the quantization means 12 as a quantization table 113.

In the initial stage, when the data length 116 of the encoded data 115 is far from the predetermined data length, the scaling factor generator 15 corrects the scaling factor 117 and updates the scaling factor 117 from the quantization table generator 16. The obtained quantization table 113 is output, and the quantization means 1
2 requantizes the spatial frequency data 112 using the quantization table 113, and updates the quantized data 114.
Further, the encoding unit 13 encodes the quantized data 114 using an encoding table.

In this processing, the data length 116 of the encoded data
Is repeated until the data length reaches the desired data length. When the data length 116 of the encoded data 115 reaches the desired data length, the encoded data 115 is output from the communication means 14.

The scaling factor generating means 15 uses the data having the data length 116 and the value of the scaling factor 117 to generate a scaling factor of a quantization table used for quantizing image data to be input thereafter. Thereby, the scaling factor generating means
From 15, a scaling factor is generated such that quantization and encoding need only be performed once for one image.

For this purpose, the scaling factor generating means 15 performs the following processing.

The scaling factor generating means 15 converts a predetermined fixed data length into a scaling factor in order to indicate a low compression ratio of image data, that is, a large number of high frequency components.
The ratio between the value divided by 117 and the data length 116 is the ratio data 11
It is calculated as 8 and stored in the storage means 17. Now, assuming that a predetermined fixed data length is a, a scaling factor is α ₁ , and a data length 116 is b ₁ , ratio data b ₁
/ (A / α ₁ ) is stored in the storage means 17.

The scaling factor generating means 15
Receives an image update signal 119 indicating that the image data to be compressed has been updated, reads out a predetermined number (n) of the ratio data 118 written in the past from the storage unit 17, and calculates the average value of the ratio data 118. Ask. This average value is as follows. (B ₁ · α ₁ + b ₂ · α ₂ + ‥ + b _n · α _n ) / na The value obtained by multiplying the average value by a predetermined value is output to the quantization table generation means 16 as a scaling factor 117 for a new image. I do.

Each time the image update signal 119 is received, the scaling factor generator 15 stores the memory 17
The ratio data 118 written in the past is read out by a predetermined number, an average value thereof is obtained, and a value obtained by multiplying the average value by a predetermined value is output to the quantization table generation means 16 as a scaling factor 117 for a new image. , Encoded data
Each time 115 is output from the encoding means 13 to the communication means 14, the ratio data is obtained from the data length 116 of the encoded data 115 at that time and the scaling factor 117, and the storage means 1
Store it in 7.

The scaling factor generating means 15 performs this series of processing each time a new image is input, so that the data length of the encoded data approaches a predetermined data length with the passage of time.

(Second Embodiment) In the second embodiment,
Another method for generating the scaling factor will be described. The block configuration of the image communication apparatus that implements this method is the same as that of the first embodiment (FIG. 1), and differs only in the operation of the scaling factor generator 15 for generating the scaling factor.

The scaling factor generating means 15
If the data length 116 output from the encoding means 13 is larger than the desired data length, a value obtained by multiplying the difference between the data length 116 and the desired data length by a predetermined number is added to the scaling factor 117, The value is stored in the storage unit 17, and if the data length 116 output from the encoding unit 13 is smaller than the desired data length, a predetermined number is set to the difference between the data length 116 and the desired data length. The multiplied value is subtracted from the scaling factor 117, and the value is stored in the storage unit 17.

When an image update signal 119 indicating that the image data to be compressed has been updated is received, the last stored scaling factor is read from the storage means 17 and the scaling factor 11 for the new image is read.
The value 7 is output to the quantization table generating means 16.

The scaling factor generating means 15 performs this series of processing each time a new image is input, so that the data length of the encoded data approaches a predetermined data length as time passes.

(Third Embodiment) The image communication apparatus of the third embodiment demodulates compressed image data, compares the demodulated image data with image data before compression, and adjusts a scaling factor. I do.

This apparatus, as shown in FIG. 2, quantizes image data 131 having a predetermined number of pixels into spatial frequency data 132 based on a quantization table 133. A quantizing unit 32 that outputs the quantized data 134; an encoding unit 33 that encodes the quantized data 134 based on a predetermined encoding table and outputs the encoded data 135; and transmits the encoded data 135 to a communication partner. Communication means 34, inverse quantization means 37 for inversely quantizing the quantized data 134 based on the quantization table 133 and outputting demodulated spatial frequency data 136, and demodulated image data by converting the demodulated spatial frequency data 136 back to image data. The spatial frequency inverse transform means 38 which outputs 138, the original image data 131 and the demodulated image data 138 are compared, and the scaling factor decrease instruction signal 139 or the scale A comparison means 39 for outputting a scaling factor increasing instruction signal 140, a scaling factor generating means 35 for generating a scaling factor 137 for multiplying the quantization table according to the instruction signal, and a quantization table 133 by multiplying a predetermined table by the scaling factor 137. It comprises a quantization table generating means 36 for outputting to the quantization means 32 and the inverse quantization means 37.

The inverse quantization means 37 is constituted by, for example, a DSP for performing a division process.
It is composed of a DSP that performs DCT operation. Also, the comparison means
Reference numeral 39 denotes a CPU having a comparison operation function.

In this device, image data 13 of a predetermined number of pixels
1 is input to the spatial frequency conversion means 31 and the comparison means 39.
The spatial frequency conversion unit 31 converts the image data 131 into spatial frequency components and outputs spatial frequency data 132.The quantization unit 32 uses the quantization table 133 output from the quantization table generation unit 36, The spatial frequency data 132 is quantized, and the quantized data 134 is output to the encoding unit 33 and the inverse quantization unit 37. The encoding means 33 outputs the quantized data 13
4 is encoded by a predetermined encoding table, and the encoded data 115 is output to the communication means 14. The communication means 34 transmits the encoded data 115 to the other party.

The inverse quantization means 37 outputs the quantized data 13
4 is inversely quantized based on the quantization table 133 and output to the spatial frequency inverse transform means 38 as demodulated spatial frequency data 136, and the spatial frequency inverse transform means 38
136 is returned to image data, and is output to the comparing means 39 as demodulated image data 138.

The comparing means 39 compares a difference value between the previous image data 131 and the demodulated image data 138 with a predetermined data value, and when the difference value is larger than the predetermined value, outputs a scaling factor reduction instruction signal 139 and When the difference value is smaller than the predetermined value, the scaling factor increasing instruction signal 140 is output to the scaling factor generating means 35.

When the scaling factor decreasing instruction signal 139 is output, the scaling factor generating means 35 reduces the current scaling factor 137 by a predetermined number and outputs it to the quantization table generating means 36, and further increases the scaling factor. When the instruction signal 140 is output, the current scaling factor 137 is increased by a predetermined number and output to the quantization table generating means 36.

The quantization table generating means 36 multiplies a predetermined table by the scaling factor 137 output from the scaling factor generating means 35, and
Is output to the quantization means 32 and the inverse quantization means 37.

The quantization means 32 is a quantization table generation means.
The spatial frequency data 132 of the new image data is quantized using the quantization table 133 output from 36.

As described above, in this image communication apparatus, the scaling factor used for the next compression is adjusted based on the magnitude of the difference between the image data before compression and the demodulated image data. When the image quality of the demodulated image data is relatively good, this difference value is smaller than the predetermined comparison value. At this time, the compression ratio is further increased, and when the image quality of the demodulated image data is relatively poor, The difference value becomes larger than a predetermined comparison value, and at this time, an adjustment to reduce the compression ratio is performed.

By doing so, the data length of the encoded data can be made closer to the desired data length while maintaining a predetermined level of image quality.

[0053]

As is clear from the above description, in the image communication apparatus and the compression encoding method of the present invention, the quantization table is adjusted according to the previous compression result every time image data is compressed. , The data length after compression can be made closer to the desired data length as time passes,
Further, the data length of the encoded data can be made closer to a desired data length while maintaining the image quality.

Therefore, the effect of keeping the number of transmitted images per unit time constant when continuously transmitting still images can be obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram of an image communication device according to first and second embodiments of the present invention;

FIG. 2 is a block diagram of an image communication device according to a third embodiment of the present invention;

FIG. 3 is a block diagram of a conventional image communication device.

[Explanation of symbols]

 1, 11, 31 Spatial frequency conversion means 2, 12, 32 Quantization means 3, 13, 33 Encoding means 4, 14, 34 Communication means 5, 15, 35 Scaling factor generation means 6, 16, 36 Quantization table generation Means 17 Storage means 37 Inverse quantization means 38 Spatial frequency inverse transformation means 39 Comparison means

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Ken Ishihara 4-3-1 Tsunashima Higashi, Kohoku-ku, Yokohama-shi, Kanagawa Prefecture Inside Matsushita Communication Industrial Co., Ltd. 3-1, Matsushita Communication Industrial Co., Ltd. (72) Inventor Takuyuki Ishii 4-3-1 Tsunashima Higashi, Kohoku-ku, Yokohama, Kanagawa Prefecture Inside Matsushita Communication Industrial Co., Ltd. (72) Toshio Kobori, Yokohama, Kanagawa Prefecture 4-3-1 Tsunashima Higashi, Kohoku-ku Matsushita Communication Industrial Co., Ltd. F term (reference) 5C059 KK22 KK23 MC15 ME01 PP01 PP04 SS06 SS10 TA48 TC08 TC10 TC20 TC38 TD03 TD11 5C078 AA04 BA32 BA42 BA64 CA31 DB00 DB07 5J064 AA01 BA01 BB13 BC BD02

Claims (7)

[Claims] An image data is converted into spatial frequency data, the spatial frequency data is quantized based on a quantization table, the quantized data is encoded based on a predetermined encoding table, and the encoded data is communicated. An image communication device for transmitting to the other party; a scaling factor generating means for generating a scaling factor; a quantization table generating means for generating the quantization table by multiplying a generated table by the generated scaling factor; Storage means for storing numerical data used by the factor generation means for generating a scaling factor, wherein the scaling factor generation means, each time the encoded data is transmitted, a compression ratio based on the amount of high frequency components of the image data Is calculated and written in the storage means. When the image data to be compressed is updated, a predetermined number of the ratio data previously written from the storage unit are read out, and the average value of the ratio data is multiplied by a predetermined value to generate the calling factor. An image communication device, comprising: 2. The method according to claim 1, wherein the scaling factor generation unit includes:
The image communication apparatus according to claim 1, wherein the ratio data is obtained from a ratio of a value obtained by dividing a predetermined fixed data length by a scaling factor and a data length of the encoded data. 3. The image data is converted into spatial frequency data, the spatial frequency data is quantized based on a quantization table, the quantized data is encoded based on a predetermined encoding table, and the encoded data is communicated. An image communication device for transmitting to the other party; a scaling factor generating means for generating a scaling factor; a quantization table generating means for generating the quantization table by multiplying a generated table by the generated scaling factor; Storage means for storing numerical data used by the factor generation means for generating a scaling factor, wherein the scaling factor generation means multiplies a difference between the data length of the encoded data and a desired data length by a predetermined number. The value is a scaling factor used for quantization of the encoded data. And the added value is written to the storage means, and when the image data to be compressed is updated, the last written value is read from the storage means and used as a scaling factor. Image communication device. 4. The image data is converted into spatial frequency data, the spatial frequency data is quantized based on a quantization table, the quantized data is encoded based on a predetermined encoding table, and the encoded data is communicated. An image communication device for transmitting to the other party; a scaling factor generating unit for generating a scaling factor; a quantization table generating unit for generating the quantization table by multiplying a generated table by the generated scaling factor; Inverse quantization means for inversely quantizing the quantized data based on the quantization table and outputting the demodulated spatial frequency data as demodulated spatial frequency data, and spatial frequency inverse transforming means for converting the demodulated spatial frequency data to image data and outputting as demodulated image data, The difference value between the image data before compression and the demodulated image data is compared with a predetermined comparison value. The scaling factor reduction instruction signal when the difference value is greater than the predetermined comparison value,
A comparing unit that outputs a scaling factor increase instruction signal to the scaling factor generation unit when the difference value is smaller than the predetermined comparison value, wherein the scaling factor generation unit is configured to output a scaling factor increase instruction signal based on the instruction signal from the comparison unit. And generating a scaling factor reduced or increased by a predetermined number from a previous scaling factor. 5. Converting image data to spatial frequency data, quantizing the spatial frequency data based on a quantization table, encoding the quantized data based on a predetermined encoding table, and communicating the encoded data for communication. In a method for compressing and encoding image data in an image communication device to be transmitted to a partner, each time encoded data is transmitted, a value obtained by dividing a predetermined fixed data length by a scaling factor and a data length of the encoded data are obtained. Compression coding, wherein ratio data is obtained and a value obtained by multiplying a predetermined value by an average value of a plurality of past ratio data is used as a scaling factor of a quantization table used for quantization of new image data. Method. 6. The image data is converted to spatial frequency data, the spatial frequency data is quantized based on a quantization table, the quantized data is encoded based on a predetermined encoding table, and the encoded data is communicated. In a method for compressing and encoding image data in an image communication device to be transmitted to a partner, a value obtained by multiplying a difference between a data length of the encoded data and a desired data length by a predetermined number is used for quantization of the encoded data. Add to the used scaling factor and add the sum to
A compression coding method characterized by using as a scaling factor of a quantization table used for quantization of new image data. 7. Converting image data into spatial frequency data, quantizing the spatial frequency data based on a quantization table, encoding the quantized data based on a predetermined encoding table, and transmitting the encoded data for communication. In the method for compressing and encoding image data in an image communication device to be transmitted to the other party, the image data is demodulated by performing inverse quantization and inverse spatial frequency transformation from the quantized data, and the image which has been demodulated with the image data before compression Comparing the difference value with the data with a predetermined comparison value, and when the difference value is larger than the predetermined comparison value,
The scaling factor of a quantization table used for quantization of new image data is reduced by a predetermined number from the previous scaling factor, and when the difference value is smaller than the predetermined comparison value, the scaling factor is used for quantization of new image data. A compression coding method characterized by increasing the scaling factor of a quantization table to be performed by a predetermined number from the previous scaling factor.