JPH09261637A - Image compressor, image pre-processing method and image transmission system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize an image compression/transmission/expansion system with high real time performance. SOLUTION: In parallel with write of image obtained by a camera to a frame memory 16 via an A/D converter 14, an maximum/minimum luminance calculation section 32 obtains a maximum luminance, a minimum luminance and a difference between the both (luminance change) in one frame of the image. Image data read out of the frame memory 16 are subjected to luminance expansion conversion based on information obtained by the maximum/minimum luminance calculation section 32 and the result is fed to an image compression section. Otherwise, a quantization coefficient is set based on the information obtained by the maximum/minimum luminance calculation section 32. A flatness degree of the image is detected and evaluated by utilizing the luminance change in one frame and the result is used to change the gradation or the quantization coefficient of the image, then a more excellent real time processing is realized because no flatness evaluation memory is required.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image compression apparatus for compressing a still image or the like prior to communication or the like, and further, an image preprocessing method executed prior to image compression or the like, or transmitting a compressed image. The present invention relates to an image transmission system.

[0002]

2. Description of the Related Art Since image information obtained by image pickup has a large amount of information and its transmission and management are not easy,
A compression / expansion process is executed at the time of communication. FIG.
An example of the image compression / transmission / decompression system is shown in FIG. In this system, an image signal from the camera 10 or the like is first converted into multi-tone image data by the A / D conversion unit 14 (see FIG. 20) of the image storage unit 12 and written in the frame memory 16. The image data on the frame memory 16 is compressed by the image compression unit 18, and is supplied to the communication line 20 or stored in the storage medium 22. The image data on the communication line 20 or the image data in the storage medium 22 is read by the image decompression unit 24 of the receiving side device, and the image decompression unit 24 releases the compression processing by the image compression unit 18 (decompression). The decompressed image data is written in the frame memory 26, and an image corresponding to the image data in the frame memory 26 is displayed on the screen of the monitor 28. In this system, since image compression is performed by the image compression unit 18, the occupied bandwidth on the communication line 20 is narrow or the transmission time is short, and the occupied capacity on the storage medium 22 is small.

Among the image compression / decompression processes, first, there is a process in which information loss does not occur during compression. While this type of image compression / expansion processing is suitable for applications that require high-quality image reception, it is difficult to meet the demands for information volume reduction, high-speed communication, and easy management. Secondly, in image compression / expansion processing, information loss occurs during compression. In this type of image compression / expansion processing, a method of removing an element which is not so important but has a relatively large amount of information in the image is usually used. For example, a small area (“conversion unit”. For example, in JPEG or MPEG, a rectangular area of 8 pixels × 8 pixels. JPEG: Joint Photog
raphic Experts Group, MPEG: MovingPicture Expe
When the brightness change appearing in the rts Group) is small, the brightness change is treated as 0 and the brightness of the pixel belonging to the conversion unit is converted to a constant value. When the amplitude of a certain spatial frequency is small, the amplitude is treated as 0. The method of replacing the luminance change or the spatial frequency component amplitude with 0 as described above is effective for reducing the amount of information, that is, image compression.
In a normal application, even if the image quality is lowered to some extent, if it is possible to replace it with a reduction in the amount of information and the like, it is practical and economical, and therefore processing involving information loss is often used.

However, the compression process with information loss has a drawback that it is not suitable for so-called flat images. First, when the brightness change in one screen (frame) of the uncompressed image is small, such as when the shooting environment is dark, the brightness change in each conversion unit is also small. Therefore, during compression, the brightness value of the pixel belonging to the conversion unit is small. Is converted to a constant value. This means that a slight change in luminance, which is originally important in a flat image, is missing.

According to the image compression apparatus disclosed in Japanese Unexamined Patent Publication No. 6-46401, it is possible to prevent and overcome these problems. In the apparatus disclosed in this publication, first, the image data is subjected to high-pass filtering in each conversion unit, then the flatness degree of the image in the conversion unit is evaluated, and the obtained score is obtained. Are summed up for one screen. The total result represents the flatness of the entire screen. In this publication, a parameter called a quantization coefficient or a quantization width is set according to the degree of flatness of the entire screen. The quantization coefficient or the quantization width here is a spatial frequency component corresponding to the number of pixels per transform unit (the size of the transform unit is 8 pixels × 8.
It is a parameter set for each of 64 components when it is a pixel), and the degree of compression is higher as the quantization coefficient is larger and lower as it is smaller. In this way, by substituting the quantization coefficient for the flatness of the entire screen,
It is possible to prevent significant deterioration of image quality due to image compression even when the brightness change in one screen of the image before compression is small, such as when the shooting environment is dark.

[0006]

However, in the image compression device disclosed in the above publication, the score evaluation is executed for each conversion unit. In order to execute this processing, the brightness value must be known in advance for all the pixels forming the conversion unit. That is, in order to realize the principle disclosed in this publication, in addition to the frame memory for storing the image data to be compressed, the image data is captured until at least the data for the score evaluation is collected. A memory (buffer) for setting the conversion coefficient is required. Furthermore, the processing delay occurs due to the acquisition of the image data in this memory, and the subsequent processing such as score evaluation and totaling, which impairs real-time performance.

Further, regardless of whether the method described in the above publication is used or not, when the luminance change in one frame of the image before compression is small, only about 30% of the prepared gradation may be used. . That is, assuming that the image data is expressed by 8 bits = 256 gradations, only 256 × 30% = 77 gradations may be used. Gradations such as gradation values 0 and 255 are rarely used. The problem that leads to such uneconomical gradation expression cannot be solved by the principle disclosed in the above publication.

In addition, in order to implement the principle disclosed in the above publication, means for predicting the code amount for one frame,
Means for applying high- and middle-pass filtering to image data,
Since the score evaluation / aggregation means is required, the hardware or software for image compression must be modified. Therefore, image compression / sold at a relatively low price
The expansion IC cannot be used, which causes problems such as difficulty in implementation and cost increase.

One of the objects of the present invention is to enable image quality deterioration prevention processing for a flat image to be executed concurrently with writing / reading related to a frame memory, thereby enabling image transmission with higher real-time characteristics. It is to be realized. One of the objects of the present invention is to make it possible to use an image compression / decompression IC that is sold at a relatively low price, by eliminating the need to modify the image compression procedure, and thus to realize an easy and inexpensive implementation. To enable. One of the objects of the present invention is to make it possible to obtain a bright screen at all times and to perform economical gradation expression by making all the prepared gradations available. One of the objects of the present invention is to make it possible to maintain the lightness and darkness of the original image by executing the image quality deterioration prevention processing without modifying the gradation value, and at that time it is necessary to transmit additional information. It is to prevent it.

[0010]

In order to achieve such an object, the image compression apparatus according to the first configuration of the present invention is
From the image signal or multi-tone image data corresponding to this,
Means for detecting the amount of change in luminance on one screen, and means for quantizing and compressing the image data by rounding the fractional part of the product of the gradation value of each pixel multiplied by the reciprocal of the quantization coefficient or the inverse matrix And means for correcting or setting the gradation value and / or the quantized coefficient according to the detected brightness change amount prior to the compression so that the product becomes large. It is characterized in that a comparatively small luminance change in the previous image data is retained or emphasized in the compressed image data.

In the image compression method premised on the present configuration, by rounding the fractional part of the product of the gradation value of each pixel and the reciprocal of the quantization coefficient or the inverse matrix, the gradation value changes even in the image data. A small portion (that is, a portion having a small change in luminance value in the image signal) is omitted or ignored, and the image data is thus compressed. Therefore, when the scale of the gradation value is large or the quantization coefficient is small, the degree of compression is small. Therefore, in the present configuration, the product of the gradation value and the reciprocal of the quantized coefficient or the inverse matrix is increased,
At least one of the gradation value and the quantized coefficient is corrected or set prior to the compression and according to the detected luminance change amount.

As described above, by correcting or setting at least one of the gradation value and the quantized coefficient so that the product becomes large, a relatively small change in brightness in the image data before compression causes a change in the image data after compression. Also, it is held or emphasized. Further, since the flatness degree of the entire screen is detected by the score evaluation for each conversion unit and the brightness change amount detection over the entire screen instead of the total score over the entire screen,
A means for storing image data for evaluation of the degree of flatness is not required, and the process of writing the image data to be compressed in the frame memory and the detection of the degree of flatness can be performed simultaneously in parallel. This improves the real-time property. Further, since the correction or setting of the gradation value and the quantization coefficient is executed as a pre-process of the image compression, it is not necessary to modify the conventional image compression procedure, and therefore, it is sold at a relatively low price. An image compression / decompression IC can be used, which enables easy and inexpensive implementation.

The image preprocessing method according to the second aspect of the present invention comprises a step of detecting a luminance change amount on one screen from an image signal or multi-tone image data corresponding thereto, and a step of detecting the brightness change amount on the one screen. A step of correcting the brightness value of the image signal or the gradation value of each pixel of the image data according to the detected brightness change amount so that a relatively small brightness change is emphasized. . In this configuration, in addition to the same operation as the first configuration, all the prepared gradations can be used, resulting in an efficient and economical gradation expression and always bright regardless of the lightness and darkness of the original image. The screen is obtained. Furthermore, since this configuration does not require manipulation of the quantization coefficient, it can be implemented regardless of the image compression method, and can also be implemented as brightness extension processing independent of image compression.

An image compression apparatus according to a third configuration of the present invention is a means for inputting an image signal, converting it into multi-tone image data, and storing it, and an input image signal or stored image data. By executing the image preprocessing method according to the second configuration as a target, the gradation value of each pixel is corrected so that a relatively small change in luminance on one screen is emphasized, and then the stored image data is stored. And a means for compressing the image data after correcting the gradation value of each pixel. In this configuration, the same operation as in the second configuration occurs regarding the compression / transmission of image data. Further, since the gradation value is corrected after storing the image data (for example, in the frame memory), an image based on the stored image data, that is, an image in which the gradation value is not corrected is transmitted via a monitor or the like. It is also possible to provide it to the user of the side device.

An image transmission system according to a fourth aspect of the present invention is a transmission side apparatus including an image compression apparatus according to the third configuration and means for outputting compressed image data onto a communication line or a storage medium. A receiving side device having means for restoring image data before compression from image data received via a communication line or a storage medium. In this configuration, the same action as the third configuration occurs. In addition to this, since the compressed image is used for transmission or storage, the uneconomical use of the transmission band is reduced.

An image transmission system according to a fifth aspect of the present invention is the image transmission system according to the fourth aspect, wherein the transmitting side device outputs the information about the luminance change together with the compressed image data onto a communication line or a storage medium. The receiving side device is provided with means for restoring the image data before gradation value correction based on the information regarding the luminance change received via the communication line or the storage medium. In the present configuration, in addition to the action that has occurred in the fourth configuration, the action that the image data before tone value correction can be restored in the receiving side device also occurs.

The image preprocessing method according to the sixth aspect of the present invention comprises the steps of inputting an image signal having a predetermined format, converting the image signal into multi-tone image data, and storing the image signal, and inputting the input image signal or storage. By executing the image preprocessing method according to the second configuration on the image data being processed, a relatively small change in brightness on one screen is emphasized.
The step of correcting the gradation value of each pixel of the image data, and the image data after correcting the gradation value of each pixel are inversely converted into an image signal of the same format as the input image signal, and then output. And a step. In the present configuration, the image data whose gradation value has been corrected by the method according to the second configuration has the original image signal format (for example, NTS).
It is converted back into an analog signal format such as C). Therefore, since the image signal after the inverse conversion can be processed in the same manner as the conventional method, the means for executing the method according to the present configuration is inserted by, for example, inserting it between the camera and the image storage unit in the conventional apparatus. The configuration can be implemented and is therefore inexpensive and easy to implement.

The image preprocessing method according to the seventh aspect of the present invention comprises a step of detecting a luminance change amount on one screen from an image signal or multi-tone image data corresponding thereto, and a compressed image. The quantization coefficient that determines the threshold value of the brightness change amount that is considered as no brightness change during image compression processing is set to the detected brightness change amount so that a relatively small brightness change on one screen remains on the data. And a step of setting according to the above. In the present configuration, in addition to the action similar to the first configuration, since the gradation value does not change, the effect that the brightness and the like in the original image also appears in the expanded image. At that time, unlike the fifth configuration, it is not necessary to transmit information indicating a change in luminance.

An image preprocessing method according to an eighth structure of the present invention is the second or seventh structure, further including a step of inputting an image signal, converting the image signal into multi-tone image data, and storing the image data. However, it is characterized in that the input and storage of the image signal and the detection of the luminance change amount are executed at least partially at the same time. In the present configuration, the image quality deterioration prevention process for a flat image is executed at least partially in parallel with the image data acquisition, so that image transmission with high real-time property is realized.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below with reference to the drawings. Note that, for simplification of description, in the following description, the reference numerals used in the description of the related art in FIGS. 19 and 20 are used. However, this means that each member in the embodiment has the same reference numeral in the related art. It does not imply that the structure is the same as that of the member of FIG. Further, although the description by the block diagram and the flowchart is adopted, these are not intended to be limited to the implementation by hardware or software, respectively.

(1) First Embodiment FIGS. 1 and 2 show image compression / compression according to the first embodiment of the present invention.
1 shows the configuration of a transmission / expansion system. In this embodiment,
A data expansion unit 30 is provided instead of the image storage unit 12 in the conventional technique of FIG. Data expansion unit 30
Is obtained by the camera 10 as shown in FIG.
The image signal conforming to the standards such as the above is converted into multi-tone image data by the A / D converter 14, and this is converted into the frame memory 1
6, the image data read from the frame memory 16 is subjected to the brightness extension processing, and then supplied to the image compression unit 18. The image data obtained from the data expansion unit 30 is compressed by the image compression unit 18, and is supplied to the receiving side device via the communication line 20 or the storage medium 32.
In the device on the receiving side, the image decompression unit 24 decompresses the received image data, writes the decompressed image data in the frame memory 26, and displays the corresponding image on the monitor 28. In addition, as shown in FIG.
The image expansion unit 24A expands the received image data, the data expansion cancellation unit 24B performs the brightness expansion cancellation process on the expanded image data, writes the result in the frame memory 26, and displays the corresponding image on the monitor 28. You may do it.

One of the features of this embodiment is the data expansion unit 30, and in particular, the brightness expansion processing executed thereby.
The data expansion unit 30 shown in FIG. 2 has a maximum / minimum luminance calculation unit 32, a subtractor 34, and a multiplier / divider 36 as means for that purpose.
have. The maximum / minimum brightness calculation unit 32 detects the maximum brightness, the minimum brightness, and the difference (brightness change amount) between the corresponding image signals in one frame from the image data obtained from the A / D converter 14, and the result is detected. Is supplied to the subtractor 34, the multiplier / divider 36, and the image compression unit 18. Subtractor 34
The multiplier / divider 36 is means for performing gradation value correction calculation on the image data read from the frame memory 16. In particular, the subtractor 34 subtracts the minimum brightness detected by the maximum / minimum brightness calculation unit 32 from the gradation value (brightness value in the case of an image signal) of each pixel read from the frame memory 16. The multiplier / divider 36 divides the gradation value of each pixel of the image data obtained from the subtractor 34 by the brightness change amount detected by the maximum / minimum brightness calculation unit 32, and further multiplies the gradation number. In the example of FIG. 2, the number of output bits of the A / D converter 14 =
8, that is, the number of gradations = 2 ⁸ = 256. By the subtractor 34 and the multiplier / divider 36, the following equation

[Formula 1] (Gradation value (luminance value) -minimum luminance) x number of gradations (=
256) / luminance change amount, the result is that the minimum luminance is the lowest gradation,
The gradation value of each pixel of the image data is corrected so that the maximum brightness is represented by the highest gradation, that is, all the prepared 0th to 255th gradations are used (brightness extension processing). .

4 to 8 show the operation procedure of this embodiment. In these figures, the image signal from the camera 10 is NT
SC-compliant grayscale image signal and JPE
Assuming still image transmission conforming to G, the number of gradations = 25
6 and the quantization coefficient is constant and invariant (that is, not variably set as in other embodiments described later).

First, the transmitting side device, as shown in FIG.
First, the image signal is taken in from the camera 10 (100), then the brightness extension processing is executed (200), and finally the image compression processing is executed (300). When capturing the image signal,
As shown in FIG. 5, first, the variables storing the maximum brightness and the minimum brightness are set to 255 (highest gradation) and 0 (lowest gradation) for each frame by the maximum / minimum brightness calculation unit 32.
Initialize to (110). The transmitting device performs steps 121-126 for all the pixels that make up the frame (120). First, in step 121, the brightness value of the pixel is converted into an 8-bit digital value (gradation value) by the A / D converter 14, and in step 122, this gradation value is written in the frame memory 16. In steps 123 and 125, the maximum / minimum brightness calculation unit 3
2 compares the gradation value obtained in step 121 with the values of variables storing the maximum brightness and the minimum brightness, respectively.
As a result, when it is found that the gradation value obtained in step 121 is larger (123) and when it is found smaller (125), the maximum / minimum luminance calculation unit 3
2 substitutes the gradation value into the corresponding variable (124,
126). Therefore, by executing steps 121 to 126 for all the pixels in one frame, the values of the maximum brightness and the minimum brightness in the frame can be detected by the above variables. The maximum / minimum brightness calculating unit 32 subtracts the minimum brightness from the maximum brightness thus obtained to obtain the brightness change amount (130).

Next, in the brightness extension processing (200),
As shown in FIG. 6, steps 211 to 215 are executed for all pixels constituting one frame (210). First, in step 211, the gradation value of each pixel is read from the frame memory 16 to the subtractor 34, and in step 212, the subtractor 34 detects the gradation value from the gradation value in the maximum / minimum luminance calculating section 32. The minimum brightness is reduced. In step 213, the number of gradations (maximum range = 256 in the figure) is multiplied by the multiplier / divider 34, and step 2
In 14, the multiplier / divider 34 divides the gradation value after multiplication by the amount of change in brightness detected by the maximum / minimum brightness calculator 32. The gradation value thus obtained, that is, the gradation value subjected to the brightness expansion processing, is output from the multiplier / divider 34 to the image compression unit 18 (215).

At the time of image compression (300), as shown in FIG. 7, steps 311 to 314 are executed for the entire one frame (310). That is, the image compression unit 18 takes in the image data after the luminance expansion processing for each conversion unit (311), and performs orthogonal transformation (312) such as DCT (Descrete Cosine Transfomation) to reduce the information amount. For quantization (313) and entropy coding (314) for Huffman transform etc. capacity reduction. When the present embodiment is actually carried out, it is not necessary to change the conventional compression procedure as described above, and therefore the detailed description of steps 312 to 314 will be omitted here.

According to the procedure shown in FIGS. 4 to 7, therefore, the process is executed at the timing shown in FIG.
As is clear from this figure, in the present embodiment, the process of writing the image data to be compressed into the frame memory 16 (the process at the uppermost stage in FIG. 8) and the detection of the amount of change in luminance, that is, the degree of flatness (the second stage). (Processing) and are simultaneously executed in parallel, so that processing with high real-time property is possible. Further, the processing by the subtractor 34 and the multiplier / divider 36 (the processing in the third stage) and the processing in the image compression unit 18 (the processing in the fourth stage) can be simultaneously executed in parallel for at least each conversion unit. Therefore, also in this aspect, high-real-time processing becomes possible. It should be noted that although the adjustment of the brightness by the aperture control is well known in the related art, the process of adjusting the brightness on the basis of the image processing as in the present embodiment and the above publication is essentially faster than the aperture control. It is unsuitable to compare etc. with aperture control.

As described above, in this embodiment, since the image data subjected to the brightness extension processing is compressed, even when the brightness change amount in one frame in the image signal output from the camera 10 is small. The luminance change in that frame is not significantly lost due to compression. Therefore, the device on the receiving side can always display a clear image with a clear contrast on the monitor 28 (emphasizing the change in brightness). This is particularly effective when it is necessary to transmit an image obtained by photographing in darkness or dusk, such as license plate photographing and number image transmission of a vehicle that violates billing. In addition, since all the prepared gradations are used in association with the brightness extension processing, uneconomical gradation expression is reduced.

Further, in the present embodiment, the degree of flatness (small change in brightness) in one frame is detected and evaluated by detecting the amount of change in brightness in the entire frame, and the subtracter 34 and The brightness extension processing by the multiplier / divider 36 is being executed. On the other hand, in the configuration of the above publication, the degree of flatness in one frame is detected and evaluated by evaluating the score for each conversion unit and summing the scores over the entire screen. Therefore, in the latter case, a memory for storing the image data which is the basis of the evaluation of the degree of flatness is required separately, but in the former case, such a memory is not necessary because successive comparison is sufficient as shown in FIG.
Therefore, the present embodiment can be implemented easily and cheaply as compared with the related art.

Further, since the conventional image compression procedure is modified in the configuration of the above publication, a dedicated image compression / decompression IC is required for implementation, but in the present embodiment, the conventional image compression is performed. The image compression / decompression IC sold at a relatively low price can be used because the image compression pre-processing is performed without any modification to the procedure described in (1) above, which enables easy and inexpensive implementation.

In addition, in this embodiment, since the gradation data is corrected after the image data is stored in the frame memory 16, a device (not shown) for displaying an image using the frame memory 16 is used. By installing in the device on the transmission side,
It is also possible to provide the user with an image based on the image data stored in the frame memory 16, that is, an image in which the gradation value is not corrected. Further, the image data subjected to the brightness extension processing tends to be data indicating a bright image regardless of whether the original image is bright or dark. Therefore, in this embodiment, a brighter image can be always displayed on the screen of the monitor 28. Further, the information obtained by the maximum / minimum brightness calculation unit 32, that is, the information related to the brightness change such as the maximum brightness, the minimum brightness, and the brightness change amount is transmitted via the image compression unit 18 together with the image data to the communication line 20. Alternatively, by outputting the image on the storage medium 22, the receiving device can display the image on the monitor 28 at the same brightness as the original image from the camera 10. For example, as shown in FIG. 3, the image data received from the communication line 20 or the storage medium 22 is used as the image expansion unit 2.
4A, and the data expansion canceller 24B based on the brightness change information received from the communication line 20 or the storage medium 22.
The brightness extension canceling process may be performed on the image data from the image expanding unit 24A.

(2) Second to Fourth Embodiments In FIG. 9, when the processing according to the present invention is executed for a color image instead of a grayscale image, the transmission side apparatus executes the processing in place of the procedure of FIG. The procedure is shown below. In the figure, the same parts as those in FIG. 5 are designated by the same reference numerals. In the second embodiment, since a color image is targeted, A
The processes after the / D conversion are executed for all pixels (120) and all colors (R, G, B) (120A).
Further, in order to prevent an imbalance between the gradation values among R, G, and B due to the brightness extension processing, the minimum brightness is set to a priori 0 in this embodiment (110A), and the step 125 for minimum brightness detection and While omitting 126, the maximum brightness of R, G, and B is treated as the maximum brightness of one frame (130A), and the obtained maximum brightness is used as it is as the brightness change amount (130B). With this modification, the same effects as those of the first embodiment can be realized for R, G, B color images without losing the balance between the colors.

FIG. 10 shows the configuration of the data expansion section 30 for detecting the maximum brightness based on the image signal of NTSC or the like obtained by the camera 10 instead of the image data obtained by the A / D conversion. Show. That is, in the third embodiment, the image signal from the camera 10 is sampled by the sample & hold circuit 38, and the maximum brightness and the minimum brightness thereof are held. In this way, it is possible to directly obtain information on the degree of image flatness from the analog image signal.

The present invention, as shown in FIG.
It can also be realized by providing a data expansion unit 30A between the camera 10 and the image storage unit 12 in the prior art of No. 9. The data expansion unit 30A in the fourth embodiment is at the final stage of the data expansion unit 30 in the first embodiment and so on.
It has a configuration provided with a D / A converter for converting the image data after the brightness enhancement processing into an analog image signal, for example, an NTSC signal. As shown in FIG. 12, the data expansion unit 3
0A captures an image signal conforming to NTSC or the like from the camera 10 as image data in the same procedure as in FIG. 4 (1
00), the brightness of the acquired image data is expanded and output in the same procedure as in FIG. 6 (200A). However, as shown in FIG. 13, the image data after the brightness expansion is processed by the data expansion unit 3.
In the D / A converter inside 0A, the original format, for example NT
It is converted back to the SC signal format (215A). Therefore,
According to the present embodiment, the present invention can be implemented by a simple and inexpensive method of adding the data expansion unit 30A between the camera 10 and the image storage unit 12 in the conventional device.

(3) Fifth Embodiment FIGS. 14 and 15 show the configuration of an image compression / transmission / decompression system according to the fifth embodiment of the present invention, and FIGS. 16-18 show the operation thereof. In this embodiment, as shown in FIG. 14, a reference changing unit 42 is provided instead of the data expansion unit 30 in the first embodiment and the like. As shown in FIG. 15, the reference changing unit 42 includes the A / D converter 14, the frame memory 16, and the maximum / minimum luminance calculating unit 32 as in the first embodiment and the like. It has a section 44. As shown in FIG. 16, the reference value setting unit 44 executes a quantization coefficient calculation process (200B) instead of the brightness extension process (200) in FIG. FIG. 17 shows the flow of this quantized coefficient calculation process (200B). In this figure, the reference value setting unit 44 refers to the table 46 based on the information obtained by the maximum / minimum luminance calculation unit 32 (220), and quantizes the reference value obtained by this (see FIG. 7). It is set in the image compression unit 18 as a parameter for determining the quantization coefficient to be used in step 313) (230). Note that steps 100 and 300 in FIG. 16 are the same as those in the first embodiment and the like, and therefore description thereof is omitted.

As described above, in the present embodiment, the quantization coefficient is variably set according to the amount of change in luminance in one frame, whereby a comparatively small change in luminance can be stored even in the compressed image. This is because the quantized coefficient is step 3
This is possible because of the coefficient that determines the degree of quantization in 13 and thus the degree of compression. That is, the quantization in step 313 is executed by obtaining the product dq ⁻¹ of the gradation value d of the image data and the inverse matrix of the quantized coefficient q, and rounding off the fractional part (discarding by the rounding operation). Therefore, if the quantized coefficient q is increased, more parts are discarded (the degree of compression is increased), and conversely, if the quantized coefficient q is decreased, less parts are discarded (the degree of compression is decreased). Therefore, when the luminance change amount appearing in one frame is relatively small (when the image is relatively flat), the luminance change amount is changed from the reference value in advance so that the quantization coefficient q becomes relatively small. By setting the contents of the conversion table 46, the effect of saving the brightness change described above can be realized. The brightness expansion process in the first embodiment is, so to speak, equivalent to the process of increasing d in the product dq ⁻¹ when the brightness change amount is small, and the quantization variable setting process in this embodiment is, so to speak, a brightness change. It should be noted that when the amount is small, it corresponds to the process of reducing q in the product dq ⁻¹ , and therefore, when viewed in step 314, the effects of the two are very close.

Further, in this embodiment, as shown in FIG. 18, the same simultaneous parallel processing as in the first embodiment is realized. In addition to this, the quantization coefficient calculation process in the present embodiment requires a shorter time than the brightness extension process, and
Since the processing is performed once for each frame, the processing becomes faster and real-time than the first embodiment. Further, in the first embodiment and the like, the image on the monitor 28 becomes bright regardless of whether the original image is dark or bright unless the information about the luminance change is transmitted as shown in FIG. 3, but in the present embodiment, If the original image is dark, the monitor 2
The image on 8 is dark and bright if bright. In addition to this, in this embodiment, the same effect as the effect obtained in the first embodiment can be obtained. Also, with respect to this embodiment, the same modifications as those of the second embodiment and the third embodiment can be applied. However, since the present embodiment uses the quantized coefficient used in image compression, it is premised that the obtained image data is compressed as it is, and the modification like the fourth embodiment is similar to the first embodiment. Not so easy. Further, in the present embodiment, the first embodiment and the like are superior in that the table 46 for determining the reference value is required.

The contents of the table 46 can be empirically determined. Further, the quantized coefficient q is a matrix having the same number of components as the number of pixels belonging to one conversion unit, and if the ratio between the components of this matrix changes, the degree of compression changes for each spatial frequency component. Therefore, in order to prevent this, when executing step 313, the quantization coefficient q is multiplied by a scaling factor (scalar), and this scaling factor is stored in the table 46.

(4) Addendum In each of the above-described embodiments, one of the gradation value of each pixel of the image data and the quantized coefficient used for compression is corrected or set. For example, both can be modified or set at the same time.
In addition, in the first embodiment and the like, (gradation value-minimum luminance) x
The brightness extension processing is executed according to the formula of the number of gradations / brightness change amount, but this is not intended to limit the present invention to linear brightness extension, and if necessary, the present invention may be performed as nonlinear brightness extension. Can also be carried out. Further, in the second embodiment, if the information about the luminance change is transmitted to the receiving side device for each color, the receiving side device can eliminate the difference in the degree of compression between the colors. In addition, the fifth
In the embodiment, the quantization coefficient matrix may be set according to the brightness change amount without using the scaling factor. As a result, although the setting of the quantized coefficient q becomes slightly complicated, it becomes possible to change the ratio between the components of the quantized coefficient matrix according to the property of the image. In addition, the gradation number of the image data is not limited to 256, the format of the image signal is not limited to NTSC, and the format of the still image transmission is not limited to JPEG. The present invention can also be applied to correction of moving images such as MPEG.

[0040]

According to the first configuration of the present invention, the amount of change in luminance on one screen is detected, and based on the result and prior to compression, the gradation value of each pixel is reciprocal or inverse of the quantization coefficient. Since the gradation value and / or the quantization coefficient are modified or set so that the product obtained by multiplying the matrix becomes large, a relatively small luminance change in the image data before compression is retained even in the image data after compression. Or can be emphasized. In addition, the degree of flatness of the entire screen is detected by detecting the amount of change in the luminance of the entire screen instead of evaluating the score for each conversion unit and the total score of the entire screen. Is no longer required. Further, the process of writing the image data to be compressed in the frame memory and the detection of the degree of flatness can be executed simultaneously in parallel, so that the real-time property is improved. Furthermore, since it is not necessary to modify the conventional image compression procedure, image compression / commercial
An extension-dedicated IC becomes available, which enables easy and inexpensive implementation.

According to the second configuration of the present invention, the luminance change amount on one screen is detected, and each luminance value of the image signal or each of the image data is emphasized so that a relatively small luminance change on one screen is emphasized. Since the gradation value of the pixel is modified according to the detected brightness change amount, the same effect as the first configuration can be obtained. Furthermore, since all the prepared gradations can be used, the economical efficiency of gradation expression is improved, and a bright screen can always be obtained regardless of whether the original image is bright or dark. Also, regardless of the image compression method, or independently of the image compression,
Can be implemented.

According to the third configuration of the present invention, the image preprocessing method according to the second configuration is executed for the input image signal or the stored image data.
With respect to image compression / transmission, the same effect as the second configuration can be obtained. According to this configuration, since the gradation value is further corrected after storing the image data, the image based on the stored image data, that is, the image with the gradation value uncorrected is used via the monitor or the like. It is also possible to provide it to a person.

According to the fourth aspect of the present invention, the compressed image data is output from the transmission side apparatus including the image compression apparatus according to the third configuration to the communication line or the storage medium, and the reception side apparatus is Since the image data before compression is restored from the image data received via the communication line or the storage medium, the same effect as the third configuration can be obtained in the transmission side device. Furthermore, since the compressed image is transmitted or stored, the transmission band can be economically used.

According to the fifth configuration of the present invention, the transmission side apparatus in the fourth configuration outputs the information regarding the luminance change together with the compressed image data onto the communication line or the storage medium, and the reception side apparatus, Since the image data before the gradation value correction is restored based on the information about the luminance change received through the communication line or the storage medium, in addition to the effect obtained by the fourth configuration, The image data before correction can be restored.

According to the sixth configuration of the present invention, the image preprocessing method according to the second configuration is executed for the input image signal or the stored image data.
The same effect as the second configuration can be obtained. In addition, since the image data after gradation value correction is inversely converted into an image signal of the same format as the input image signal and then output, the means for implementing the method according to the present configuration can be implemented by a conventional device, for example. This configuration can be implemented by a method of inserting it between the camera and the image storage unit, and therefore the implementation is inexpensive and easy.

According to the seventh aspect of the present invention, the brightness change amount on one screen is detected, and a relatively small brightness change on one screen is performed on the compressed image data according to the detected brightness change amount. Since the quantization coefficient is set so as to remain, the effect similar to that of the first configuration is obtained, and in addition to the effect that the gradation value does not change, the brightness and the like in the original image also appear in the decompressed image. To be At that time, it is not necessary to transmit information regarding the change in brightness.

According to the eighth structure of the present invention, when the second or seventh structure is carried out, a step of inputting an image signal, converting the image signal into multi-tone image data and storing the image data, and a luminance change Since the step of detecting the amount is executed at least partially at the same time, the same effect as the second or seventh configuration can be obtained, and in addition, image transmission with high real-time property can be realized.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an image compression / transmission / decompression system according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a data expansion unit in the present embodiment.

FIG. 3 is a block diagram showing a modified example of the image decompression unit in the present embodiment.

FIG. 4 is a flowchart showing an operation procedure of a transmission side device in the present embodiment.

FIG. 5 is a flowchart showing an image capturing / luminance detecting procedure of the operation procedure of the transmitting side apparatus in the present embodiment.

FIG. 6 is a flowchart showing a luminance extension procedure of the operation procedure of the transmission side apparatus in the present embodiment.

FIG. 7 is a flowchart showing an image compression procedure of the operation procedure of the transmission side apparatus in the present embodiment.

FIG. 8 is a timing chart showing the operation timing of the transmission side apparatus in the present embodiment.

FIG. 9 is a flowchart showing an image capturing / brightness detecting procedure of the operation procedure of the transmitting side apparatus according to the second embodiment of the present invention.

FIG. 10 is a block diagram showing a configuration of a data expansion unit according to the third embodiment of the present invention.

FIG. 11 is a block diagram showing a configuration of a data expansion unit according to the fourth embodiment of the present invention.

FIG. 12 is a flowchart showing an operation procedure of the transmission side device in the present embodiment.

FIG. 13 is a flowchart showing a luminance extension procedure of the operation procedure of the transmission side apparatus in the present embodiment.

FIG. 14 is a block diagram showing a configuration of an image compression / transmission / decompression system according to a fifth embodiment of the present invention.

FIG. 15 is a block diagram showing a configuration of a data expansion unit in the present embodiment.

FIG. 16 is a flowchart showing an operation procedure of the transmission side device in the present embodiment.

FIG. 17 is a flowchart showing a quantization coefficient setting procedure of the operation procedure of the transmitting side apparatus in the present embodiment.

FIG. 18 is a timing chart showing the operation timing of the transmission side apparatus in the present embodiment.

FIG. 19 is a block diagram showing a configuration of an image compression / transmission / decompression system according to a conventional technique.

FIG. 20 is a block diagram showing a configuration of an image storage unit in the conventional technique.

[Explanation of symbols]

10 camera, 14 A / D converter, 16, 26 frame memory, 18 image compression unit, 20 communication line, 22
Storage medium, 24, 24A image expansion unit, 24B data expansion cancellation unit, 28 monitor, 30, 30A data expansion unit, 32 maximum / minimum brightness calculation unit, 34 subtractor, 3
6 multiplier / divider, 38 sample & hold circuit, 42
Reference changing unit, 44 Reference value setting unit, 46 table.

Claims (8)

[Claims] 1. A means for detecting a luminance change amount on one screen from an image signal or multi-tone image data corresponding thereto, and a tone value of each pixel is multiplied by an inverse number or an inverse matrix of a quantization coefficient. Means for quantizing and compressing the image data by rounding the fractional part of the product, and the detected luminance change of the gradation value and / or the quantization coefficient prior to the compression and so that the product becomes large. An image compression apparatus comprising: means for correcting or setting according to the amount; and by the correction or setting, a relatively small luminance change in the image data before compression is held or emphasized in the image data after compression. . 2. A step of detecting a brightness change amount on one screen from an image signal or multi-tone image data corresponding to the image signal, and the image so that a relatively small brightness change on one screen is emphasized. An image preprocessing method comprising: a step of correcting a luminance value of a signal or a gradation value of each pixel of image data according to a detected luminance change amount. 3. A method for inputting an image signal, converting the image signal into multi-tone image data and storing the converted image data, and an image preprocessing method according to claim 2, which is applied to the input image signal or the stored image data. By executing the correction, the gradation value of each pixel is corrected so that a relatively small change in luminance on one screen is emphasized, and then the means for outputting the stored image data and the gradation value of each pixel are set. An image compression apparatus comprising: a unit configured to compress the corrected image data. 4. The image compression apparatus according to claim 3, and a transmission side apparatus provided with means for outputting the compressed image data onto a communication line or a storage medium, and image data received via the communication line or the storage medium. An image transmission system comprising: a receiving side device having a unit for restoring image data before compression. 5. The image transmission system according to claim 4, wherein the transmission side device is provided with means for outputting information regarding a luminance change together with the compressed image data onto a communication line or a storage medium, and the reception side device. An image transmission system comprising means for restoring the image data before gradation value correction based on the information on the brightness change received via a communication line or a storage medium. 6. The method according to claim 2, wherein a step of inputting an image signal having a predetermined format, converting the image signal into multi-gradation image data and storing the image data, and the input image signal or the stored image data are targets. By executing the image pre-processing method, the step of correcting the gradation value of each pixel of the image data and the gradation value of each pixel are corrected so that a relatively small change in brightness on one screen is emphasized. An image preprocessing method, comprising the step of inversely converting the subsequent image data into an image signal of the same format as the input image signal and outputting the image signal. 7. A step of detecting a brightness change amount on one screen from an image signal or multi-gradation image data corresponding thereto, and a relatively small brightness change on one screen on the compressed image data. And a step of setting a quantization coefficient that determines a threshold value of the brightness change amount that is considered to be no brightness change during the image compression processing so as to remain according to the detected brightness change amount. Image preprocessing method. 8. The image preprocessing method according to claim 2 or 7, further comprising a step of inputting an image signal, converting the image signal into multi-tone image data, and storing the image data. An image pre-processing method, characterized in that the detection of the amount of change in luminance is executed at least partially at the same time.