JPH03163957A - Picture processing unit - Google Patents

Abstract

PURPOSE:To binarize a gradation picture data without being affected with a luminance level distribution of the gradation picture data by detecting the luminance distribution of an inputted video data by one pattern, normalizing the inputted and outputted video data based on the luminance distribution and outputting it after binarizing the data. CONSTITUTION:When reading and transmission of a picture data are commanded, a reproduced picture data by one pattern is written in a memory circuit 109. Thus, the picture data by one pattern is stored in a RAM of a control circuit 115 and the distribution state of the luminance level of a luminance signal is analyzed. Thus, when Lmax, Lmin are decided, number of picture elements of the sender side and number of picture elements of the receiver side are matched. When an interpolation picture element level is decided, the Lmax, Lmin are used and the normalizing calculation of the interpolation picture element is implemented to obtain the transmission picture element and the data is normalized between the density level '0' and '255' and the picture data binarized by a coding circuit 120 is outputted to a line 122 via a modulation circuit 121.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image processing device that binarizes and outputs video data such as a still image television signal such as a still video.

[Prior Art] In recent years, still image video signals, which are played back from a floppy disk that stores video data, for one screen are temporarily stored in a memory, and then sequentially read out and transmitted in accordance with the transmission speed. Transmission devices have been proposed and developed. Among such transmission devices, those that perform digital transmission have become mainstream due to the recent spread of personal computer communications and modem devices becoming faster and cheaper. Among these transmission devices, those capable of transmitting color images are
It is common to transmit color information using GB or Y and color difference signals (R-Y) and (B-Y), but if the communication protocol is combined, it can also be used with the currently widely used binary facsimile machine. Transmission possible.

In this case, it is necessary to convert the television signal, which is usually expressed in 256 gradations, into a black and white binary signal. The simplest method of this binary conversion method is to convert the luminance signal Y into two values.
There is a binarization method that compares the image with the center level (127) of 56 gradations and determines whether it is black or white based on the comparison result. Another common method is to perform dither processing by associating, for example, a 4×4 dither pattern with each sampling pixel level.

[Problems to be Solved by the Invention] When gradation image data recorded on a video floppy is binarized using the method described above, when the gradation image data is captured under appropriate exposure conditions, There is no problem, but if there is underexposure or overexposure during imaging, the luminance distribution of the imaging signal will be biased on the level axis.

FIG. 2 is a diagram showing such an example, in which the horizontal axis represents the imaging brightness level, and the vertical axis represents the frequency distribution of each imaging brightness level. FIG. 2A shows the frequency distribution under proper exposure conditions, and in this case the frequency distribution is uniform. FIG. 2B shows a case of underexposure, in which the imaging brightness level is biased toward a low portion. FIG. 2C shows the case of single-bar exposure, and the imaging brightness level is biased towards the high part.

Therefore, in the conventional method of binarizing television signals, the binarized image data may become dark or overexposed due to variations in the exposure conditions during image capture, resulting in the optimal There was a problem that a valued image could not be obtained.

The present invention has been made in view of the above-mentioned conventional example, and by changing the binarization method based on the luminance level distribution of gradation image data, it is possible to generate binary data without being influenced by the luminance level distribution of gradation image data. The purpose is to provide an image processing device that can

[Means for Solving the Problems] In order to achieve the above object, an image processing apparatus of the present invention has the following configuration. That is, it is an image processing device that binarizes and outputs multivalued video data, and includes a detection means for detecting the brightness distribution of input video data for one screen, and a detection means based on the brightness distribution detected by the detection means. The apparatus further includes normalizing means for normalizing the video data, and encoding means for binarizing the normalized video data.

[Operation] In the above configuration, the brightness distribution of the input video data for one screen is detected, and the input video data is normalized based on the brightness distribution. The normalized video data is then binarized by the encoding means and output.

[Embodiments] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[Description of Image Transmission Apparatus (FIG. 1)] FIG. 1 is a block diagram showing a schematic configuration of an image transmission apparatus according to an embodiment.

100 is a video floppy disk storing still video signals, etc.; 102 is this video floppy disk 10;
This is a motor that rotates the 0.

Reference numeral 103 denotes a reproducing magnetic head for reading out video signals recorded on the video floppy 100. The video signal thus read out is amplified by the reproduction amplifier 104 and input to the reproduction process circuit 105 . Regeneration process circuit 105
Then, the baseband level video signal is restored, and the luminance signal Y and line-sequential color difference signal (R-Y)/(B-Y)
is output to the NTSC encoder 106, while the luminance signal YllO and color difference signal 1l1 converted into digital signals are output to the NTSC encoder 106.
is output to the memory circuit 109.

The NTSC encoder 106 creates an NTSC signal based on the input luminance signal Y and color difference signal, and outputs it to the monitor circuit 107.
The image is displayed on the CRT device 108 through the image. As a result, the reproduced image signal reproduced by the reproduction magnetic head 103 is transmitted to the CRT.
This can be confirmed using the device 10B.

112 is a memory control circuit which receives the horizontal synchronization signal H and vertical synchronization signal V from the reproduction process circuit 105 as input signals, and outputs the address data 1l3 and the write signal 114.
It outputs and controls data writing to the memory circuit 109. 115 is a control circuit that controls the entire device; a CPU that outputs various control signals and controls the entire device;
It also includes a ROM that stores the CPU control program and various data shown in the flowchart of FIG. 3, and a RAM that is used as a work area for the CPU.

The control circuit 115 outputs a freeze start signal 116 to the memory control circuit 112 when instructed to transmit the image data currently displayed on the CRT 108 from an instruction unit such as a keyboard (not shown). As a result, the memory control circuit 112 is activated, and the horizontal synchronization signal H
and address data 113 in synchronization with the vertical synchronization signal V.
and outputs the write signal 114 to write 1 to the memory circuit 109.
Writes digital image data such as brightness signals for the screen. On the other hand, when reading from the memory circuit 109, the control circuit 115 outputs a read signal 118 and an address signal 113,
Image data 117 is read from memory circuit 109.

The read image data is normalized by performing a calculation described later in the control circuit 115, and the image data 11
It is output as 9 to the encoding circuit 120 at the subsequent stage. The encoding circuit 120 binarizes the input image data 119 using a dither matrix, converts it into serial data, and outputs it to the modulation circuit 12l. The modulation circuit 121 modulates manually input serial data into a 9-signal format that can be transmitted via a telephone line, for example, and outputs it to the line 122.

[Description of transmission data creation process (FIGS. 1 and 3)] FIG. 3 is a flowchart showing the transmission data creation process in the control circuit 115 of the embodiment. Stored in ROM.

As described above in step S1, when reading and transmitting image data is instructed from a keyboard (not shown) or the like, the process proceeds to step S2, where a freeze start signal 116 is output to the memory control circuit 112. As a result, one screen worth of reproduced image data is written into the memory circuit 109.

In step S3, the address signal 113 and the read signal 118 are
and reads one screen worth of image data from the memory circuit 109. At this time, if it is known that the destination of the transmission is a binary facsimile machine, 1 ○ Only the luminance signal Y may be written into the memory circuit 109, or only the luminance signal Y may be read out. .

In this way, one screen worth of image data is transferred to the RA of the control circuit 1l5.
The distribution state of the brightness level of the brightness signal is analyzed. In this way, as shown in FIG. 2, the minimum value L m l n and the maximum value L mau of the brightness level are determined. In this case, there is a method of checking the entire screen and all pixel levels, but it is also possible to select several representative pixels and determine the maximum and minimum values from the selected pixels.

In addition, from among all the pixels in the pixel data, remove a number of pixels equivalent to 5% of the total number of pixels in order from the pixel with the lowest brightness level, and then remove the number of pixels equivalent to 5% of the total number of pixels in order from the pixel with the highest brightness level among all the pixels. If you remove a number of pixels equivalent to 5% of the image data and then determine the maximum and minimum values of the brightness level of the image data, it will be possible to eliminate pixels with extremely high or extremely low brightness levels due to noise etc. The impact will be less.

Thus L. . , Lmln are determined, step S4
Proceed to step 2 to match the number of pixels on the transmitting side and the number of pixels on the receiving side.
conduct. This is an expansion that satisfies the relationship 4XkXNT≦NR, where NR is the number of dots (pixels) per line on the receiving side facsimile, NT is the number of pixels at the sampling level on the sending side, and k is the expansion rate. Determine the rate. Here, ”4
” is processed by a 4×4 dither matrix, so
This is used as a multiplier because the number of pixels in one line after encoding is multiplied by 4 in the main scanning direction. For example, if the image size on the transmitting side is 480 pixels (horizontal) x 640 lines 1 9 (vertical) as shown in FIG. 4, then NT is "4 80".

Specific numerical examples of each of these counts are shown in the table below.

Once the enlargement ratio is determined in this way, the vertical address of the memory circuit 1O9 is set to "O" (step S5), and the horizontal address is set to "0°" (step S6).

Next, in step S7, interpolated pixel levels 4(i) and r(j) are calculated according to the enlargement ratio k. here,
12 (i) indicates pixel data in the main scanning direction, r (
j) indicates pixel data in the sub-scanning direction. However, i
．． j is a numerical value indicating the pixel position and is 1 3 .

[Explanation of pixel interpolation (FIGS. 5 and 8)] FIG. 5 is a diagram showing interpolation in the main scanning direction, taking as an example the case of k=5/4 in the above table.

1 (0), I2 (1), . . . indicate pixels in the main scanning direction, and A' (0), I2' (2) indicate pixel levels with the number of pixels expanded 5/4 times. Here, we assume that each pixel is interpolated by one-dimensional linear interpolation, and β'
(0) = flood (0)! ' (1) = {β(0) +4 xA (1
)} /5β' (2) = {2 xl (1)
+3 X℃(2))/5β'(3) = {3
xI2(2) +2 Xβ(3))/5 flood' (4)
= {4 xn (3) +12(4) ) /5, and the same cycle is repeated after 1'(5).

1 4 Instead of one-dimensional linear interpolation, interpolation using high-dimensional interpolation, convolution that realizes an ideal LPF (low-pass filter), etc. may be used.

FIG. 8 is a diagram showing how the interpolated data in the main scanning direction shown in FIG. 5 is further interpolated in the sub-scanning direction.

r(0) to r(3) indicate pixels in the sub-scanning direction, and r'
(0) to r' (3) indicate pixel levels whose number of pixels has been expanded 5/4 times. As in the case of the main scanning direction, the conversion is performed as shown in the equation below, and this cycle is repeated thereafter.

r' (0) = r (0) r' (1) = {r (0) +4 Xr (
1) ) /5r' (2) = {2 Xr (1
) +3 Xr (2) ) /5r' (3) =
{3 Xr (2) +2 Xr (3) } /5
r' (4) = (4 Xr (3) +r (4
) ) /5? Ultimately, these r' (j) become the transmission pixel level, but 42' (i) and r' (j
) is uniquely determined by the position of the transmitted pixel in the line and dot directions in the main scanning and sub-scanning directions, and therefore is performed by selecting an appropriate calculation algorithm.

When the interpolation pixel level is determined in step S7, the process proceeds to step S8, where L m m obtained in step S3
A normalization operation is performed on these interpolated pixels using M + L m l n to obtain transmission pixels.

This is the interpolation pixel level (n'(t)r' (
j) ) is L7 and the transmission pixel is LT', then LT'
■ ((LT - Lffl+7)/(L---
The transmission pixel level LT′ is normalized by −L−+, ))×2 5 5, and L mln
Pixel data distributed between and L max is normalized to a density level between ゜゜O゜'' and ゜'255''.

l 5 Once the transmission pixel L7' is determined in this way, the process proceeds to step S9, and LT' is outputted as data 119 to the encoding circuit 120. The image data binarized by this encoding circuit 120 is output to a line 122 via a modulation circuit 121.

When the processing of a predetermined amount of pixel data in the main scanning direction is completed as described above, the horizontal address is increased by +1 in step S10.
Then, in step Sll, it is checked whether processing of one line of pixel data has been completed. If the processing for one line has not been completed, the process returns to step S7 and the above-mentioned processing is executed. When the processing of the pixel data of the l line is completed, the process proceeds to step S12, the vertical address is increased by 1, and the process proceeds to step S13. If the processing for the pixel data of all lines has not been completed in step S13, the process returns to step S6, and normalization processing of the pixel data from the beginning of the next main scanning line is performed, and if necessary, the somale processing etc. do.

[Explanation of Binarization Process (Figures 6 and 7)] Figure 6 shows a 4x4 dither matrix used for binarization in the encoding circuit 120, and each number in this matrix is , when each pixel level is higher than the brightness level 16 times that number, the pixel part corresponding to that position is black ("1")
It is binarized so that

FIG. 7 is a diagram schematically showing a binarization algorithm using dither in an encoding circuit, and black pixel generation patterns in the dither pattern of FIG. 6 are shown corresponding to each luminance level range of pixels. There is.

In this way, in the encoding circuit 120, for example, G I
When performing run length according to the II standard, the first line of pixel data of the data 119 is dithered into 1 8 codes, and when transmitted, it is sequentially run length encoded in the horizontal direction starting from the top row of dots. Next stage modulation circuit 1
Output to 21. This is performed for four lines before proceeding to pixel data processing for the next line, so it is sufficient to perform dither encoding for the next line while run-length encoding for four lines is being performed.

In the above embodiment, the binarization in the encoding circuit is performed by dither encoding, but the invention is not limited to this, and the binarization may be performed using the simplest median value as the threshold. good. Furthermore, although the dither pattern is 4×4, it is also possible to use a dither pattern of 3×3 or other numbers.

As explained above, according to this embodiment, after normalizing the luminance level according to the distribution of luminance components of the reproduced video signal, the video signal is encoded and converted into 21 9 values. This has the effect of further improving the reproducibility of the binarized luminance signal.

[Effects of the Invention] As described above, according to the present invention, there is an effect that image data can be binarized without being affected by the luminance level distribution of the gradation image data.

[Brief explanation of the drawing]

Figure 1 is a block diagram showing the schematic configuration of the image transmission device of this embodiment. Figure 2 is a diagram showing an example of the distribution of brightness levels in video data for one screen. Figure 3 is creation of transmission data in the control circuit. A flowchart showing the processing, FIG. 4 is a diagram showing an example of the configuration of image data used in this embodiment, FIG. 5 is a diagram showing horizontal pixel interpolation, and FIG. 6 is an example of a dither matrix. FIG. 7 is a diagram showing the relationship between density level and dither pattern, and FIG. 8 is a diagram showing pixel interpolation in the column direction. In the figure, 100...floppy disk, 104...
Amplifier, 105... Regeneration process circuit, 106...
NTSC encoder, 107...monitor circuit, 108
...CRT, 109...Memory circuit, 110...
Luminance signal, 111... Color difference signal, 112... Memory control circuit, 113... Address signal, 115
... Control circuit, 116 ... Freeze start signal, 120 ... Encoding circuit, 121 ... Modulation circuit, 1
22... It is a line.

Claims (1)

[Scope of Claims] An image processing device that binarizes and outputs multivalued video data, comprising: a detection means for detecting a luminance distribution of input video data for one screen; and a luminance detected by the detection means. An image processing device comprising: normalizing means for normalizing the video data based on distribution; and encoding means for binarizing the normalized video data.