JP2738685B2 - Scan line density conversion method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a scanning line density conversion method for performing conversion such as enlargement and reduction of an image. 2. Description of the Related Art Conventionally, the following two methods have been introduced as methods of scanning line density conversion. (Literature: Arai, Yasuda, "A Study of Facsimile Linear Density Conversion," The Institute of Image Electronics Engineers of Japan, Vol. 7, No. 1, pp. 11-18 (1978
Year)) Method 1: Enlargement / reduction by repeating or thinning out input data Method 2: Interpolation of input data or enlargement / reduction by average. Figure 7 (A), (B) using the above two will be described scaling method, data d ₁₁ of the solid line circle in Fig _{(A), d 13, d} 15, d 31, d 33, Method 1 with d ₃₅ as input data
If both the main and sub-scans are enlarged by a factor of two, then d ₁₁ = d ₁₂ = d ₂₁ = d ₂₂ d ₁₃ = d ₁₄ = d ₂₃ = d ₂₄ and so on. On the other hand, in the method _{2, d 12 = (d 11} + d 13) / 2, d 14 = (d 13 + d 15) / 2, d 32 = (d 31 + d 32) / 2, d 34 = (d 33 + d 35 ) / 2, d ₂₁ = (d ₁₁ + d ₃₁ ) / 2, d ₂₂ = (d ₁₂ + d ₃₂ ) / 2. In the case of a three-fold enlargement in the same manner in FIG. 7B, in the method 1, d ₁₁ = d ₁₂ = d ₁₃ = d ₂₁ = d ₂₂ = d ₂₃ = d ₃₁
= D ₃₂ = d ₃₃ ,... In the method 2, for example, to calculate d ₂₂ by interpolation, each input data of d ₁₁ , d ₁₄ , d ₄₁ , and d ₄₃ is converted to d ₂₂
Therefore, the calculation is complicated by adding the weights in inverse proportion to the spatial distance from In the case of reduction, when all data is input data and solid circles are reduced data in FIGS. 7A and 7B, the dotted circle data is thinned out in the method 1, and in FIG. In (A), d ₁₁ = (d ₁₁ = d ₁₂ + d ₂₁ + d ₂₂ ) / 4 In FIG. (B), d ₁₁ = (d ₁₁ + d ₁₂ + d ₁₃ + d ₂₁ + d ₂₂ + d ₂₃ + d ₃₁ + d ₃₂ + d
₃₃ ) The average is calculated as in / 9. Problems to be Solved by the Invention However, method 1 has a drawback in that, although the arbitrary magnification conversion can be implemented by an easy algorithm, the speed of the hardware operation circuit can be increased, but the distortion of the converted data is large. On the other hand, in method 2, the distortion of the converted data is small, and 2 ⁿ times or 2 ⁻ⁿ times (n
Is a positive integer), it is sufficient to repeat 2 times or 1/2 times for the magnification conversion, and the operation algorithm is simple. However, for other magnifications, the operation becomes complicated and the operation circuit is compared with the method 1. It is difficult to increase the speed. SUMMARY OF THE INVENTION The present invention solves the above-mentioned problems, and provides a scanning line density conversion method capable of changing a scanning line density at an arbitrary magnification with a small distortion of converted data and capable of high-speed processing. Means for Solving the Problems The present invention provides a method for converting an input image signal having a scanning line density of 1 lines / mm into an image signal having a scanning line density of m lines / mm by using a designated magnification M
(However, an interpolation magnification S (where S ≦ m / l) of 2 ⁿ (n is a positive integer) corresponding to m / l) is set, and the l / mm input image signal is subjected to the interpolation magnification. Interpolated by S times averaging operation and S
After converting the image signal to l lines / mm, determine whether to newly input and output the image signal or repeatedly output the previous image signal based on DDA (Digital Differential Analysis) and output. The S / l lines / mm image signal interpolated at the interpolation magnification S times is enlarged to M / S times by the
The above object is achieved by converting the image signal into a book / mm image signal. According to the present invention, in the above-described method, in the scanning line density conversion at an arbitrary magnification, an input image signal can be rapidly converted to scanning line density data with small distortion by a simple algorithm corresponding to a specified magnification. At the time of scanning density conversion, the interpolation magnification S (S≤m / l) is used to temporarily perform complementary conversion to S times, and then to convert the enlarged data to the scanning line density specified by a simple algorithm, thereby setting the designated magnification setting area. However, it is possible to perform high-speed scanning line density conversion with little distortion. Embodiment First, the reason why the distortion is small is shown in FIGS.
For simplicity, a description will be given using a one-dimensional signal using (D). FIGS. 7A to 7D are examples of triple magnification. Analog image signal f in FIG (A) is the quantized in terms of S ₀ to S _4, the quantized signals of the shaded area is obtained. The signal obtained by converting the quantized signal into a triple scanning line density by the method 1 has a larger distortion (error) than a signal (dotted line) quantized normally at a triple scanning line density. FIG. 3B shows a case where the scanning line density is converted to three times the scanning line density by the method 2. The distortion is small, but the conversion is not a power-of-two conversion, so that the operation circuit becomes complicated (especially the operation circuit in two-dimensional processing). As described above, the processing takes time. FIGS. 3C to 3D show an embodiment of the present invention, and FIG. 3C shows a signal obtained by performing data interpolation twice in the above-described method. FIG. 3D is a diagram in which the signal of FIG. 3C is enlarged by 3/2 times by the above-mentioned method 1 and the signal is rewritten in an equal gap. Compared to the signal (dotted line) quantized at three times the scanning line density in FIG.
The distortion of the converted signal (shaded area) according to the present invention is small.
It is clear in comparison with FIG. Furthermore, a brief description will be given of the fact that conversion processing with small distortion can be performed even when the variable region of the designated magnification is wide. When the variable area of the designated magnification is widened from 1.0 to 8.0 times, to satisfy all the designated magnifications in this area, interpolation is performed by 1.0 times and then enlarged to the designated magnification to perform the conversion for a while. However, after the 1.0-fold (2 ⁰ times) interpolation, when expanded by the method 1 to 8 times, the distortion of the converted signal increases. This is apparent from the fact that the method 1 is equivalent to the method of expanding the data up to three times after the one-time interpolation. The distortion of the converted signal depends on the number of continuous repetitive processes at the time of interpolation. For example, in the case of the triple conversion according to the method 1, the number of consecutive repetitive processes is two (FIG. 6 (A)). On the other hand, in the example of FIG. This difference leads to the magnitude of the distortion. Therefore, if the interpolation magnification is set as follows according to the specified magnification so that the number of repetitions is reduced, a converted signal with little distortion can be obtained. When 1.0 times ≤ M <2.0 times, the interpolation magnification is 1.0 times (maximum continuous repetition times: 1) When 2.0 times ≤ M <4.0 times, the interpolation magnification is 2.0 times (maximum continuous repetitions: 1 time) 4.0 times ≤ M <8.0 When the interpolation factor is 4.0, the interpolation magnification is 4.0 (maximum continuous repetition frequency is 1). When M = 8.0, the interpolation magnification is 8.0 (maximum continuous repetition frequency is 1). M is the designated magnification. Hereinafter, a specific example of the present invention will be described. FIG. 1 is a block diagram of an apparatus for realizing a scanning line density conversion method by enlargement processing according to an embodiment of the present invention. In FIG. 1, 1 is an input terminal for a designated magnification in the main scanning direction, 2 is an input terminal for a designated magnification in the sub-scanning direction, 3 is an interpolation magnification setting section in the main scanning direction, and 4 is an interpolation magnification setting section in the sub-scanning direction. 5 is an input terminal of a quantized input image signal, 6 is a sub-scan data interpolator, 7 is a main scan data interpolator, 8 is a sub-scan data repeater, 9 is a main scan data repeater, and 10 is an output image signal. Reference numeral 11 denotes a timing signal generator. Now, the scanning line density of the input image signal is set to l
/ mm, the scanning line density of the output image signal is
mm, and when the main and sub-scans are tripled,
The operation will be described below. The designated magnifications of the input terminals 1 and 2 are input to the main scanning interpolation magnification setting unit 3 and the sub-scanning interpolation magnification setting unit 4, respectively, and the interpolation magnification is set. In this embodiment, the designated magnification M
The relationship between and the interpolation magnification S is as follows. When 2 ⁿ ≦ M <2 ^{n + 1} , S = 2 ⁿ (where n is a positive integer given by n = INT (log ₂ M)) Since the specified magnification M is 3 times, the interpolation magnification is Is doubled. Further, the interpolation magnification setting units 3 and 4 output the repetition coefficients (3/2 in this example) to the repetition units 8 and 9, respectively. The input image signal from the input terminal 5 is input to the sub-scanning data interpolation unit 6, and the data in the sub-scanning direction is increased by 2 times by the interpolation magnification from the sub-scanning interpolation magnification setting unit 4, and is input to the main scanning data interpolation unit 7. I do. In the main scanning data interpolation section 7, the data in the main scanning direction is further increased by a factor of 2 by the interpolation magnification from the main scanning interpolation magnification setting section 3. Thereafter, the sub-scanning repetition unit 8
The data in the sub-scanning direction is enlarged to 3/2 by the repetition coefficient from the width-scanning interpolation magnification setting unit 4 and the main-scanning repetition unit 9
Then, the data in the main scanning direction is enlarged to 3/2 by the repetition coefficient from the main scanning interpolation magnification setting unit 3 to become an output image signal of the output terminal 10. Each of the sections 6 to 9 is controlled by each timing signal from the timing signal generation section 11.
Hereinafter, more detailed examples of the respective units 6 to 9 will be described. Since the main scanning direction and the sub-scanning direction have basically the same configuration, only the main scanning direction will be described for the sake of simplicity. In addition, it is considered that the input / output lines of each block need only be able to understand the signal flow, and are represented by one line for simplicity. The variable area of the designated magnification M is 1 for simplicity of explanation.
≦ M <8. FIG. 2 is a detailed configuration diagram of the main scanning data interpolation unit 7 in FIG. 1, and shows a case where the supply of input image data can be sufficiently fast. FIG. 3 shows the timing. In Figure 2, 20 denotes a signal line of the input image signal sequence D _0, 2
1,22,26 latch circuit, the averaging circuit 23 to 25, the selector circuits 27, 29 signal lines of the output image signal sequence D _1, 30 to 33
Are signal lines for timing signals P ₁ , P ₂ , P ₄ and P ₀ , respectively. On the other hand, (1) to (3) of FIG. 3 respectively show the input image signal sequence D ₀ when the specified magnification M is 4 ≦ M <8, 2 ≦ M <4, 1 ≦ M <2, and (4) P ₁ signal, (5) P ₂ signal, (6) P ₄ signals, (7) - (12), respectively designated magnification M has 4 ≦ M <
8, the output of the latch 21 and the output of the latch 22 when 2 ≦ M <4, 1 ≦ M <2, and (13) to (15) indicate that the designated magnification M is 4
≦ M <represents the timing of the output image signal sequence D ₁ of the case of 8,2 ≦ M <4,1 ≦ M < 2. The operation of the above configuration will be described below. The selector 27 outputs the timing signal P ₁ of the signal lines 30 to 32,
P ₄ , 2 when 1 ≦ M <2 according to designated magnification M among P ₂ and P ₄
A selector circuit that selects P ₂ when ≦ M <4 and P ₁ when 4 ≦ M <8 and outputs it as a signal line 34, and is supplied from the main scanning interpolation magnification setting unit 3 in FIG. It is switched in the interpolation magnification signal P _0. The input image signal sequence D ₀ ｛d ₁ , d ₂ ,
d _3, a ......} to the latch 21 is set in synchronization with the rising timing signal of the signal line 34 to output to the latch 22 of the latch 21 (of FIG. 3 (7) - (12)). At this time, the input timing of the input image signal sequence D ₀ may vary depending specified magnification (in FIG. 3 (1) to (3)) which is to handle the faster. The averaging circuits 23 to 25 are circuits for calculating and outputting the averaging of two input image data. The averaging circuit 23 outputs the output of the latch 21 and the output of the latch 22, and the averaging circuit 24 outputs the output of the latch 22. The output of the averaging circuit 23 and the averaging circuit 25 calculate the averaging of the output of the latch 21 and the output of the averaging circuit 23, respectively. now,
Assuming that the output of the latch 21 is d ₂ and the output of the latch 22 is d ₁ , the outputs of the averaging circuits 23 to 25 are as follows. Latch the outputs of the averaging circuits 23 to 25 and the output of the latch 22
26 is set in synchronization with the rise of the timing signal on the signal line 34. The selector 28 is connected to the interpolation magnification signal P ₀ and the signal line
By 32 the timing signal P ₄ of a selector circuit for switching the output of the latch 26, thereby, FIG. 3 (13) -
Output shown in (15) is obtained as the output image signal sequence D ₁ of the signal line 29. That is, when the designated magnification M is 1 ≦ M <2, only the output of the latch 22 is selected, and when 2 ≦ M <4, the output of the latch 22 and the output of the averaging circuit 23 are selected.
<When the 8 select all the outputs of the averaging circuits 23 to 25 of the latch 22 with the interpolation magnification signal P _0, in the period of the timing signal P ₄ each signal line 32, when 1 ≦ M <2 ｛D ₁ , d ₂ , d ₃ ……｝ When 2 ≦ M <4 ｛d ₁ , (d ₁ + d ₂ ) / 2, d ₂ , (d ₂ + d ₃ ) / 2, ...
...} {d ₁ when 4 ≦ M <8, (3d 1 + d 2) / 4, (d 1 + d 2) / 2, (d 1
+ 3d ₂ ) / 4, d ₂ ,. In the present embodiment, the designated magnification M is 1 ≦ M <8, and the number of stages of the averaging circuit is two (the first stage is the averaging circuit 23, and the second stage is the averaging circuits 24 and 25). If four averaging circuits are added to the third stage in the configuration of the third stage, it is possible to reduce the number to less than 16 times, and similarly to the configuration of the nth stage, it is possible to reduce the number to less than 2 ^{n + 1} times. By the above structure, a constant regardless of the period specified magnification of the output image signal sequence D _1, the smaller the interpolation magnification, to perform interpolation processing at a high speed since the period of the input image signal sequence D ₀ can short Can be. In the present embodiment, the interpolation magnification is set as follows with respect to the specified magnification M.
It is clear from FIG. When the specified magnification M is 1 ≦ M <2, the interpolation magnification is 1 × When the specified magnification M is 2 ≦ M <4, the interpolation magnification is 2 × When the specified magnification M is 4 ≦ M <8, the interpolation magnification is 4 FIG. 4 shows an operation flowchart of the main scanning repetition section 9 of FIG. The enlargement process in the repetition unit is the same as that of the conventional method 1, but will be described in detail with reference to the operation flowchart of FIG. In FIG. 4, R is an operation register, A and B are constant registers, and the enlargement ratio of the repetition part is represented by A / B (A ≧ B). The operation flowchart will be described below in order. 41 is 1
This is the starting point of the line image processing operation. Reference numeral 42 denotes an initial setting unit for clearing the register R to 0 and performing initial setting. Reference numeral 43 denotes a subtraction unit for subtracting the contents of the register B from the contents of the register R. The comparison unit 44 determines whether the result is negative or not. In the case of the present embodiment, since B> 0, the output of the subtraction unit 43 is always negative in the first processing. If the comparison unit 44 determines that the value is negative, the process goes to the comparison unit 45 to determine whether the processing of one line is completed. If the determination by the comparing section 45 is completed, the process goes to the image processing operation end point 49 for one line, and if not, the process goes to 46 to input data of one pixel. In the first process, the data of the first pixel is input here, and thereafter, the data of the second pixel and the third pixel are sequentially input each time the data passes therethrough.
After the data input, the process goes to the adder 47, where the content of the register A is added to the content of the register R, and then goes to 48 to output one pixel data. On the other hand, if the comparator 44 determines that the value is not negative, the process goes to 48 and outputs data of one pixel. That is, if it is determined that the input data is not negative, the input data is not updated by inputting one pixel data, so that the input data of the previous process is output as it is. 48
After outputting one pixel data in, the process returns to the subtraction unit 43. The processing from the subtraction unit 43 to the one-pixel data output 48 is 1
This processing unit is a pixel, and the same processing unit is repeated until the comparison unit 45 determines that the image processing operation of one line is completed. The above processing is performed by DDA (Digital Differential Analyzer: digita
l differential analyzer) to determine whether to newly input and output the image signal or to repeatedly output the image signal. The input data is updated only when the comparator 44 determines negative, Otherwise, 1
The data input in the process before the pixel is repeatedly output, and the enlargement process is performed by this repetition. FIG. 5 shows an example when the enlargement is repeated in the operation flow of FIG. FIG. 5A shows an example when the enlargement ratio is 3/2, and FIG. 5B shows an example when the enlargement ratio is 7/4. In FIG. 5, the comparison content of R indicates the value of the register R input to the comparison unit 44 in FIG. 4, and the input data sequence and the output data sequence are respectively the input data and the output data for each processing unit of one pixel. Is shown. As is clear from FIG. 5, when the comparison content of R is negative, the input data string becomes the output data string as it is, and when it is not negative, the same data as the output data one pixel before is repeatedly output to the output data string. The operation of the sub-scanning repetition unit 8 is basically the same as the operation of the main-scanning repetition unit 9 and will not be described. As described above, according to the present invention, when an input image signal having a scanning line density of 1 lines / mm is converted into an image signal having a scanning line density of m lines / mm, the designated magnification M (where m / l) is obtained. Then, an interpolation magnification S (where S ≦ m / l) of 2 ⁿ (n is a positive integer) is set, and the l / mm input image signal is interpolated by the averaging operation of the interpolation magnification S times. After converting to an image signal of S · l / mm, it is determined whether the image signal is newly input and output or the previous image signal is repeatedly output based on DDA (Digital Differential Analysis). And, by the output enlargement process,
The image signal of S · l lines / mm interpolated at the interpolation magnification S times is M
By converting the image data into the m lines / mm image signal by multiplying the image data by a factor of / S, the distortion can be easily and quickly converted to a small value even at a wide range of magnifications, and the effect is large. Furthermore, by dividing the variable region of the designated magnification into a plurality of regions and setting the interpolation region S (S ≦ m / l) which can be expressed by 2 ⁿ (n is a positive integer) for each divided region. , High-speed processing can be realized, and the effect is great.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of an apparatus for realizing a scanning line density conversion method according to an embodiment of the present invention. FIGS. 2 and 3 are diagrams of a main scanning data interpolation unit in the apparatus. FIG. 4 is a block connection diagram and timing chart, and FIG. 4 is an operation flowchart of a main scanning repetition unit in the apparatus, and FIGS. 5 (A) and (B).
FIGS. 6A to 6D are diagrams each showing enlarged data by repetition in the same apparatus, FIGS. 6A to 6D are conceptual diagrams illustrating that the data conversion according to the present invention has little distortion using a one-dimensional signal, and FIGS. 2) and 2 (B) are conceptual diagrams showing conventional scanning line density conversion. 3 main scanning interpolation magnification setting unit 4 sub scanning interpolation magnification setting unit 6 sub scanning data interpolation unit 7 main scanning data interpolation unit 8 sub scanning repetition unit 9 main Scanning repeater.

Continuation of front page    (72) Inventor Hiroyoshi Tsuchiya               Matsu, 3-10-1 Higashimita, Tama-ku, Kawasaki-shi               Shimogiken Co., Ltd. (72) Inventor Yuji Maruyama               Matsu, 3-10-1 Higashimita, Tama-ku, Kawasaki-shi               Shimogiken Co., Ltd. (72) Inventor Toshiharu Kurosawa               Matsu, 3-10-1 Higashimita, Tama-ku, Kawasaki-shi               Shimogiken Co., Ltd.                (56) References JP-A-60-22871 (JP, A)

Claims (1)

(57) [Claims] An input image signal having a scanning line density of 1 lines / mm is converted to a scanning line density of m
When converting to an image signal of lines / mm, the specified magnification M (where m
/ l) and an interpolation magnification S of 2 ⁿ (n is a positive integer) (where
S ≦ m / l), and the l / mm input image signal is interpolated by the averaging operation with the interpolation magnification S times S · l / mm
After converting the image signal into a new image signal, the image signal is newly input and output based on DDA (Digital Differential Analysis), or it is determined whether to repeatedly output the previous image signal. The image signal of S · l lines / mm interpolated at the interpolation magnification S times is expanded to M / S times and the m lines / m
A scanning line density conversion method for converting into an m image signal.