JPH1141467A - Filter processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To leave gradations in a filter process result such as image data through a correcting process for putting the filter process result within a limited range. SOLUTION: The filter processor consists of a filter process means 1 which filters an input signal, a decision means 2 which decides whether or not the output of the filter process means 1 is within the limited range, and a limiting process means 3 which corrects the output of the filter process means 1. When the filter process result of the filter process means 1 is outside the limited range, the difference between the last filter process result corrected value and the upper limit value (lower limit value) of the limited range and the difference between the last filter process result correction value and a data maximum value (minimum) are found and the filter process result is corrected into the limited range according to the ratio of the two differences and outputted as a filter process result correction value. Even if the filter process result deviates continuously from the limited range, the filter process result correction value approximates the upper limit value (lower limit value) asymptotically, so the filter process result correction value can be put within the limited range while having gradations left.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a filter processing device used in an image processing device or a graphics processing device, and more particularly, to a method in which a gradation result can be left even if the filter processing result of image data is corrected within a limited range. And a filter processing device.

[0002]

2. Description of the Related Art In image processing apparatuses and graphics processing apparatuses, filter processing is performed for various purposes such as noise removal and contour enhancement. Since the data value changes when the original image data is filtered, the data value may fall outside a specified limit range. If the data is out of the restricted range, the data becomes meaningless. Therefore, the result of the filtering process is corrected by clipping or the like so as not to be out of the restricted range.

[0003] A filter circuit of a data processing device in which input / output is performed by a signal line having a fixed bit width and a range of usable data values is determined will be described below. FIG. 4 shows a conventional filter circuit. The filter processing means 1 is a circuit for performing a filtering process on input data, and includes a multiplication processing means 4 (4-1 to 4-7) and an addition means 7. The judging means 42 is an arithmetic circuit for judging the result of the filter processing, and is composed of an upper limit judging means 9, a lower limit judging means 10, an upper limit flag output means 13, and a lower limit flag output means 14. The restriction processing means 43 is a selection circuit for keeping the filter processing result within the restriction range, and includes a selector 45 and an output data storage unit 46.

The multiplication processing means 4 (4-1 to 4-7) is an operation circuit for one data in the filter processing, and includes a data storage means 5 (5-1 to 5-7) and a coefficient storage means 6 ( 6-1 to 6
-7) and a multiplication circuit. Data storage means 5
(5-1 to 5-7) are the data X1 to X in the filtering process.
7 is a storage circuit for storing the data. Coefficient storage means 6 (6-1 to 6
-7) is a storage circuit for storing coefficients A1 to A7 for multiplying data in the filter processing. Addition means 7
Is an addition circuit for adding the result of multiplication processing of the input data by the multiplication processing means 4 (4-1 to 4-7).

[0005] The addition result storage means 8 is a storage circuit for storing the result added by the addition means 7. Upper limit judgment means 9
Is a circuit for determining whether the value of the data in the addition result storage means 8 exceeds the upper limit of the restriction range. The lower limit judging means 10 is a circuit for judging whether the value of the data in the addition result storage means 8 is below the lower limit of the limit range. The upper limit value storage unit 11 is a storage circuit that stores the upper limit value of the limit range. The lower limit value storage means 12 is a storage circuit for storing a lower limit value of the limit range. Upper limit flag output means
Reference numeral 13 denotes a circuit that outputs a flag signal indicating whether or not the filtering result exceeds the upper limit of the limit range. The lower limit flag output means 14 is a circuit that outputs a flag signal indicating whether or not the result of the filter processing is below the lower limit of the limit range.

The selector 45 selects and outputs one of the addition result storage means 8, the upper limit value storage means 11, and the lower limit value storage means 12 according to the output flags of the upper limit flag output means 13 and the lower limit flag output means 14. It is a selection circuit to be performed.

The operation of the filter processing device having the above-described configuration will be described with reference to the block diagram of FIG. 4 and the flowchart of FIG. First, a case where the filtering result exceeds the upper limit value of the restriction range will be described. In step 51, the filter coefficients A1 to A7 of the filter processing device are set in the coefficient storage means 6, the upper limit value Xmax of the limit range is set in the upper limit value storage means 11, and the lower limit value of the limit range is stored in the lower limit value storage means 12. Xmin is set.

Next, at step 52, the input data X1 to X6 are sequentially input from the outside to the multiplication processing means 4-2 to 4-7. Next, in step 53, X7 is input to the multiplication processing means 4-7. At the same time, the data X1 to X6 of the multiplication processing means 4-2 to 4-7 are transferred to the multiplication processing means 4-1.
.., And data X1 to X7 necessary for performing the filtering process. Therefore, in step 54, the filtering process is performed. That is, multiplication processing means 4-1 to 4-7 multiply input data X1 to X7 by filter coefficients A1 to A7, and the result is added by addition means 7.

Next, at step 55, the value of the filter processing result is assigned to a variable called Xtmp. That is, the output of the addition means 7 is stored in the addition result storage means 8. next,
In step 56, Xmax and Xtmp are compared. If Xtmp is larger than Xmax, the process proceeds to step 60 because the filtering result exceeds the limit range. In FIG. 4, the processing of the upper limit judging means 9 is performed, and Ra is set to 1 if Xtmp is larger than Xmax. In step 60, since the filtering result exceeds the upper limit of the limiting range, the upper limit value of the limiting range is stored in the output data storage unit 46 instead of the filtering result. In FIG. 4, the values of Xmax, Xtmp, and Xmin are input to the selector 45, and the output changes depending on the combination of Ra and Rb. Finally, in step 59, the data in the output data storage section 46 is output. In FIG. 4, Ra is set to 1,
Further, Rb calculated at the same time as Ra is set to 0, and Xmax is selected and output.

Second, a case where the result of the filtering process is below the lower limit of the limit range will be described. The processing of steps 51 to 55 in FIG. 5 is the same as in the first case. In step 56, Xmax is compared with Xtmp, and Xtmp is set to Xmax.
If it is smaller, the result of the filter processing is within the limit range, and the process proceeds to step 57. In FIG. 4, the calculation result Ra of the upper limit judgment means 9 is set to 0. In step 57, Xmin and Xtmp
Are compared, Xtmp is smaller than Xmin, and the filtering result is below the lower limit of the limit range.
Go to 61. In other words, (Xtm
Since p-Xmin) was negative, Rb of the lower limit flag output means 14 is set to 1. In step 61, since the filter processing result is lower than the lower limit of the restriction range, the lower limit value of the restriction range is stored in the output data storage unit 46 instead of the filter processing result. In FIG. 4, Xmax of the upper limit value storage unit 11, Xtmp of the addition result storage unit 8, and Xmin of the lower limit value storage unit 12 are input to the selector 45. Then, in a step 59, the data of the output data storage section 46 is output. That is, the selector 45 selects and outputs Xmin of the lower limit value storage means 12 in a combination of Ra = 0 and Rb = 1.

Third, a case where the filtering result is within the limit range will be described. The processing of steps 51 to 55 in FIG. 5 is the same as in the first case. In step 56, X
Max is compared with Xtmp. Since Xtmp is smaller than Xmax, the routine proceeds to step 57. In FIG. 4, the calculation result Ra of the upper limit judgment means 9 is set to 0. In step 57, Xmin and Xtmp are compared. If Xtmp is larger than Xmin, the result of the filtering process is within the limit range, and the process proceeds to step 58. That is, (Xtmp-Xmin) is calculated by the lower limit determining means 10, and Rb is set to 0 since the result is positive (non-negative). In step 58, since the result of the filtering is within the limited range, the result of the filtering is stored in the output data storage unit 46. In FIG. 4, Xmax of the upper limit value storage unit 11, Xtmp of the addition result storage unit 8, and Xmin of the lower limit value storage unit 12 are input to the selector 45. In step 59, the data in the output data storage unit 46 is output. That is, Ra is 0 and Rb is 0
, The selector 45 selects the X
Select and output tmp.

[0012]

However, in the above-mentioned conventional method, when the filtering result continuously exceeds the upper limit of the limiting range (below the lower limit), the corrected value of the filtering result becomes flat. There was a disadvantage. That is, when the correction value of the filter processing result is confirmed on a screen or the like, a screen that actually has a gradation becomes a uniform screen.

For example, consider data represented by an 8-bit signal line. The range of the data value represented by this signal line is 256 gradations with a minimum value of 0 to a maximum value of 255. It is assumed that this data value is limited to a range where the lower limit is 16 and the upper limit is 235. Such a prescribed signal is shown in Table 1 below.
Is input to a filter processing device having a filter coefficient of

[0014] Consider a case where the data of Table 2 below is input to this filter processing device.

[0015]

The result of filtering the data in Table 2 is shown in a line graph 65 in FIG. When the filter processing result exceeds the upper limit value of the restriction range (t3, t4), the filter processing result is corrected to the same value as the upper limit value. The result is shown in a line graph 66 in FIG. When the line graphs 65 and 66 are compared, there is a change in the graph 65 between the data numbers t3 and t4, but there is no change in the graph 66 and the graph is flat.

If the result of the filtering process continuously exceeds the upper limit as in the above example and the clipping process is performed to make the same value as the upper limit value, the image becomes flat and the gradation is lost. Although the filter coefficient is corrected in advance so that the filter processing result does not deviate from the limit range, it is not practical because the dynamic range of the filter processing result becomes too narrow. Further, a correction for compressing a data value in an area close to an upper limit value or a lower limit value is also performed. For example, a logarithmic compression operation or an arctangent operation is performed on the filter processing result so that the result does not deviate from the limit range. However, these operations are complicated and require a long operation time, and the scale of the operation circuit is large. Therefore, it is difficult to realize high-speed filtering at low cost.

The present invention solves the above-mentioned problem, and in a filter processing device used in an image processing device or the like, it is possible to easily keep a tone that should be originally present in a filter processing result and to keep the filtering processing result within a limited range. It is intended to be realized by a simple circuit.

[0019]

In order to achieve this object, a filter processing apparatus according to the present invention comprises a filter processing means for filtering an input signal, and a filter processing result which exceeds an upper limit value or is within a limit range. Determination means for determining whether the current value is below the lower limit value or not, and the difference between the immediately preceding filtering result corrected value and the upper limit value, and the difference between the immediately preceding filtering result corrected value and the data maximum value. , 2
Upper limit correcting means for correcting the filtering result within the limited range based on the ratio of the two differences, a difference between the immediately preceding filtering result corrected value and the lower limit value, and a difference between the immediately preceding filtering result corrected value and the data minimum value. Calculating the difference, selecting the lower limit correcting means for correcting the filtering result within the limited range based on the ratio of the two differences, and selecting the output of the upper correcting means when the filtering result exceeds the upper limit value; If the filter processing result is lower than the lower limit value, the output of the lower limit correction means is selected. If the filter processing result is within the limit range, the filter processing result is selected and set as the filter processing result correction value. And a restriction processing means.

According to the present invention, with the above-described configuration, even if the filter processing result continuously goes out of the limit range, the filter processing result correction value corrected within the limit range does not become flat, and the filter Even if the processing result falls within the limited range, the gradation can be left as the filter processing result correction value.

[0021]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention according to claim 1 of the present invention relates to a filter processing device of a data processing device in which input / output is performed by a signal line having a fixed bit width and a limited range of a data value is determined. Filter processing means for filtering the signal, and whether a filtering result of the filter processing means exceeds an upper limit value of the restriction range, is within the restriction range, or falls below a lower limit value of the restriction range. Determining means for determining whether a first difference between the immediately preceding filtering result correction value and the upper limit value and a second difference between the immediately preceding filtering result correction value and the data maximum value; Upper limit correcting means for correcting the filtering result within the limit range based on the ratio of the second difference, a third difference between the immediately preceding filtering result corrected value and the lower limit value, and Lower limit correcting means for obtaining a fourth difference between the filter processing result correction value and the data minimum value, and correcting the filter processing result within the limited range based on a ratio of the third difference to the fourth difference; If the processing result exceeds the upper limit value, select the output of the upper limit correction means, if the filter processing result is less than the lower limit value, select the output of the lower limit correction means,
When the filter processing result is within the limit range, the filter processing result is selected as a filter processing result correction value. Has the effect that the gradation can be corrected to be within the limit range while leaving the gradation of the filter processing result.

According to a second aspect of the present invention, there is provided an image processing method for inputting / outputting a signal line having a fixed bit width and processing an image having a limited data value range. Is filtered, it is determined whether the filter processing result is above the upper limit of the limit range, within the limit range, or below the lower limit of the limit range, the immediately preceding filter processing result Calculating a first difference between a corrected value and the upper limit value and a second difference between the immediately preceding filter processing result corrected value and a data maximum value;
The filter processing result is corrected to be within the limit range based on the ratio of the difference and the second difference to obtain an upper limit correction result, and a third difference between the immediately preceding filter processing result correction value and the lower limit value and the immediately preceding filter processing result are corrected. Calculating a fourth difference between the filter processing result correction value and the data minimum value, correcting the filter processing result within the limit range based on a ratio of the third difference to the fourth difference to obtain a lower limit correction result; When the processing result exceeds the upper limit, the upper limit correction result is selected. When the filter processing result is lower than the lower limit, the lower limit correction result is selected. If the value is within the range, the filter processing result is selected and used as a filter processing result correction value. It has the effect that it is Rukoto.

According to a third aspect of the present invention, there is provided an apparatus for processing an image in accordance with ITC-R601 in which input / output is performed by a signal line having a fixed bit width and a limited range of image data values is determined. In the image processing apparatus, a filter processing unit that performs a filtering process on an input image signal, and a filter processing result of the filter processing unit that exceeds an upper limit value of the restriction range, is within the restriction range, or sets a lower limit value of the restriction range. Determining means for determining which of the following is below;
A first difference between the immediately preceding filtering result correction value and the upper limit value and a second difference between the immediately preceding filtering result correction value and the data maximum value are determined, and a second difference between the first difference and the second difference is determined. Upper limit correcting means for correcting the filter processing result within the limit range, a third difference between the immediately preceding filter processing result corrected value and the lower limit value, and a third difference between the immediately preceding filter processing result corrected value and the data minimum value. The fourth difference is obtained, and the third difference is obtained.
A lower limit correction unit that corrects the filter processing result within the limit range based on a ratio of the difference and the fourth difference, and an output of the upper limit correction unit when the filter processing result exceeds the upper limit value. Selecting, selecting the output of the lower limit correction means if the filtering result is below the lower limit, selecting the filtering result if the filtering result is within the limiting range, A filter processing result correction value limiting means, which is capable of keeping the filter processing result correction value within the limit range while leaving the tone which should be present in the filter processing result of the image data. Having.

Hereinafter, an embodiment of the present invention will be described with reference to FIG.
This will be described with reference to FIG. Here, a case will be described in which the data in Table 2 is subjected to filter processing for each of the seven data by the filter coefficients in Table 1. The data is 8 bits, the upper limit is 235, and the lower limit is 16.

(First Embodiment) In a first embodiment of the present invention, the difference between the immediately preceding filtering result correction value and the upper limit value of the limit range, and the immediately preceding filtering result correction value and the data maximum. Upper limit correction means for obtaining a difference between the filter processing result and the difference between the immediately preceding filter processing result correction value and the filter processing result, and correcting the filter processing result within a limited range; The difference between the processing result correction value and the lower limit value of the limit range and the difference between the immediately preceding filtering result correction value and the data minimum value are obtained, and the ratio of the two differences is calculated. Multiplying the difference between the filter processing result and the lower limit correction means for correcting the filter processing result within the limit range, and selecting the output of the upper limit correction means when the filter processing result exceeds the upper limit value. If you have Selects the output of the limited correction means, if the filtering processing result is within limits a filtration device that includes a restriction processing means for selecting the filtering result to the filtering result correction value.

FIG. 1 is a block diagram showing the configuration of the filter processing device according to the first embodiment of the present invention. The filter processing means 1 is a circuit for performing a filter processing on input data such as image data, and is composed of multiplication processing means 4-1 to 4-7 and an addition means 7. The judging means 2 is a circuit for judging the result of the filter processing, and includes an upper limit judging means 9, a lower limit judging means 10, an upper limit flag output means 13, and a lower limit flag output means 14.
And upper limit correcting means 15 and lower limit correcting means 16. The restriction processing means 3 is a selection circuit for keeping the filter processing result within the restriction range.
It consists of 18 and.

Each of the multiplication processing means 4-1 to 4-7 is an arithmetic processing circuit for one data in the filter processing, and includes data storage means 5-1 to 5-7 and coefficient storage means 6-1 to 6-7. And a multiplication circuit. The data storage units 5-1 to 5-7 are storage circuits that store input data X1 to X7 in the filter processing. The coefficient storage means 6-1 to 6-7 are storage circuits for storing coefficients A1 to A7 for multiplying data in the filter processing. The addition means 7 is an addition circuit for adding the results of the input data multiplied by the multiplication processing means 4-1 to 4-7.

The addition result storage means 8 is a storage circuit for storing the result added by the addition means 7. Upper limit judgment means 9
Is a circuit for determining whether or not the value of the filter processing result in the addition result storage means 8 exceeds the upper limit of the restriction range. The lower limit judging means 10 is a circuit for judging whether or not the value of the filter processing result in the addition result storing means 8 is lower than the lower limit of the limit range. The upper limit value storage unit 11 is a storage circuit that stores the upper limit value of the limit range. The lower limit value storage means 12 is a storage circuit for storing a lower limit value of the limit range. The upper limit flag output means 13 is a circuit that outputs a flag signal indicating whether or not the filtering result exceeds the upper limit of the limit range. The lower limit flag output means 14 is a circuit that outputs a flag signal indicating whether or not the result of the filter processing is below the lower limit of the limit range.

The upper limit correcting means 15 is an arithmetic circuit for rounding the result of the filtering process into the limit range when the result exceeds the upper limit of the limit range. The lower limit correction means 16 is an arithmetic circuit for rounding the result of the filter processing into the limited range when the result of the filtering is below the lower limit of the limited range. The selector 17 is a selection circuit for selecting and outputting one of the addition result storage means 8, the upper limit correction means 15 and the lower limit correction means 16 according to the output flags of the upper limit flag output means 13 and the lower limit flag output means 14. . The storage unit 18 is a storage circuit for storing a previous filter processing result correction value.

The operation of the filter processing device configured as described above will be described with reference to the block diagram of FIG. 1 and the flowchart of FIG. First, the case where the filtering result exceeds the upper limit value (235) of the restriction range will be described. In step 21, the filter coefficients A1 to A7 are set in the coefficient storage means 6-1 to 6-7, the limit range upper limit value Xmax is set in the upper limit value storage means 11, and the limit range lower limit value Xmin is set in the lower limit value storage means 12. Set.

Next, in step 22, the input data X1 to X6 are sequentially stored in the data storage means 5-2 to 5-7. Next, in step 23, the data storage means 5-2 to 5
-7 is transferred to the data storage means 5-1 to 5-6, and the input data X7 enters the data storage means 5-7. That is, the data D1 to D7 in Table 2 are represented by X1 to X7 as data storage units.
Stored in 5-1 to 5-7. Next time, data D2 of Table 2
D8 is stored in the data storage means 5-1 to 5-7 as X1 to X7.

Next, in the filter processing in step 24, the multiplication of the data X1 to X7 of the data storage means 5-1 to 5-7 and the coefficients A1 to A7 of the coefficient storage means 6-1 to 6-7 is performed for each data. The output data of the multiplication means 4-1 to 4-7 are added in the addition means 7. Next, in step 25,
The result Xtmp added by the addition means 7 is stored in the addition result storage means 8. That is, X1 × A1 + X2 × A2 +
.. + X7 × A7 are held in Xtmp.

Next, in step 26, the addition result storage means 8 is read from Xmax stored in the upper limit value storage means 11.
Is subtracted, and if the result is negative, 1 is substituted for Ra of the upper-limit flag output means 13; if the result is positive (non-negative), 0 is substituted for Ra of the upper-limit flag output means 13. Since the case where the filtering result exceeds the upper limit of the limit range is considered, 1 is substituted for Ra, and the routine proceeds to step 31.

In step 31, the upper limit correcting means 15
When the filter processing result exceeds the upper limit of the limit range, a process of setting the value within the limit range is performed. This processing is obtained by processing the filter processing result Xtmp held in the addition result storage means 8 using the following equation 1. X = (Xtmp−X ′) (Xmax−X ′) / (Xlimita−X ′) + X ′ (1) (where X = filter processing result correction value corrected within the limit range, Xmax = upper limit value of the limit range) (235), X '= correction value of previous filter processing result, Xlimita = maximum data value (255))

Next, at step 29, X is output. This is output when the selector 17 selects the upper limit correcting unit 15 and the storage unit 18 stores the output value of the upper limit correcting unit 15. Finally, in step 30, X 'of the upper limit correcting means 15 and the lower limit correcting means 16 is replaced with X.

Next, a case where the filtering result is smaller than the lower limit value (16) of the limit range will be described. Steps 21 to 25 are the same as the case where the value exceeds the upper limit value of the above-described restriction range. In step 26, Xtmp is subtracted from Xmax.
Now, it is assumed that the filter processing result is below the lower limit of the limit range, so that Xtmp is smaller than Xmax. As a result,
Proceed to step 27. In step 27, the lower limit determining means
In step 10, Xmin is subtracted from Xtmp, and if the result is negative, 1 is substituted for Rb of the upper limit flag output means 13, and if positive (non-negative), 0 is substituted for Rb of the upper limit flag output means 13. . Now, consider the case where the filter processing result is below the lower limit of the limit range.
Proceed to 2. In step 32, when the filter processing result falls below the lower limit of the limit range, a process of setting the value within the limit range is performed. This processing is obtained by processing the filter processing result Xtmp held in the addition result storage means 8 using the following equation (2). X = (Xtmp−X ′) (Xmin−X ′) / (Xlimitb−X ′) + X ′ (2) (where, X = filtered result corrected value within the limit range, Xmin = lower limit value of the limit range) (16), X '= correction value of the immediately preceding filter processing result, Xlimitb = minimum data value (0))

Next, at step 29, X is output.
This means that the selector 17 selects the lower limit correction means 16 and the storage means
The output value is output when the output value of the lower limit correcting means 16 is stored in 18. Finally, in step 30, the upper limit correcting means 15
And X ′ of the lower limit correcting means 16 is replaced with X.

Finally, a case where the result of the filtering process is within the limited range (16 to 235) will be described. Step 21
Are the same as when exceeding the upper limit of the restriction range. In step 26, Xmax and Xtmp are compared.
Since Xtmp is within the limit range, the upper limit flag output means 13
Becomes 0, and the process proceeds to step 27. In step 27, Xtmp and Xmin are compared. Since Xtmp is within the limit range, Rb of the lower limit flag output means 14 becomes 0, and the routine proceeds to step 28. In step 28, the output value of the addition result storage means 8 is input to the selector 17, and the process proceeds to step 29. In step 29, X is output. This is output when the addition result storage means 8 is selected by the selector 17 and the output value of the addition result storage means 8 is stored in the storage means 18. Finally, in step 30, X 'of the upper limit correcting means 15 and the lower limit correcting means 16 is replaced with X.

An example in which the data of Table 2 is input to the filter processing device having the filter coefficients of Table 1 will be described. The filter processing result Xtmp and the filter processing result correction value X are
Table 3 below.

[0040]

FIG. 3 shows a line graph of the data in Table 3. The line graph 35 is an uncorrected filter processing result Xtmp, and the line graph 36 is a filter processing result correction value X after the correction processing. As described above, at t3 and t4 when the filter processing result Xtmp exceeds the upper limit (235), the filter processing result correction value X becomes 226 and 234, and the gradation remains. At t7, the output of the adding means 7 becomes negative, but becomes 0 when converted to 8-bit data, so that the lower limit value (16) is output. When the filtering result exceeds the maximum value (255) of the 8-bit data or falls below the minimum value (0), the upper limit value (235) is set as the filtering result correction value.
Alternatively, the lower limit (16) is output.

In the present embodiment, the flowchart of FIG. 2 shows a processing procedure in which the filter processing result is compared with the upper limit value of the limit range, and the filter processing result is compared with the lower limit value of the limit range in order. Higher-speed processing can be realized by performing the processing at the same time and calculating the filter processing result within the limited range at the same time.

As described above, according to the filter processing apparatus of the present embodiment, the difference between the immediately preceding filtering result correction value and the upper limit (or lower limit) of the limit range, and the difference between the immediately preceding filtering result correction value and the immediately preceding filtering result correction value. The difference between the maximum value (or the minimum value) of the data is obtained, the ratio of the two differences is multiplied by the difference between the immediately preceding filtering result correction value and the filtering result, and the filtering result is corrected within the limit range. Since the correction value is selected in accordance with the result of the determination means, the correction value asymptotically approaches the upper limit value (or lower limit value) of the limit range even if the filter processing result continuously goes out of the limit range. , And the gradation can be left in the filter processing result correction value.

The present invention is effective when applied to a filter processing device in an image processing device, particularly, an image processing device for processing an image conforming to ITC-R601. The same effect is exhibited when applied to a device or another signal processing device.

[0045]

According to the present invention, as apparent from the above embodiment, when the filtering result is out of the limit range, the difference between the immediately preceding filtering result correction value and the upper limit value (lower limit value) of the limit range. , And the difference between the immediately preceding filter processing correction value and the maximum value (minimum value) of the data is determined, and the filter processing result is corrected within the limit range based on the ratio of the two differences. Therefore, even if the value falls outside the limit range, it is possible to obtain the effect that the gradation can be left in the filter processing result correction value within the limit range.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a filter processing device according to a first embodiment of the present invention;

FIG. 2 is a flowchart showing processing according to the first embodiment of the present invention;

FIG. 3 is a graph showing a filter processing result of the present invention;

FIG. 4 is a block diagram showing a configuration of a conventional filter processing apparatus.

FIG. 5 is a flowchart showing processing of a conventional filter processing apparatus;

FIG. 6 is a graph showing a filter processing result of a conventional example.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Filter processing means 2 Judgment means 3, 43 Restriction processing means 4-1 to 4-7 Multiplication processing means 5-1 to 5-7 Data storage means 6-1 to 6-7 Coefficient storage means 7 Addition means 8 Addition result storage Means 9 Upper limit judgment means 10 Lower limit judgment means 11 Upper limit value storage means 12 Lower limit value storage means 13 Upper limit flag output means 14 Lower limit flag output means 15 Upper limit correction means 16 Lower limit correction means 17, 45 Selector 18 Storage means 46 Output data storage

Claims (3)

[Claims] 1. A filter processing device for performing input / output on a signal line having a fixed bit width and having a fixed data value limiting range, the filter processing means for filtering an input signal, and the filter processing means Determining means for determining whether the filter processing result is above the upper limit value of the restriction range, within the restriction range, or below the lower limit value of the restriction range, and correcting the immediately preceding filter processing result Calculating a first difference between the value and the upper limit value and a second difference between the immediately preceding filtering result correction value and the data maximum value, and calculating the filtering result based on a ratio of the first difference and the second difference. Upper limit correction means for correcting within the limit range, a third difference between the immediately preceding filter processing result correction value and the lower limit value, and the immediately preceding filter processing result correction value and data minimum value. And a lower limit correcting means for correcting the filtering result within the limit range based on the ratio of the third difference and the fourth difference, and the filtering result exceeds the upper limit value. In the case, the output of the upper limit correction means is selected, and when the filter processing result is lower than the lower limit, the output of the lower limit correction means is selected, and the filter processing result is within the limit range. In some cases, the filter processing apparatus further comprises a restriction processing unit that selects the filter processing result and sets the result as a filter processing result correction value. 2. An image processing method for performing input / output on a signal line having a fixed bit width and processing an image having a limited data value range, wherein an input image signal is filtered, and It is determined whether the value exceeds the upper limit value of the limit range, is within the limit range, or is lower than the lower limit value of the limit range. Calculating a difference and a second difference between the immediately preceding filtering result correction value and the data maximum value, correcting the filtering result within the limited range based on a ratio of the first difference to the second difference, and setting an upper limit. As a correction result, a third difference between the immediately preceding filtering result correction value and the lower limit value and a fourth difference between the immediately preceding filtering result correction value and the data minimum value are obtained, and the third difference and the fourth difference are obtained. Based on the ratio of The filter processing result is corrected to be within the limit range as a lower limit correction result, and when the filter processing result exceeds an upper limit value, the upper limit correction result is selected, and the filter processing result falls below the lower limit value. The lower limit correction result is selected when the filter processing result is within the limit range, and the filter processing result is selected and set as a filter processing result correction value when the filter processing result is within the limit range. . 3. An ITC-R601 in which input / output is performed by a signal line having a fixed bit width, and a limit range of image data values is determined.
In an image processing apparatus for processing an image conforming to the above, a filter processing unit for filtering an input image signal, and a filter processing result of the filter processing unit exceeds an upper limit value of the limit range or falls within the limit range. Determining means for determining whether the value is below the lower limit value of the restriction range, a first difference between the immediately preceding filter processing result correction value and the upper limit value, and the immediately preceding filter processing result correction value. Upper limit correction means for obtaining a second difference from a data maximum value, and correcting the filtering result within the limited range based on a ratio of the first difference to the second difference; And a fourth difference between the immediately preceding filtering result correction value and the data minimum value, and performing the filtering on the basis of the ratio of the third difference to the fourth difference. The lower limit correcting means for correcting the result within the limit range, and when the filter processing result exceeds the upper limit value, the output of the upper limit correcting means is selected, and the filter processing result falls below the lower limit value. If the filter processing result is within the limit range, the filter processing result is selected and the filter processing result is corrected. An image processing apparatus characterized by the above-mentioned.