WO2003001793A1 - Procédé et dispositif de traitement d'images - Google Patents
Procédé et dispositif de traitement d'images Download PDFInfo
- Publication number
- WO2003001793A1 WO2003001793A1 PCT/JP2002/006106 JP0206106W WO03001793A1 WO 2003001793 A1 WO2003001793 A1 WO 2003001793A1 JP 0206106 W JP0206106 W JP 0206106W WO 03001793 A1 WO03001793 A1 WO 03001793A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- value
- pixel
- threshold
- image processing
- image
- Prior art date
Links
- 238000003672 processing method Methods 0.000 title claims description 17
- 238000012545 processing Methods 0.000 claims abstract description 42
- 238000006243 chemical reaction Methods 0.000 claims description 75
- 238000000034 method Methods 0.000 claims description 35
- 230000008569 process Effects 0.000 claims description 21
- 238000001914 filtration Methods 0.000 claims description 16
- 230000007423 decrease Effects 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 4
- 238000005286 illumination Methods 0.000 abstract description 76
- 230000006835 compression Effects 0.000 abstract description 22
- 238000007906 compression Methods 0.000 abstract description 22
- 230000014509 gene expression Effects 0.000 description 27
- 238000010586 diagram Methods 0.000 description 22
- 230000008859 change Effects 0.000 description 17
- 238000007796 conventional method Methods 0.000 description 11
- 230000000694 effects Effects 0.000 description 8
- 238000012937 correction Methods 0.000 description 7
- 230000005540 biological transmission Effects 0.000 description 6
- 239000000284 extract Substances 0.000 description 6
- 230000001965 increasing effect Effects 0.000 description 5
- 230000001186 cumulative effect Effects 0.000 description 4
- 230000002349 favourable effect Effects 0.000 description 4
- 238000009499 grossing Methods 0.000 description 3
- 238000007792 addition Methods 0.000 description 1
- 230000001427 coherent effect Effects 0.000 description 1
- 102000007336 epsin Human genes 0.000 description 1
- 108010032643 epsin Proteins 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000002310 reflectometry Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/14—Picture signal circuitry for video frequency region
- H04N5/20—Circuitry for controlling amplitude response
- H04N5/205—Circuitry for controlling amplitude response for correcting amplitude versus frequency characteristic
- H04N5/208—Circuitry for controlling amplitude response for correcting amplitude versus frequency characteristic for compensating for attenuation of high frequency components, e.g. crispening, aperture distortion correction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20004—Adaptive image processing
- G06T2207/20012—Locally adaptive
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20172—Image enhancement details
- G06T2207/20192—Edge enhancement; Edge preservation
Definitions
- the present invention can be suitably used for various types of image input / output devices such as a television, a video tape recorder, a still camera, a video camera, and a printer.
- the present invention relates to an image processing method and apparatus for reproducing in an image apparatus having a narrow dynamic range.
- a function having an input / output relationship as shown by a solid line in FIG. 18 for each pixel of an input image (hereinafter referred to as a “level conversion function”) ) (Hereinafter referred to as “level conversion”).
- the horizontal axis represents the pixel level (input level) 1 of the input image
- the vertical axis represents the pixel level (output level) T (1) of the output image obtained by the level conversion process.
- L max represents the maximum level that each pixel of the input / output image can take.
- the contrast of the image after level conversion increases as the slope of the level conversion function increases. In the example in Fig.
- the slope of the straight line indicating the level conversion function between the high level side with the input level 1b as the boundary and the low level side with the input level 1s as the middle level (the input level 1 The slope is smaller than the slope of s ⁇ 1b). Therefore, in the level conversion using the function shown in Fig. 18, the contrast at the intermediate level is increased by sacrificing the contrast at the high level and the low level.
- the level conversion function is not limited to the one shown in FIG. 18, and for example, the one shown by a solid line in FIG. 19 can be used. In the level conversion function shown in Fig. 19, the slope of the straight line at the high level is smaller than the slope at the low level and the middle level at the input level lk. Therefore, in the level conversion using the function shown in Fig.
- the contrast at the low level and the intermediate level can be increased by sacrificing the contrast at the high level.
- the gun shown in Equation 1 In some cases, a more continuous level conversion function is used than the functions shown in FIGS. 18 and 19, such as the MA function and the LOG function shown in Equation 2.
- G in Equation 1 is a parameter for adjusting the slope of the function.
- FIG. 2A and FIG. 20B illustrate the principle of this histogram equalization.
- the horizontal axis represents the pixel level (input level) 1 of the input image
- the vertical axis represents the frequency (frequency or cumulative frequency).
- Fmax is the maximum value of the cumulative frequency, and is the total number of pixels used for calculating the frequency.
- a frequency distribution H (1) for pixel level 1 of the input image is generated, and then, a cumulative frequency distribution C (1) is generated using Equation 3. .
- the level conversion function T (1) is generated by normalizing the vertical axis of the cumulative frequency distribution C (1) to the level range that the output image can take using Equation 4 (see FIG. 20B). ). By using this function T (1), it is possible to increase the contrast of a region (a region with a large area) constituted by levels with a high frequency of appearance.
- a level conversion function having a monotonous increasing property is used in order to avoid generation of an unnatural image. For this reason, when the contrast (gradient of the level conversion function) in one of the level ranges is increased, the contrast is reduced in the other level ranges.
- the Muliiscale retine method makes it possible to reproduce images with higher contrast by sacrificing monotonicity, but when the lighting conditions change rapidly, the linear filter extracts the change And there is a problem that subjectively undesirable noise occurs.
- Equation 5 is equivalent to dividing the input signal by the illumination component, meaning that the larger the illumination component is, the greater the compression is, and as a result, the BNB region of the reproduced image (the thick broken line in the figure) Overshoot occurs.
- C Multiscale retine In order to avoid this problem, the method uses multiple linear low-pass filters with different scales and combines the results obtained from each low-pass filter with a linear load, but the weight for each scale is fixed. The above problems have not been sufficiently suppressed.
- Epsilon filter Is superior in its ability to preserve edges compared to linear filters, and can extract illumination components more effectively even in images with different illumination light.
- a fixed threshold epsilon filter which is generally used for noise removal, etc., produces a discontinuous waveform near the edge of its output, and is used for compression of the dynamic range. In this case, an unnatural image pattern that is not in the original image may occur in the compressed reproduced image.
- the present invention has been made in view of such a problem, and an object of the present invention is to make it possible to appropriately extract the boundaries of a plurality of different illuminations using an epsilon filter even when a plurality of different illuminations exist.
- An object of the present invention is to provide an image processing method and apparatus capable of realizing a subjectively preferable compression of a dynamic range while suppressing the occurrence of a natural image pattern. Disclosure of the invention
- An image processing method includes: an edge strength calculating step of calculating an edge strength for each position on an input image; and a threshold control for controlling a threshold of an epsilon filter based on the calculated edge strength.
- a filtering process in which an input image is filtered by an epsilon filter using the threshold value controlled in the threshold control process, and a pixel value is converted according to an output value of the filtering process.
- a pixel value conversion process of converting a pixel value for each pixel based on the calculated coefficient.
- An image processing apparatus includes: an edge strength calculation unit configured to calculate an edge strength for each position on an input image; and an epsilon filter configured to perform a filter process on the input image using a set threshold.
- a threshold value controlling means for controlling a threshold value used in the epsilon filter based on the edge strength calculated by the edge strength calculating means; and a pixel value according to an output value from the epsilon filter.
- Pixel value conversion means for calculating a coefficient for converting the pixel value, and converting a pixel value for each pixel using the calculated coefficient.
- the edge strength is calculated for each position on the input image, and the edge strength is calculated based on the calculated edge strength.
- the threshold used is controlled.
- a coefficient for converting the pixel value is calculated according to the output value from the epsilon filter, and the calculated coefficient is used to convert the pixel value for each pixel.
- the threshold value of the epsilon filter is controlled, for example, such that the value becomes smaller as the edge strength becomes larger.
- the threshold value may be controlled to increase as the pixel value of the input image increases.
- two thresholds having different magnitudes may be calculated and the thresholds may be controlled.
- filter processing is performed using different threshold values depending on whether the value of the neighboring pixel is larger or smaller than that of the pixel of interest. Also in this case, in the filtering process, the influence of the illumination level on the change in the pixel value is reduced, and the illumination component is more appropriately extracted.
- FIG. 1 is a block diagram showing a configuration of an image processing apparatus according to a first embodiment of the present invention.
- FIG. 2 is an explanatory diagram showing a scanning direction of an image.
- FIG. 3 is an explanatory diagram showing an example of a level conversion function.
- FIG. 4 is a block diagram showing a configuration of an epsilon filter in the image processing apparatus shown in FIG.
- 5A to 5C are explanatory diagrams showing the difference between the reproduced image by the conventional method and the reproduced image by the image processing device shown in FIG.
- 6A and 6B are explanatory diagrams showing the relationship between the level conversion function and the coefficient calculation function.
- FIG. 7A and FIG. 7B are explanatory diagrams showing the effect of epsilon fill.
- FIGS. 8A and 8B are explanatory diagrams showing the behavior of an epsilon filter with a fixed threshold value around an edge.
- FIG. 9A and FIG. 9B are explanatory diagrams showing a reversal phenomenon of a level gradient occurring in an epsilon filter having a fixed threshold.
- FIG. 10 is an explanatory diagram showing the relationship between the level conversion function used for the output of the epsilon filter and its differential value.
- FIG. 11A and FIG. 11B are explanatory diagrams showing the magnitude function of the input level and the compression ratio in the level conversion curve.
- FIG. 12 is a block diagram showing a configuration of an image processing apparatus according to a second embodiment of the present invention.
- FIG. 13 is an explanatory diagram showing an example of a coefficient calculation function used in a coefficient calculator in the image processing device shown in FIG.
- FIG. 14 is a block diagram showing a configuration of an image processing apparatus according to the third embodiment of the present invention.
- FIG. 15 is a block diagram showing a configuration of the image processing apparatus according to the fourth embodiment of the present invention.
- FIG. 16 is an explanatory diagram showing a configuration of an epsilon filter in the image processing apparatus shown in FIG.
- FIG. 17 is a block diagram showing a configuration of an image processing apparatus according to a fifth embodiment of the present invention.
- FIG. 18 is an explanatory diagram showing an example of a conventionally used level conversion function.
- FIG. 19 is an explanatory diagram showing an example of another level conversion function conventionally used.
- FIG. 2OA and FIG. 20B are explanatory diagrams showing the principle of histogram equalization.
- FIG. 21 is a diagram for explaining a problem of the Multiscale int method. BEST MODE FOR CARRYING OUT THE INVENTION
- the input image signal processed by the present image processing apparatus is a signal of time-series pixel values obtained by scanning a two-dimensional digital image in the horizontal and vertical directions in the order shown in FIG.
- a pixel value corresponding to an arbitrary position (X, y) on a two-dimensional image is represented as I (x, y), and this is processed as an input image signal.
- the image processing apparatus includes an edge strength calculator 10, a threshold controller 11, an epsilon filter 12, a divider 13, and a level converter 14. And a multiplier 15.
- the edge strength calculator 10 has a function of calculating the edge strength G (x, y) of the pixel value I (X, y) at each position of the input image.
- the edge strength G (x, y) for example, the first derivative of I (X, y) given by Equation 6 can be used.
- Equation 7 it is also possible to use a value according to Equation 7 having a smoothing effect as the edge strength G (X, y) in order to suppress the influence of noise.
- d is a constant indicating a minute distance for calculating the derivative.
- the threshold controller 11 calculates a threshold E (x, y) used in an epsilon filter 12 described later. It has a function of determining the size for each pixel. By the function of the threshold controller 11, the threshold value E (X, y) is controlled so as to become smaller as the edge strength G (x, y) becomes larger, for example, by using Expression 8. You.
- Grain, Graax, Emin, and Ema are constants for converting the edge strength G (, y) into a threshold ⁇ (x, y). Indicates the minimum value, maximum value, and minimum and maximum values of the threshold value E (x, y).
- the threshold value E (x, y) determined by the threshold value controller 11 is sent to the epsilon filter 12.
- the epsilon filter 12 includes, for example, a differentiator 20, an absolute value calculator 21, a comparator 22, and a linear one-pass filter (in the figure, denoted as LPF. 2) and 3).
- the epsilon filter 12 is a two-dimensional filter, and performs a nonlinear filtering process on an input image using the threshold value E (x, y) determined by the threshold value controller 11. have.
- the output R (X, y) of the Ebsilon filter 12 is sent to a divider 13 and a level converter 14 as an illumination component.
- the divider 13 converts each pixel value I (X, y) of the input image into the illumination component R as shown in Expression 9. Divide by (x, y). The non-lighting component S (X, y) obtained as a result of the division is sent to a multiplier 15.
- the level converter 14 compresses the illumination component R (X, y) calculated by the epsilon filter 12 by performing level conversion using a level conversion function T (1) as shown in Expression 10. It has a function of calculating the corrected illumination component CR (x, y).
- level conversion function T (1) used in the level converter 14 for example, a function as shown in FIG. 3 can be used.
- Rmax and C Rmax represent the maximum values of the input level and the output level, respectively.
- the multiplier 15 restores the image signal by multiplying the non-illumination component S (X, y) by the correction pig light component CR (X, y).
- the image signal 0 (X, y) indicating the restored result is output to a transmission line, a storage device, or a display device (not shown).
- a signal indicating an input image is input to an edge strength calculator 10, an epsilon filter 12 and a multiplier 13.
- the edge strength calculator 10 calculates the size of the edge for each position of the input image, that is, the edge strength G (X, 5).
- the edge strength calculator 10 calculates, for example, Equation 6 or Equation 6.
- the edge intensity G (x, y) is calculated such that the larger the first derivative of the pixel value in the vicinity of the target pixel becomes, the larger the edge intensity G (x, y) is.
- the threshold controller 11 controls the threshold E of the epsilon filter 12 based on the edge strength G (x, y). More specifically, the threshold value controller 11 determines the magnitude of the threshold value E (X, y) for each pixel using, for example, Equation 8, and the edge intensity G (x, y) is Control is performed so that the threshold value E (X, y) decreases as the value increases. The threshold controller 11 outputs the determined threshold E (x, y) to the epsilon filter 12.
- the filtering process in the epsilon filter 12 is performed as follows, for example, by the configuration shown in FIG.
- the differentiator 20 includes a signal indicating the current value I (X, y) of the target pixel and the value I (X x + dx, y + dy) and are input.
- the differentiator 20 calculates a difference between the value I (X, y) of the target pixel and the value I (x + dx, y + dy) of the pixel in the neighboring area NB.
- the differentiator 20 sequentially calculates this difference value for all the pixels in the neighboring area NB, and outputs the value D (d X, dy) to the absolute value calculator 21 in association with each neighboring pixel. I do.
- the absolute value calculator 21 calculates an absolute value AD (dx, dy) of each difference value D (dX, dy) sent from the differentiator 20.
- the absolute value calculator 21 outputs the calculated absolute value AD (dX, dy) to the comparator 22.
- the comparator 22 receives the absolute value AD (d X, dy) calculated by the absolute value calculator 21 and outputs a signal indicating the value I (x, y) of the pixel of interest and its neighboring area N A signal indicating the value I (x + dx, y + dy) of the pixel in B and the threshold value E (X, y) determined by the threshold controller 11 are input.
- the comparator 22 compares the absolute value AD (d X, dy) with the threshold value E (, y) as shown in Expression 12, and according to the result, the value I (X, y ) Or the value I (x + d X, y + dy) of the neighboring pixel, and outputs it to the linear low-pass filter 23 as the value J (d, dy).
- the linear low-pass filter 23 calculates the weighted average value R (x, y) according to Expression 13 when the value J (dx, dy) corresponding to all the pixels in the neighboring area NB is calculated by the comparator 22. I do.
- NB is a set of relative coordinates that define a neighboring area in the filtering process.
- a (dX, dy) is a weighting coefficient for each pixel value.
- the linear low-pass filter 23 for example, an average value filter as shown in Expression 14 can be used.
- Equation 14 N represents the number of NB pixels in the neighboring area.
- the purpose of the Ebsilon filter 12 is to remove fine structures in an image and to extract a coherent area, and it is desirable that the area in the vicinity is large.
- the value R (x, y) obtained in the epsilon fill 12 as described above is considered to approximately represent the illumination component contained in the image.
- the epsilon filter 12 outputs the value R (x, y) to the divider 13 and the level converter 14 as an illumination component.
- the divider 13 divides each pixel value I (X, y) of the input image by the illumination component R (X, y) as shown in Expression 9, thereby obtaining the input image by the epsilon filter 12.
- the calculated illumination component is removed, and the resulting non-illumination component S (X, y) is output to the multiplier 15.
- the level converter 14 performs level conversion of the illumination component R (X, y) calculated by the epsilon filter 12 using, for example, a level conversion function T (1) as shown in FIG. Compress and calculate the corrected illumination component CR (x, y).
- the level converter 14 outputs the calculated corrected illumination component CR (X, y) to the multiplier 15.
- the multiplier 15 multiplies the non-illumination component S (x, y) output from the divider 13 by the corrected illumination component CR (x, y) output from the level converter 14. Restore the image signal.
- the non-illumination component S (x, y) is multiplied by the corrected illumination component CR (x, y).
- a coefficient F (R (x, y)) for converting a pixel value according to the output value R (x, y) from the epsilon filter 12 is calculated, and By multiplying by the corresponding input pixel value I (x, y), it is equivalent to converting the pixel value for each pixel and compressing the dynamic range.
- the image signal ⁇ (x, y) output from the multiplier 15 is an image device having a relatively smaller dynamic range than the input image, that is, an environment in which the number of bits representing the pixel level is small. It is used in cases such as transmission on a transmission line with a small number of bits, display on a display device, or storage in a storage device.
- FIG. 5A is a diagram showing the pixel value I (X, y) of the input image and the output R (x, y) from the epsilon filter 12 in the present embodiment as a one-dimensional signal. is there.
- FIG. 5B shows the result (reproduced image) of the dynamic range compression performed by the conventional level conversion method on the input image shown in FIG. 5A
- FIG. 5C shows the dynamic image in the present embodiment. The result of range compression is shown.
- Equation 16 shows that the dynamic range compression by the compression of the illumination component R (, y) reduces the coefficient F (R (X, y)) calculated for each pixel to the corresponding input pixel value I (X, y).
- the coefficient calculation function F (1) has a function of converting the output value of the epsilon filter 12 into a gain coefficient to be applied to each pixel. Note that the minimum value C min of the value of the coefficient calculation function F (1) in FIG. 6B is given by Expression 17.
- the conventional method can preserve the contrast in the low-level area (the area composed of levels 11 and 13 shown in FIG. 5A), In the level region (region composed of level 14 and level 16), the contrast is reduced. This is the result of the direct effect of the slope of the level conversion function T (1) for higher levels above the inflection point lk. In order to improve the contrast in the conventional method, it is necessary to increase the slope of the level conversion function T (1).
- a single correction coefficient given by the coefficient calculation function F (1) is applied to each of the high-level region and the low-level region.
- the contrast in each area depends on the magnitude of this correction coefficient.
- the correction coefficient determined by the average level 12 is uniformly applied to the low-level area, but the value is the same as that for levels 11 and 13. 0, and the same level of contrast as the conventional method can be obtained.
- a constant correction coefficient c5 determined by the average value 15 is applied, so that the contrast between the level 14 portion and the level 16 portion is determined by this gain. Will be secured.
- the straight line corresponding to the inflection point level 1k or more is expressed by equation 18
- the contrast is defined as "(maximum level minimum level) / average level”.
- the correction coefficient c5 applied to the high-level region is given by: Given in nineteen.
- the contrast in the region extracted by the epsilon filter 12 is determined by the value of the correction coefficient itself given by the coefficient calculation function F (1), and the level conversion function T (1)
- the slope of the graph affects the contrast between regions. Therefore, according to the present embodiment, by compressing the contrast between the regions, the contrast within the region can be preserved, and a subjectively favorable output image can be obtained.
- the large difference absolute value AD (d X, dy) exceeding the threshold value E (x, y) is set in the area NB near the pixel of interest of the epsin filter 12.
- the pixel value I (x + dx, y + dy) is calculated by the comparator 22 (FIG. 4) as shown by the thick broken line (I ′ (dX, dy)) in FIG. 7A.
- the value of the pixel of interest (the value at the center of the neighborhood NB) is replaced by I (X, y), so that it does not significantly contribute to the smoothing by Expression 13, and as a result, the edge shape is preserved.
- I the value at the center of the neighborhood NB
- AD the absolute difference value AD (dX, dy) becomes smaller than the threshold value E (, y) over the entire neighborhood NB of the target pixel.
- J (d X, dy) in Equation 12 are all equal to the input pixel value I (x + dx, y + dy), and the epsilon filter is equivalent to a simple linear low-pass filter, and the neighborhood area Smoothing is performed over the entire area NB.
- the epsilon filter 12 is superior in the ability to preserve edges as compared with the linear filter, and can extract the illumination component more effectively even in an image in which different illumination light exists.
- a discontinuous waveform is formed near the edge of the output, and thus the dynamic range is reduced.
- Figs. 8A and 8B show the output of a conventional epsilon filter with a fixed threshold value at the modeled edge and its surroundings.
- the edge part where the level changes rapidly is approximated by a straight line with a slope a, and the other edges are approximated by a flat straight line with a slope of zero.
- the input signal is indicated by a thin solid line 81
- the output of the ipsi filter is indicated by a thick solid line 82
- the output of the corresponding linear low-pass filter is indicated by a thick broken line 83.
- the linear low-pass filter used in the epsilon filter is an average filter.
- FIG. 8A shows the power of the epsilon filter when the relationship of equation 20 is established between a, N, and E.
- the horizontal axis represents spatial position coordinates, and the coordinate value increases as going to the right.
- ⁇ indicates the position N / 2 away from the edge falling point pe.
- the epsilon filter behaves the same as a linear mouth-pass filter.
- the level change on the image due to the change in the illumination light occurs steeply (the slope a is large), and a large filter must be used to effectively compress the dynamic range as described above.
- FIG. 8B shows the output of the epsilon filter when Expression 20 does not hold.
- the output from ⁇ to P0-E / a is the same as that of a linear filter, and then a constant value is output up to pe. From pe to pe—E / a, it descends according to the quadratic curve, but under the condition that Equation 20 does not hold, it always intersects the straight line 81 indicating the input signal during this time.
- pe-E / a the value becomes discontinuously the same as the input signal, and then the input signal is output as it is.
- This behavior of the fixed threshold epsilon filter is for a simplified edge model, but it is clear that it will output a complex waveform near the edge. In particular, it changes discontinuously around pe-E / a and outputs a waveform with a slope greater than that of the input signal.
- the local shape (the direction of the spatial gradient) at each position on the image must be preserved. That is, as shown in Equation 21, the sign of the differential value O ′ (X, y) of the output image 0 (X, y) and the differential value I ′ (x, y) of the input image I (X, y) The sign of must match.
- sign (x) represents the sign of x.
- the condition of Expression 21 is not satisfied, the level gradient is reversed between the input image and the output image, and an image pattern that does not exist in the input image appears in the output image.
- the input image shown in FIG. 9A is output as shown in FIG. 9B.
- a pattern 90 that does not exist in the input image appears due to the reversal of the level gradient.
- the discontinuous behavior of the epsilon filter near the edge shown in Fig. 8B may make it difficult to satisfy the condition of Equation 21, and the level gradient may be reversed.
- Equation 21 Whether Equation 21 is satisfied or not depends on the behavior of the epsilon filter and the level conversion function T (1) applied to its output. To clarify this, first, Expression 9 is substituted into Expression 11 to obtain Expression 22 of the relation between the input image I (X, y) and the output image ⁇ (X, y).
- FIG. 10 shows the relationship between the output of the epsilon filter and the output of the level conversion function T (1).
- the horizontal axis represents the output of the epsilon filter (input to the level conversion function T (1)), and the vertical axis represents the output of the level conversion function T (1).
- the output of the epsilon filter of a pixel be R
- the value after its level conversion be CR
- the ratio CRZR be a. That is, a corresponds to the inclination of the straight line passing through the origin and (R, CR) (Equation 24).
- Equation 26 can be obtained which consists of a part R'ZR related to the output of the epsilon filter and a part related to the property of the level conversion function T (1) (1-1 b / a).
- Equation 27 Since the level conversion function T (1) for compressing the illumination component usually has a monotonically increasing property, a and b are positive values, and Equation 27 is satisfied.
- 11 ba ⁇ 0 means a ⁇ b.
- the higher the level the greater the compression.
- the compression ratio of level R is a and the area around it is uniform; when compressing, the level; the slope of the level conversion function T (1) in the vicinity must be a, The actual slope is smaller b, and as shown in Fig. 11A, the higher the level, the greater the compression.
- 1-b / a that is, if a ⁇ b, as shown in FIG.
- the vertical and horizontal axes in FIGS. 11A and 11B are the output (input to the level conversion function T (1)) and the level conversion function T ( 1) represents the output.
- Equation 29 is obtained from I ′ / I ⁇ R VR obtained from Equation 28A and Equation 28B.
- Equation 26 holds when I ′ is zero.
- Equation 30 holds in this part. It is clear that 26 is satisfactory.
- the epsilon filter 12 is removed and R (X, y)-I (x, y) is obtained, the result is equivalent to the conventional method using only level conversion. This corresponds to the case where the condition of Equation 30 is satisfied over the entire image, and in the conventional method, the reversal of the level gradient does not occur.
- Equation 26 is satisfied depends on the output of the ipsi filter 12 and the characteristics of the level conversion function T (1). Since the characteristics of the level conversion function T (1) should be determined by how to compress the illumination components, here we will examine how the output of the epsilon filter 12 affects the establishment of Equation 26. Are explained in four cases (1) to (4).
- Equation 26 always holds because the right side is positive and the left side is negative.
- Equation 26 always holds because the right side is negative and the left side is positive.
- Equation 26 always holds regardless of the value of R'ZR.
- Equation 26 is as described above.
- the threshold value E of the epsilon filter 12 is made variable, and a signal as close as possible to the input image is output at the edge portion, that is, the edge intensity G (X, y)
- the threshold value so that the threshold value E of the epsilon filter 12 becomes smaller as the value is larger, the reversal of the gradient is suppressed to the minimum, and the unnaturalness as shown in Fig. 9B This makes it possible to reproduce a natural image so that the pattern 90 does not occur.
- level conversion function T (1) shown in FIG. 3 is merely an example, and it goes without saying that any level conversion function can be used depending on the purpose.
- the functions shown in Equations 1 and 2 may be used.
- the edge intensity G (X, y) is calculated for each position on the input image, and the edge intensity G (x, y) is calculated based on the edge intensity G (x, y).
- the input image is filtered by controlling the threshold value E (x, y) of the epsilon filter 12, and calculated according to the output value R (X, y) from the epsilon filter 12.
- the coefficient F (R (x, y)) is multiplied by the input pixel value I (X, y) as shown in Equation 16 to convert the pixel value for each pixel and Compression, so that even if there are multiple different illuminations, it is possible to extract the boundaries of those illuminations appropriately, suppress the occurrence of unnatural image patterns, and provide a subjectively favorable dynamic range. Compression can be realized. That is, by using the epsilon filter 12 to extract the illumination component from the input image and compressing the illumination component, the overall dynamic range is reduced while preserving local contrast. A subjectively favorable reproduced image can be obtained.
- the threshold value E of the epsilon filter 12 is adaptively changed according to the local gradient of the pixel value I (X, y) based on the edge strength G (X, y). Therefore, the illumination boundary can be extracted more accurately than when a linear low-pass filter is used or when an epsilon filter with a fixed threshold is used.
- FIG. 12 shows the configuration of an image processing apparatus according to the second embodiment of the present invention.
- the overall function of the image processing apparatus according to this embodiment is the same as that of the first embodiment, but in this embodiment, the divider 13 and the level converter 14 (FIG. 1) Instead, a point that a coefficient calculator 16 having these functions is provided is different from the first embodiment. That is, in the present embodiment, the dynamic range is compressed based on Equation 16 shown in the first embodiment.
- the coefficient calculator 16 in the present embodiment performs a coefficient calculation function F (1) as shown in FIG. 13 on the output R (X, y) of the epsilon filter 12 to obtain a coefficient C Calculate (X, y).
- the coefficient calculation function F (1) is obtained by Expression 15 using the level conversion function T (1).
- the coefficient C (x, y) calculated by the coefficient calculator 16 is sent to the multiplier 15.
- a signal indicating an input image is directly input, and the coefficient C (X, y) calculated by the calculator 16 is input.
- the threshold value E (x, y) of the epsilon filter 12 is controlled based on the edge strength G (x, y) to perform a filtering process on the input image. Calculated in accordance with the output value R (x, y) from the epsilon filter 12 The coefficient F (R (x, y)) is multiplied by the input pixel value I (x, y) to convert the pixel value for each pixel and compress the dynamic range. The same effect as in the first embodiment can be obtained.
- FIG. 14 shows the configuration of an image processing apparatus according to the third embodiment of the present invention.
- the configuration of the image processing apparatus according to the present embodiment is almost the same as that of the first embodiment (FIG. 1), except that the threshold controller 11 has an output from the edge strength calculator 10.
- the difference is that the pixel value I (x, y) of the input image is directly input in addition to the edge strength G (x, y).
- the threshold E (x, y) used in the subsequent epsilon filter 12 is determined not only by the edge intensity G (x, y) but also by the input image. Is also controlled by the pixel level. More specifically, in the threshold controller 11 A, as the pixel value I (X, y) of the input image increases, the threshold E (x, y) increases, and the edge intensity G (X, The threshold value E (x, y) is controlled so that the threshold value E (X, y) decreases as y) increases.
- Such threshold control can be realized as follows. For example, first, using a pre-set positive coefficient r (r ⁇ 1.0), Equation 31 shows that the larger the pixel value I (X, y) of the input image, the larger the value. Set a temporary threshold E imp (X, y) such that
- the threshold E trap (x, y) is modified by the edge strength G (x, y) to determine the threshold E (x, y) to be actually used.
- the constants Emin and Emax for normalization in Equation 8 are set to 0.0 and 1.0, respectively, and the coefficient G (x, y) according to the edge strength is calculated according to Equation 8. This is added to the provisional threshold value ⁇ ⁇ ( ⁇ , y) as in Equation 32 to obtain the final threshold value E (X, y).
- the threshold value E (x, y) is controlled.
- the threshold E (X, y) of the epsilon fill 1 or 2 plays the role of separating whether the spatial change of the input pixel is due to the change in the illumination component or the change in the reflectivity of the object surface. Even when the change in the reflectance of the surface is small, the change in the pixel value I (X, y) is large if the illumination level itself is large. In other words, how to distinguish between a large change in illumination level and a small change in reflectance under strong illumination is a problem.
- the threshold value E (X, y) of the epsilon filter 12 is set to be larger as the pixel level is larger.
- the effect of the illumination level on the change of (x, y) can be reduced, and the illumination component can be extracted more appropriately.
- the illumination component can be more appropriately extracted in the epsilon filter 12.
- FIG. 15 shows a configuration of an image processing apparatus according to the fourth embodiment of the present invention.
- the configuration of the image processing apparatus according to the present embodiment is almost the same as that of the third embodiment (FIG. 14).
- two types of threshold values Elo (x, y) are provided for each pixel.
- Eup ( ⁇ , y) is different from the third embodiment in that the threshold control of the epsilon filter is performed.
- the threshold controller 11 B in the present embodiment calculates two types of thresholds Elo (x, y) and Eup (x, y) having different sizes for each pixel, and uses them to calculate the thresholds. Performs threshold control of the Psilon filter 12 A. That is, in the threshold controller 1 IB, for example, using two coefficients r 1 and ru (0.0.0 ⁇ r 1, ru ⁇ 1.0) of different sizes set in advance, the equation 33 A Then, as shown in Equation 33B, two types of provisional threshold values E tmplo (x, y) and E trapup (x, y) are calculated. Then, they are output to the epsilon filter 12 A as the first threshold E lo (x, y) and the second Eup (x, y).
- threshold processing is performed by two thresholds E lo (x, y) and Eup (x,) calculated by the threshold controller 11B. Do. More specifically, the processing in the epsilon filter 12A is performed as follows, for example, by the configuration shown in FIG. In the present embodiment, the difference value D (x, y) calculated by the differentiator 20 is also sent to the sign determiner 24 in addition to the absolute value calculator 21.
- the sign determination unit 24 determines the sign of the difference value D (X, y), and sends the result to the switch 27.
- the first comparator 25 uses the first threshold value Elo (x, y) sent from the threshold controller 11 B to calculate the equation (1) as in the first embodiment. Select the signal with. That is, the value AD (d X, dy) calculated by the absolute value calculator 21 is compared with the threshold value Elo (x, y), and the value I (x, y) or One of the neighboring pixel values I (x + dx, y + dy) is selected and output as the value J (dX, dy).
- the second comparator 26 uses the second threshold value Eup (x, y) sent from the threshold controller 11B to calculate the signal by the equation 12 as in the first embodiment. Make a selection. That is, the value AD (dx, dy) calculated by the absolute value calculator 21 is compared with the threshold value Eup (X, y), and according to the result, the value I (x, y) or Select one of the neighboring pixel values I (x + dx, y + dy) and output it as value J (dx, dy).
- the switch 27 selects either the output of the first comparator 25 or the output of the second comparator 26 based on the judgment result of the sign judging unit 24 and sends it to the linear port-pass filter 23. .
- the switch 27 selects the output of the first comparator 25 when, for example, the sign determination result indicates that it is positive. Conversely, if the determination result is negative, the output of the second comparator 26 is selected.
- the threshold is reached.
- Value E up (x, y) is used, and if smaller, Elo (x, y) is used. That is, in the epsilon filter 12 A, it is possible to set different threshold values on the high level side and the low level side.
- the second threshold value Eup (x, y) is set to be larger than the first threshold value Elo (x, y)
- the pixel value change as in the third embodiment is achieved. Therefore, it is possible to reduce the influence of the illumination component included in the image, and to extract the illumination component more appropriately.
- FIG. 17 shows the configuration of an image processing apparatus according to the fifth embodiment of the present invention.
- the configuration of the image processing apparatus according to the present embodiment is similar to that of the first embodiment (FIG. 1).
- a non-linear conversion such as a logarithmic conversion is performed on the pixel level of the input image. The difference is that they are applied.
- a logarithmic converter 17 is installed at the input stage of the circuit, and first, the logarithmic conversion shown in Expression 1 is performed on each pixel value I (X, y) of the input image. It has become.
- a divider 13 for subtracting the illumination component R (x, y) obtained by the epsilon filter 12 sends a compressed illumination component CR (X, y) to a subtractor 18.
- a multiplier 15 for multiplying the non-illumination component S (X, y) has been replaced with an adder 19.
- the pixel value IL (X, y) of the input image after logarithmic conversion by the logarithmic converter 17 is sent to the edge strength calculator 10, the epsilon filter 12, and the subtractor 18.
- the edge strength calculator 10 and the epsilon filter 12 perform the same processing as in the first embodiment based on the pixel value IL (X, y).
- the subtracter 18 subtracts the illumination component R (x, y) obtained by the epsilon filter 12 from each pixel value IL (x, y), thereby obtaining the illumination component from the input image.
- the non-illumination component S (x, y) obtained as a result is output to the adder 19.
- the adder 19 adds the corrected illumination component CR (x, y) to the non-illumination component S (x, y), The image signal is restored, and the resulting image signal O (x, y) is output to a transmission line, a storage device, a display device, or the like (not shown), as in the first embodiment.
- the logarithmic transformation used in the present embodiment itself has the effect of compressing the dynamic range, and the pixel value I (X, y) is compressed more as the input level becomes higher.
- the effect of the illumination level on the spatial change of the pixel value is reduced, and the illumination component can be more appropriately extracted. It becomes possible.
- the threshold value used in the epsilon filter is controlled based on the edge strength calculated for each position on the input image.
- a coefficient for converting the pixel value is calculated according to the output value from the epsilon filter whose threshold value is controlled, and the pixel value is converted for each pixel by the calculated coefficient.
- E (x, y) (Emax-Emin) + EminGmin ⁇ G (x, y) ⁇ Gmax
- R (x, y) ⁇ 7j- ⁇ J (dx, dy)
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Image Processing (AREA)
- Picture Signal Circuits (AREA)
- Facsimile Image Signal Circuits (AREA)
Description
Claims
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US10/362,201 US7127122B2 (en) | 2001-06-20 | 2002-06-19 | Image processing method and image processing apparatus |
KR1020037002412A KR100895207B1 (ko) | 2001-06-20 | 2002-06-19 | 화상 처리 방법 및 장치 |
EP02743646A EP1404120A4 (en) | 2001-06-20 | 2002-06-19 | METHOD AND DEVICE FOR PROCESSING IMAGES |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2001187146A JP4649781B2 (ja) | 2001-06-20 | 2001-06-20 | 画像処理方法および装置 |
JP2001-187146 | 2001-06-20 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2003001793A1 true WO2003001793A1 (fr) | 2003-01-03 |
Family
ID=19026482
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2002/006106 WO2003001793A1 (fr) | 2001-06-20 | 2002-06-19 | Procédé et dispositif de traitement d'images |
Country Status (5)
Country | Link |
---|---|
US (1) | US7127122B2 (ja) |
EP (1) | EP1404120A4 (ja) |
JP (1) | JP4649781B2 (ja) |
KR (1) | KR100895207B1 (ja) |
WO (1) | WO2003001793A1 (ja) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7773824B2 (en) | 2003-04-25 | 2010-08-10 | Sony Corporation | Signal processing device and method, recording medium, and program |
US7995839B2 (en) | 2006-04-14 | 2011-08-09 | Fujifilm Corporation | Image processing device and method with distance calculating on color space |
Families Citing this family (58)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4045422B2 (ja) * | 2002-07-23 | 2008-02-13 | 株式会社日立メディコ | 画像処理装置 |
US7599578B2 (en) | 2004-09-17 | 2009-10-06 | Nikon Corporation | Apparatus, program, and method for image tone transformation, and electronic camera |
WO2006040960A1 (ja) * | 2004-10-08 | 2006-04-20 | Matsushita Electric Industrial Co., Ltd. | 画像処理装置および画像処理プログラム |
JP4127262B2 (ja) * | 2004-11-29 | 2008-07-30 | ソニー株式会社 | 画像処理装置及び画像処理方法 |
US7525579B2 (en) * | 2004-12-27 | 2009-04-28 | Konica Minolta Holdings, Inc. | Image sensing apparatus and image processing method for use therein |
US7961966B2 (en) * | 2005-01-04 | 2011-06-14 | Etron Technology, Inc. | Digitized image stabilization using energy analysis method |
JP4538358B2 (ja) * | 2005-03-30 | 2010-09-08 | パイオニア株式会社 | 画像処理装置 |
US8014034B2 (en) * | 2005-04-13 | 2011-09-06 | Acd Systems International Inc. | Image contrast enhancement |
JP2007011926A (ja) * | 2005-07-04 | 2007-01-18 | Fujifilm Holdings Corp | 画像処理方法および装置並びにプログラム |
JP4736598B2 (ja) * | 2005-07-27 | 2011-07-27 | ソニー株式会社 | ノイズ除去回路及びノイズ除去方法 |
US7791656B2 (en) * | 2005-08-16 | 2010-09-07 | Konica Minolta Holdings, Inc. | Image sensing apparatus and image processing method |
JP4736939B2 (ja) * | 2005-08-16 | 2011-07-27 | コニカミノルタホールディングス株式会社 | 撮像装置及び画像処理方法 |
JP4135737B2 (ja) | 2005-08-30 | 2008-08-20 | ブラザー工業株式会社 | 画像処理プログラム |
US7590303B2 (en) * | 2005-09-29 | 2009-09-15 | Samsung Electronics Co., Ltd. | Image enhancement method using local illumination correction |
JP4157592B2 (ja) * | 2005-10-12 | 2008-10-01 | 松下電器産業株式会社 | 視覚処理装置、表示装置、視覚処理方法、プログラムおよび集積回路 |
US7755670B2 (en) * | 2005-10-24 | 2010-07-13 | Nikon Corporation | Tone-conversion device for image, program, electronic camera, and tone-conversion method |
US7995058B2 (en) * | 2006-01-27 | 2011-08-09 | Tandent Vision Science, Inc. | Method and system for identifying illumination fields in an image |
US7843493B2 (en) * | 2006-01-31 | 2010-11-30 | Konica Minolta Holdings, Inc. | Image sensing apparatus and image processing method |
CN101427559B (zh) | 2006-04-19 | 2011-11-30 | 松下电器产业株式会社 | 视觉处理装置、视觉处理方法、显示装置和集成电路 |
WO2007125732A1 (ja) | 2006-04-28 | 2007-11-08 | Panasonic Corporation | 視覚処理装置、視覚処理方法、プログラム、記録媒体、表示装置および集積回路 |
EP2053557B1 (en) * | 2006-08-17 | 2013-11-06 | Fujitsu Limited | Image processing device, image processing program, and image processing method |
JP4208909B2 (ja) | 2006-08-24 | 2009-01-14 | 株式会社東芝 | 画像処理装置と撮影装置 |
JP4859632B2 (ja) * | 2006-11-15 | 2012-01-25 | 富士通セミコンダクター株式会社 | 画像処理装置及び画像処理方法 |
JP4288623B2 (ja) * | 2007-01-18 | 2009-07-01 | ソニー株式会社 | 撮像装置、ノイズ除去装置、ノイズ除去方法、ノイズ除去方法のプログラム及びノイズ除去方法のプログラムを記録した記録媒体 |
US8396290B1 (en) | 2007-03-09 | 2013-03-12 | Pixar | Saturation varying color space |
US8942476B1 (en) | 2007-03-09 | 2015-01-27 | Pixar | Saturation varying and lighting independent color color control for computer graphics |
JP4746575B2 (ja) * | 2007-03-16 | 2011-08-10 | 株式会社東芝 | 画像処理装置及びその方法 |
JP4772721B2 (ja) * | 2007-03-26 | 2011-09-14 | 株式会社東芝 | 画像処理装置及びその方法 |
JP4527750B2 (ja) * | 2007-05-30 | 2010-08-18 | 三菱電機株式会社 | 画像処理装置及び方法並びに画像表示装置 |
KR100901354B1 (ko) * | 2007-05-31 | 2009-06-05 | 주식회사 코아로직 | 영상의 에지 보정장치 및 그 방법 |
EP2179589A4 (en) * | 2007-07-20 | 2010-12-01 | Fujifilm Corp | IMAGE PROCESSING DEVICE, IMAGE PROCESSING AND PROGRAM |
US8330768B2 (en) * | 2007-07-27 | 2012-12-11 | Sharp Laboratories Of America, Inc. | Apparatus and method for rendering high dynamic range images for standard dynamic range display |
JP5012333B2 (ja) * | 2007-08-30 | 2012-08-29 | コニカミノルタアドバンストレイヤー株式会社 | 画像処理装置および画像処理方法ならびに撮像装置 |
US8300974B2 (en) | 2007-11-12 | 2012-10-30 | Mitsubishi Electric Corporation | Image processing device, image display device, and image processing method therefor |
JP4502055B2 (ja) * | 2007-11-12 | 2010-07-14 | 三菱電機株式会社 | 画像処理装置、画像表示装置、および画像処理方法 |
EP2216988B1 (en) * | 2007-12-04 | 2013-02-13 | Sony Corporation | Image processing device and method, program, and recording medium |
US7995854B2 (en) * | 2008-03-28 | 2011-08-09 | Tandent Vision Science, Inc. | System and method for identifying complex tokens in an image |
US8139883B2 (en) * | 2008-07-29 | 2012-03-20 | Sony Corporation | System and method for image and video encoding artifacts reduction and quality improvement |
US8103120B2 (en) | 2008-09-22 | 2012-01-24 | Solomon Systech Limited | Method and apparatus of local contrast enhancement |
EP2166509A1 (en) * | 2008-09-22 | 2010-03-24 | Solomon Systech Limited | Method and apparatus of local contrast enhancement |
KR100975221B1 (ko) * | 2008-11-05 | 2010-08-10 | 매그나칩 반도체 유한회사 | 샤프니스 보정장치 및 그 방법 |
GB2468304A (en) * | 2009-03-03 | 2010-09-08 | Sharp Kk | Video image processing method that smooths the jagged diagonal edges created by epsilon filtering |
JP5233757B2 (ja) * | 2009-03-09 | 2013-07-10 | 三菱電機株式会社 | 画像処理装置および画像表示装置 |
JP4585602B1 (ja) | 2009-09-18 | 2010-11-24 | 株式会社東芝 | 画像処理装置、表示装置及び画像処理方法 |
JP2011090569A (ja) * | 2009-10-23 | 2011-05-06 | Sony Corp | 画像処理装置および画像処理方法 |
CN101930599B (zh) * | 2010-08-24 | 2013-02-13 | 黄伟萍 | 一种医学图像增强方法及系统 |
DE102010039807A1 (de) * | 2010-08-26 | 2012-03-01 | Siemens Aktiengesellschaft | Vorrichtung und Verfahren zur Filterung in Röntgenbildern |
JP5470415B2 (ja) * | 2012-03-30 | 2014-04-16 | Eizo株式会社 | イプシロンフィルタの閾値決定方法およびローパスフィルタの係数決定方法 |
JP5267704B2 (ja) * | 2012-06-06 | 2013-08-21 | コニカミノルタアドバンストレイヤー株式会社 | 画像処理装置および画像処理方法ならびに撮像装置 |
JP2013017240A (ja) * | 2012-10-10 | 2013-01-24 | Konica Minolta Advanced Layers Inc | 画像処理装置、画像処理方法、及び撮像装置 |
JP6625165B2 (ja) | 2013-02-14 | 2019-12-25 | キヤノン株式会社 | 画像処理装置、画像処理方法、およびプログラム |
JP6362333B2 (ja) * | 2013-02-14 | 2018-07-25 | キヤノン株式会社 | 画像処理装置、画像処理方法、およびプログラム |
JP6303816B2 (ja) * | 2014-05-28 | 2018-04-04 | 富士ゼロックス株式会社 | 画像処理装置及びプログラム |
JP2016086347A (ja) | 2014-10-28 | 2016-05-19 | 三星ディスプレイ株式會社Samsung Display Co.,Ltd. | 画像処理装置、画像処理方法、及びプログラム |
JP6548403B2 (ja) | 2015-02-24 | 2019-07-24 | 三星ディスプレイ株式會社Samsung Display Co.,Ltd. | 画像処理装置、画像処理方法、及びプログラム |
JP6548907B2 (ja) | 2015-02-24 | 2019-07-24 | 三星ディスプレイ株式會社Samsung Display Co.,Ltd. | 画像処理装置、画像処理方法、及びプログラム |
KR102495754B1 (ko) | 2016-10-25 | 2023-02-02 | 한화테크윈 주식회사 | 광역 영상의 동적 범위를 압축하는 영상 처리 장치 및 방법 |
JP6742231B2 (ja) * | 2016-12-09 | 2020-08-19 | キヤノン株式会社 | 画像処理装置及び方法、及び撮像装置 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS62155679A (ja) * | 1985-12-27 | 1987-07-10 | Olympus Optical Co Ltd | 画像信号処理回路 |
JPH07212759A (ja) * | 1994-01-27 | 1995-08-11 | Sharp Corp | 画像信号復号器 |
JPH0865673A (ja) * | 1994-08-22 | 1996-03-08 | Kyocera Corp | 画像信号復号装置 |
JPH09247674A (ja) * | 1996-03-11 | 1997-09-19 | Matsushita Electric Ind Co Ltd | 輪郭および輪郭近傍歪み除去装置 |
Family Cites Families (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4527265A (en) * | 1983-06-30 | 1985-07-02 | Staar S. A. | Position memory for laser reproduction unit |
BE902324A (fr) * | 1985-04-30 | 1985-08-16 | Staar Sa | Appareil assurant la reproduction et/ou l'enregistrement selectif et/ou successif de supports d'informations. |
JPS62214576A (ja) * | 1986-03-14 | 1987-09-21 | Pioneer Electronic Corp | ピツクアツプの位置制御方法 |
US4878129A (en) * | 1986-08-27 | 1989-10-31 | Pioneer Electronic Corporation | Method for reproduction of stored video signals with disk reproducing apparatus |
JP2609542B2 (ja) * | 1988-10-04 | 1997-05-14 | パイオニア株式会社 | ディスクプレーヤ |
CA2020059C (en) * | 1989-06-30 | 1995-05-23 | Shigemi Maeda | Modular ceramic igniter |
JPH04254959A (ja) * | 1991-02-05 | 1992-09-10 | Sony Corp | 記録媒体の再生装置 |
JP2970008B2 (ja) * | 1991-02-22 | 1999-11-02 | ソニー株式会社 | ディスク状記録媒体再生装置 |
JP3396894B2 (ja) * | 1992-05-14 | 2003-04-14 | ソニー株式会社 | 音響再生装置 |
JPH07326175A (ja) * | 1994-05-30 | 1995-12-12 | Sony Corp | 記録又は再生装置 |
JPH08147702A (ja) * | 1994-11-11 | 1996-06-07 | Mitsumi Electric Co Ltd | 光ディスク書き込み方法 |
JPH08241565A (ja) * | 1995-03-03 | 1996-09-17 | Toshiba Corp | ポータブルコンピュータ |
JPH10508138A (ja) * | 1995-08-22 | 1998-08-04 | フィリップス エレクトロニクス ネムローゼ フェンノートシャップ | 媒体のオーディオ信号及び/又はビデオ信号読取り用プレーヤ |
US6463601B1 (en) * | 1995-10-12 | 2002-10-15 | Randy E. Fetty | Self-contained waste and refuse disposal unit |
KR100279147B1 (ko) * | 1996-07-27 | 2001-01-15 | 구자홍 | 에지강조 오차확산법의 에지검출 방법 |
KR100283948B1 (ko) * | 1997-09-12 | 2001-05-02 | 윤종용 | 디스크플레이어제어장치및그방법 |
JP3913356B2 (ja) * | 1998-03-17 | 2007-05-09 | 富士フイルム株式会社 | 画像処理方法 |
JP3873523B2 (ja) * | 1999-05-21 | 2007-01-24 | ソニー株式会社 | 再生装置 |
JP3515019B2 (ja) * | 1999-07-26 | 2004-04-05 | 日本電信電話株式会社 | 画像補正方法及びその装置及びその方法を記録した記録媒体及びその装置を組み込んだ画像撮影装置及びその装置を組み込んだ画像表示装置 |
EP1137267A3 (en) * | 2000-02-07 | 2005-09-14 | Sony Corporation | Device and method for image processing |
JP4556276B2 (ja) * | 2000-03-23 | 2010-10-06 | ソニー株式会社 | 画像処理回路及び画像処理方法 |
US6396777B1 (en) * | 2000-05-30 | 2002-05-28 | Delphi Technologies, Inc. | Compact disk player instant-on feature |
-
2001
- 2001-06-20 JP JP2001187146A patent/JP4649781B2/ja not_active Expired - Fee Related
-
2002
- 2002-06-19 EP EP02743646A patent/EP1404120A4/en not_active Withdrawn
- 2002-06-19 KR KR1020037002412A patent/KR100895207B1/ko not_active IP Right Cessation
- 2002-06-19 WO PCT/JP2002/006106 patent/WO2003001793A1/ja active Application Filing
- 2002-06-19 US US10/362,201 patent/US7127122B2/en not_active Expired - Lifetime
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS62155679A (ja) * | 1985-12-27 | 1987-07-10 | Olympus Optical Co Ltd | 画像信号処理回路 |
JPH07212759A (ja) * | 1994-01-27 | 1995-08-11 | Sharp Corp | 画像信号復号器 |
JPH0865673A (ja) * | 1994-08-22 | 1996-03-08 | Kyocera Corp | 画像信号復号装置 |
JPH09247674A (ja) * | 1996-03-11 | 1997-09-19 | Matsushita Electric Ind Co Ltd | 輪郭および輪郭近傍歪み除去装置 |
Non-Patent Citations (1)
Title |
---|
See also references of EP1404120A4 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7773824B2 (en) | 2003-04-25 | 2010-08-10 | Sony Corporation | Signal processing device and method, recording medium, and program |
US7995839B2 (en) | 2006-04-14 | 2011-08-09 | Fujifilm Corporation | Image processing device and method with distance calculating on color space |
Also Published As
Publication number | Publication date |
---|---|
JP4649781B2 (ja) | 2011-03-16 |
US7127122B2 (en) | 2006-10-24 |
KR20030025294A (ko) | 2003-03-28 |
EP1404120A4 (en) | 2006-10-04 |
JP2003008935A (ja) | 2003-01-10 |
KR100895207B1 (ko) | 2009-05-04 |
US20030156761A1 (en) | 2003-08-21 |
EP1404120A1 (en) | 2004-03-31 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2003001793A1 (fr) | Procédé et dispositif de traitement d'images | |
US8654221B2 (en) | Image processing device and method, and program | |
US7570390B2 (en) | Image processing device and method | |
EP1137258B1 (en) | Image processing circuit and method for processing image | |
JP4210577B2 (ja) | 選択的空間フィルタを使用するディジタル画像の階調及び空間特性の向上方法 | |
EP1111907B1 (en) | A method for enhancing a digital image with noise-dependent control of texture | |
EP2833317B1 (en) | Image display device and/or method therefor | |
KR101031153B1 (ko) | 신호 처리 장치 및 방법 | |
EP1111906A2 (en) | A method for enhancing the edge contrast of a digital image independently from the texture | |
JP4874184B2 (ja) | 画像処理装置及び画像処理方法 | |
JP4497160B2 (ja) | 画像処理装置および方法、並びにプログラム | |
WO2001041425A1 (fr) | Dispositif de correction de contour | |
US7916967B2 (en) | Image processing apparatus and image processing method | |
US7421146B2 (en) | Apparatus and method of processing shoot artifacts of image signals | |
US7894686B2 (en) | Adaptive video enhancement gain control | |
JP5295854B2 (ja) | 画像処理装置及び画像処理プログラム | |
KR101007840B1 (ko) | 신호 처리 장치 및 방법, 기록 매체 | |
JPH0944648A (ja) | 画像処理装置及び画像処理方法 | |
JP2003009178A (ja) | 画像処理方法および装置 | |
JP2004007292A (ja) | モスキート歪み低減装置およびモスキート歪み低減方法 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): KR US |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2002743646 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1020037002412 Country of ref document: KR |
|
WWE | Wipo information: entry into national phase |
Ref document number: 10362201 Country of ref document: US |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWP | Wipo information: published in national office |
Ref document number: 1020037002412 Country of ref document: KR |
|
WWP | Wipo information: published in national office |
Ref document number: 2002743646 Country of ref document: EP |