JP4648242B2 - Image processing device - Google Patents

Description

  The present invention relates to an image processing apparatus having a contour correction function.

  Conventionally, various techniques for correcting the contour of an image are known. For example, in Patent Document 1, the level of the contour component is made constant based on the peak level of the high-frequency component (contour component) extracted from the video signal, and the gain of the high-frequency component is controlled based on the accumulated value of the contour component in a predetermined period. I am going to do that. However, since the technique described in Patent Document 1 integrates contour components in a predetermined period, the same integrated value is used for an image having a small number of contour components and a large number of contour components having a small level. May be calculated, and the same contour correction may be performed on the two types of images. For this reason, the technique described in Patent Document 1 may not be able to perform appropriate contour correction according to the image.

As a technique for solving the above problems, a technique described in Patent Document 2 is disclosed. In Patent Document 2, the contour component included in the input signal is distributed for each of a plurality of predetermined level ranges, and a histogram is created based on the appearance frequency. Then, the contour correction in the video of one field or one frame period is performed using the created histogram.
Japanese Patent Laid-Open No. 11-177852 JP 2006-13735 A

  However, when comparing the contour component in the vertical direction with the contour component in the horizontal direction in the contour of the image, the frequency of the contour component in the horizontal direction is generally higher than the frequency of the contour component in the vertical direction. It is difficult to see. For this reason, when the gain of the contour component is uniformly controlled regardless of the direction of the contour as in the invention described in Patent Document 2, the vertical contour is emphasized compared to the horizontal contour in the corrected image. As a result, there is a problem that the outline looks unnatural as a whole, such as the fact that it becomes difficult to see the outline in the horizontal direction.

  Further, the gain control amount for correcting the contour cannot be calculated until the histogram of the contour component is created, and the contour correction must be performed in units of one field or one frame period. For this reason, it is necessary to delay the video signal by a unit of one field or one frame period, which causes a problem that the delay circuit becomes large.

  The present invention has been made in view of the above problems, and an object thereof is to perform contour correction suitable for an image by performing gain control in accordance with the direction of the contour.

  In order to solve the above problems, in the present invention, a pixel matrix information acquisition unit that acquires pixel matrix information, a contour direction information acquisition unit that acquires contour direction information that is information indicating a contour direction from the acquired pixel matrix information, The present invention proposes an image processing apparatus comprising: a contour pixel correction unit that corrects a contour pixel of the pixel matrix information using the acquired contour direction information to obtain new pixel matrix information with enhanced contours.

  The contour direction information acquisition unit includes coefficient matrix holding means for holding a coefficient matrix sensitive to the contour direction of the pixel matrix information for the contour direction pattern type, and sensitivity of the held coefficient matrix to the acquired pixel matrix information. The image processing apparatus may include a sensitivity result acquisition unit that acquires the result and a contour direction information acquisition unit that acquires the contour direction information corresponding to the pixel matrix information according to the sensitivity result.

  The contour pixel correction unit acquires vertical contour correction information indicating a vertical contour correction amount and horizontal contour correction information indicating a horizontal contour correction amount based on the acquired contour direction information. Information acquisition means, corrected vertical contour pixel information acquisition means for correcting vertical contour pixel information of the pixel matrix information by using vertical contour correction information and acquiring corrected vertical contour pixel information, and horizontal contour correction information Corrected horizontal contour pixel information acquisition means for correcting the horizontal contour pixel portion of the pixel matrix information and acquiring corrected horizontal contour pixel information, acquired corrected vertical contour pixel information, and corrected Horizontal contour pixel information, and corrected contour pixel information generating means for generating corrected contour pixel information corresponding to the contour pixel of new pixel matrix information from the horizontal contour pixel information It may be.

  In addition, the contour pixel correction unit acquires a contour enhancement coefficient acquisition unit that acquires a contour enhancement coefficient that is a coefficient that determines the enhancement degree in the new pixel matrix of the corrected contour pixel information according to the sensitivity intensity of the acquired sensitivity result. May be an image processing apparatus.

  Furthermore, in order to solve the above problems, in the present invention, a pixel matrix information acquisition step for acquiring pixel matrix information, and contour direction information acquisition for acquiring contour direction information that is information indicating the contour direction from the acquired pixel matrix information. The present invention proposes a contour correction method comprising: a step and a contour pixel correction step that corrects a contour pixel of the pixel matrix information using the acquired contour direction information to obtain new pixel matrix information in which the contour is emphasized.

  The contour direction information acquisition step includes a sensitivity result acquisition substep for acquiring a sensitivity result of a coefficient matrix for a contour direction pattern type sensitive to the contour direction of the pixel matrix information and the acquired pixel matrix information, and the sensitivity The contour correction method may include a contour direction information acquisition substep for acquiring contour direction information corresponding to the pixel matrix information according to the result.

  The contour pixel correcting step acquires vertical contour correction information indicating a vertical contour correction amount and horizontal contour correction information indicating a horizontal contour correction amount based on the acquired contour direction information. An information acquisition sub-step; a corrected vertical contour pixel information acquisition sub-step for correcting corrected vertical contour pixel information of the pixel matrix information by using vertical contour correction information; A corrected horizontal contour pixel information acquisition sub-step for acquiring corrected horizontal contour pixel information by correcting the horizontal contour pixel portion of the pixel matrix information using correction information; and the acquired corrected vertical contour pixel information; , Corrected horizontal pixel information, and corrected contour pixel information for generating corrected contour pixel information corresponding to the contour pixel of the new pixel matrix information And forming sub-step may be a contour correction method having.

  Further, the contour pixel correction step acquires a contour emphasis coefficient that is a coefficient that determines an emphasis degree in the new pixel matrix information of the corrected contour pixel information according to the sensitivity intensity of the acquired sensitivity result, and has been corrected. A contour correction method having a contour adjustment substep for adjusting contour pixel information may be used.

  In the image processing apparatus of the present invention having the above-described configuration, it is possible to acquire contour direction information from pixel matrix information and perform contour correction according to the contour direction information. For this reason, the vertical contour component and the horizontal contour component can be separately gain controlled, and the contour of the corrected image becomes unnatural by increasing the gain control amount of the horizontal contour component. This can be prevented. In the present invention, since the delay for acquiring the pixel matrix information is small, the delay circuit can be made small.

The best mode for carrying out the present invention will be described in detail with reference to the drawings. In addition, this invention is not limited to these embodiment at all, and can be implemented with various aspects in the range which does not deviate from the summary. In addition, the relationship between the following embodiment and a claim is as follows. The first embodiment will mainly describe claims 1 and 4 . The second embodiment will mainly describe claims 1, 2, 4, 5, and the like. The third embodiment will mainly describe claims 3 and 6 .

  << Embodiment 1 >> (Concept of Embodiment 1) Embodiment 1 will be described. FIG. 1 shows the concept of the image processing apparatus in this embodiment. For example, when the contour a and the contour b shown in FIG. 1A are obtained as image information, when the contour a and the contour b are compared, the contour shown in a is closer to the horizontal than the contour shown in b. I understand that. For this reason, if the gain control amount of the contour component of the image having the contour a is made larger than the gain control amount of the contour component of the image having the contour b, and the pixel matrix information is corrected, a contour that is difficult to see and that is almost horizontal can be clearly seen. The entire image can be corrected so as not to be unnatural.

  (Configuration of First Embodiment) FIG. 2 shows an example of a functional block diagram of an image processing apparatus according to the present embodiment. The “image processing apparatus” (0200) according to the present embodiment includes a “pixel matrix information acquisition unit” (0201), a “contour direction information acquisition unit” (0202), a “contour pixel correction unit” (0203), Consists of.

  (Description of Configuration of Embodiment 1) An “image processing apparatus” (0200) is an apparatus having a function of correcting the contour of an image such as a still image or a moving image. The image processing apparatus may be a circuit device that performs only contour correction or video correction, or may be a television receiver or a video reproduction device that incorporates a contour correction circuit or a video correction circuit.

  The “pixel matrix information acquisition unit” (0201) is configured to acquire pixel matrix information. The pixel matrix is a collection of pixels represented in an n × m matrix. Note that n = m and n is desirably an odd number of 3 or more, but is not limited thereto. Throughout the embodiment, for convenience of explanation, the case of n = m and n = 3 is shown. For example, when the contour shown in FIG. 1A is enlarged to each pixel level, it can be shown as in FIG. FIG. 1B shows the outline a and the outline b in a 3 × 3 pixel matrix. The pixel matrix information is a numerical representation of luminance information and color difference information (RGB) of the pixel matrix. This is illustrated in FIG. In the pixel matrix information acquisition unit (0201), the image is digitized in this way to acquire n × m pixel matrix information.

  The “contour direction information acquisition unit” (0202) is configured to acquire contour direction information which is information indicating the contour direction from the pixel matrix information acquired by the pixel matrix information acquisition unit (0201). The contour direction is the direction in which the contour is directed, and can be represented by the direction of inclination of the contour, the direction of the normal of the contour, the direction of changing from dark color to light color, or the like. If the direction from dark color to light color is the contour direction, the contour directions of contour a and contour b in FIG. 1A can be expressed as directions indicated by arrows. An example of deriving the contour direction will be described in detail with reference to FIG. 3 separately from FIG.

  FIG. 3A shows only the pixels in the contour portion extracted from the pixel matrix of the contour a and the contour b shown in FIG. The pixels of the contour portion extracted at this time are called contour pixels. When the contour pixels are extracted, an approximate straight line connecting these contour pixels is assumed and expressed as a straight contour. This is shown in FIG. A straight line contour is obtained by assuming a straight line that passes through the contour pixels most frequently. If possible, a straight contour may be obtained directly from the original image shown in FIG. Then, an angle θ (0 ≦ θ <π) formed by the reference line (0301) and the straight line outline (0302) is calculated, and the contour direction can be derived as a direction forming a relative angle from the reference line. The method for calculating the contour direction is not limited to such a method, and the method described in the second embodiment may be used.

  The “contour pixel correction unit” (0203) is configured to correct the contour pixel portion of the pixel matrix information using the contour direction information acquired by the contour direction information acquisition unit (0202). As described above, the contour pixel is a pixel in the contour portion. “For the contour pixel of the pixel matrix information” is obtained by extracting only the value of the portion corresponding to the contour pixel from the pixel matrix information. FIG. 1D shows a result obtained by extracting only the contour pixel from the pixel matrix information of FIG.

The contour pixel correction unit (0203) calculates an amplification amount (gain) control value of the contour pixel from the contour direction information acquired by the contour direction information acquisition unit (0202). The gain control value G can be calculated by, for example, the following calculation formula (Formula 1) using the angle θ between the straight line outline and the reference line acquired by the outline direction information acquisition unit (0202). .
(Formula 1)
G = α × (β + | cos θ |) (α and β are constants, and θ is contour direction information)

  That is, the gain control value is increased when the straight line contour extracted from the pixel matrix information is a horizontal contour, and the gain control value is decreased when the straight line contour is a vertical contour. Since the reference line in the contour direction information acquisition unit (0202) is horizontal, the cos function is used in the calculation formula. However, if the reference line is in the vertical direction, sin is used instead of the cos function of the calculation formula. Use a function.

More specifically, if α = 2 and β = 1,
When θ = 0 and π (when the straight line contour is horizontal), G = 2 × (1 + 1) = 4
When θ = π / 2 (when the straight contour is vertical), G = 2 × (1 + 0) = 2
It becomes. Further, the gain control value G changes in the range of 2 <G <4 according to the change in the value of θ.

Here, the gain control values of the pixel matrix information of the contour a and the contour b shown in FIG. 3 are calculated in the same manner as follows. When the angles formed by the contour a and the contour b with the reference lines are θa and θb, respectively, and θa = 3π / 4 and θb = 2π / 3, the gain control values Ga and Gb are calculated as follows.
Ga = 2 × (1+ | cos (3π / 4) |) ≈3.4
Gb = 2 × (1+ | cos (2π / 3) |) = 3.0
The calculated gain control values Ga and Gb correct the contour pixel portion of the pixel matrix information of the contour a and the contour b shown in FIG. This is shown in FIG. As shown in FIG. 1E, each numerical value for the contour pixel of the contour a in FIG. 1D is corrected to 3.4 times by the gain control value Ga. Similarly, each numerical value of the contour pixel of the contour b is corrected to 3.0 times by the gain control value Gb. When the numerical values of the obtained contour pixels are added to the numerical values of the pixel matrix information shown in FIG. 1C, pixel matrix information in which the contour pixels are emphasized is obtained as shown in FIG. Obtainable.

  In the above example, α = 2 is set in order to clarify the enhancement amount for the contour pixel, but α, β are set to values such that the gain control values Ga and Gb are 2% to 7%. It is desirable to set This is because a general contour correction circuit performs correction so that the contour component to be added is 2% to 7% of the input signal level. In the above example, for example, by setting α = 0.02 and β = 1, it is possible to set the gain control values Ga and Gb in which the enhancement amount for the contour pixel satisfies the above range.

  (Specific Example of Embodiment 1) FIG. 4 shows a specific contour correction circuit (0400) for realizing the image processing apparatus according to this embodiment. The contour correction circuit (0400) shown in FIG. 4 delays the input image and outputs the line memory (0401) to the pixels for each line, and the n × n pixel matrix information from the output from the line memory (0401). A pixel matrix information acquisition circuit (0402) to be acquired, a contour extraction circuit (0403) that extracts only the contour pixel of the pixel matrix information, and a contour direction information acquisition circuit (0404) that acquires contour direction information from the pixel matrix information ), A contour control circuit (0405) for calculating the gain control value from the contour direction information, a correction circuit (0406) for correcting the contour pixel of the pixel matrix information by the gain control value calculated by the contour control circuit, A coring circuit (0407) for coring the contour corrected by the circuit, and a pixel from the line memory A delay circuit (0408) to cast, consisting of an adder circuit for adding the pixel output from the coring circuit (0407) and modified contour pixels output from the delay circuit (0408) (0409).

  The functional blocks of the image processing apparatus (0200) shown in FIG. 2 correspond to the contour correction circuit (0400) as follows. The pixel matrix information acquisition unit (0201) corresponds to the line memory (0401) and the pixel matrix acquisition circuit (0402). The contour direction information acquisition unit (0202) corresponds to the contour direction information acquisition circuit (0404). The contour pixel correction unit (0203) corresponds to a contour extraction circuit (0403), a contour control circuit (0405), a correction circuit (0406), a delay circuit (0408), and an addition circuit (0409). Note that a coring circuit (0407) may be included in the contour pixel correction unit. Note that these are merely forms for realizing the image processing apparatus of the present embodiment, and may be realized by other similar circuits as long as the configuration requirements of the present embodiment are provided. The operation of these circuits will be described below.

  When the video signal is input to the line memory (0401), the line memory (0401) delays and outputs the video signal for each line. The output from the line memory (0401) is output to the contour extraction circuit (0403), the pixel matrix information acquisition circuit (0402), and the delay circuit (0408). If the pixel matrix information is information represented by a 3 × 3 matrix, as shown in FIG. 4, the contour extraction circuit (0403) and the pixel matrix information acquisition circuit (0402) have pixel information for three lines. Is entered. The pixel matrix information acquisition circuit (0402) acquires pixel matrix information from the luminance information and color difference information of the input pixel information, and outputs this. The contour extraction circuit (0403) extracts only the contour pixels from the pixel matrix information. The contour extraction circuit (0402) is directly input from the line memory (0401), but may be input from the pixel matrix information acquisition circuit (0403). The contour direction information acquisition circuit (0404) acquires the contour direction information from the angle formed by the straight line contour represented by the input pixel matrix information and the reference line, and outputs it. The contour direction information may be acquired from the contour pixels extracted by the contour extraction circuit (0402) (indicated by a broken line in FIG. 4) or may be acquired by another method. . When the contour direction information is input from the contour direction information acquisition circuit (0404), the contour control circuit (0405) calculates and outputs a gain control value for correcting the contour. The correction circuit (0406) multiplies the contour pixel of the pixel matrix information output from the contour extraction circuit (0403) by the gain control value output from the contour control circuit (0405) to thereby obtain the contour pixel of the pixel matrix information. Correct the minutes.

  The corrected pixel matrix information corresponding to the contour pixel is input to the coring circuit (0407). As shown in FIG. 5, the coring circuit (0407) is provided with a dead zone δ (0501), and an output signal corresponding to an input signal in the range of the dead zone δ is set to zero. In FIG. 5, the horizontal axis indicates the contour intensity of the contour pixel, and the vertical axis indicates the output of the contour pixel. Since a contour pixel that does not have a certain contour strength is not likely to form a contour, the contour pixel in such pixel matrix information should not be emphasized. For this reason, the noise component which does not form an outline can be removed in the coring circuit. The contour strength can be indicated by using the sum of values such as digitized luminance information or the maximum value, and a specific calculation method will be described in detail in the third embodiment. By adding the contour pixels of the pixel matrix information output from the coring circuit (0407) to the original image delayed by the delay circuit (0408), it is possible to output a video signal in which the contour portion is corrected. . The delay circuit (0408) is also directly input from the line memory (0401) similarly to the contour extraction circuit (0402), but may be input from the pixel matrix information acquisition circuit (0403).

  (Processing Flow of Embodiment 1) The basic processing flow of the contour correction method in this embodiment is shown in the processing flow of FIG. In the contour correction method according to the present embodiment, first, n × n pixel matrix information digitized according to luminance and color difference is acquired from a video signal (S0601: pixel matrix information acquisition step). Next, contour direction information indicating the direction of the contour is acquired from the acquired pixel matrix information (S0602: contour direction information acquisition step). When the contour direction information is acquired, a gain control value for correcting the contour pixel portion of the acquired pixel matrix information is calculated according to the contour direction information, and the contour pixel of the pixel matrix information acquired based on the calculated gain control value The minute is corrected (S0603: contour pixel correcting step). In this way, it is possible to generate an image modified so as to emphasize the contour pixels.

  (Effect of Embodiment 1) In the image processing apparatus of this embodiment having the above-described configuration, it is possible to acquire contour direction information from pixel matrix information and perform contour correction according to the contour direction information. For this reason, it is possible to prevent the contour of the corrected image from becoming unnatural by increasing the gain control amount of the contour component whose contour is almost horizontal.

  << Embodiment 2 >> (Concept of Embodiment 2) An image processing apparatus according to this embodiment is based on the image processing apparatus according to Embodiment 1, and is characterized by a contour direction information acquisition method in a contour direction information acquisition unit. The contour direction information acquisition unit of the present embodiment prepares a plurality of predetermined contour direction patterns and determines the contour direction by determining which contour direction pattern the contour of the video signal is sensitive to. Furthermore, the gain control is performed by dividing the contour pixel into the horizontal direction and the vertical direction based on the acquired contour direction information.

  (Configuration of Second Embodiment) FIG. 7 shows an example of a functional block diagram of an image processing apparatus according to this embodiment. The “image processing apparatus” (0700) according to the present embodiment includes a “pixel matrix information acquisition unit” (0701), a “contour direction information acquisition unit” (0702), a “contour pixel correction unit” (0703), The contour direction information acquisition unit (0702) includes “coefficient matrix holding means” (0704), “sensitivity result acquisition means” (0705), and “contour direction information acquisition means” (0706). The pixel correction unit (0703) includes “vertical horizontal contour correction information acquisition unit” (0707), “corrected vertical contour pixel information acquisition unit” (0708), and “corrected horizontal contour pixel information acquisition unit” (0709). And “corrected contour pixel information generating means” (0710). Since the “pixel matrix information acquisition unit” (0701) has already been described in the first embodiment, the description thereof is omitted.

  (Explanation of Configuration of Embodiment 2) “Coefficient matrix holding means” (0704) is means for holding a coefficient matrix corresponding to the outline direction of pixel matrix information for the contour direction pattern type. The “contour direction pattern” is a basic contour shape for classifying the contour direction. For example, FIG. 8A illustrates contour direction patterns A to H representing eight contour directions. The contour direction patterns shown in FIG. 8A are: A: upward direction (0801), B: upper left direction (0802), C: left direction (0803), D: lower left direction (0804), E: lower direction (0805). ), F: Lower right direction (0806), G: Right direction (0807), H: Upper right direction (0808). As the contour direction pattern, a plurality of finer contour direction patterns may be prepared. The “coefficient matrix that is sensitive to the contour direction of the pixel matrix information” is a matrix-like coefficient determined so that a unique sensitivity result can be acquired according to the contour direction of the pixel matrix information. The same number of coefficient matrix information as the number of types of contour direction patterns is prepared in correspondence with the contour direction patterns. A coefficient matrix corresponding to the contour direction patterns of FIGS. 8A to 8H is shown in FIGS. The coefficient matrix holding means (0704) holds coefficient matrices for these contour direction pattern types. The coefficient matrix corresponds to a predetermined contour direction pattern, and which contour direction pattern the pixel matrix information is close to is reflected in the sensitivity result of the pixel matrix information and the coefficient matrix.

  "Sensitivity result acquisition means" (0705) is a means for acquiring a sensitivity result with the held coefficient matrix for the acquired pixel matrix information. A method of acquiring the sensitivity result by the sensitivity result acquisition means (0705) will be specifically described below with reference to FIGS.

  When the pixel matrix is acquired for the contour c and the contour d shown in FIG. 9A, it is expressed as shown in FIG. 9B. When pixel matrix information is acquired from this pixel matrix, it is represented as shown in FIG. Next, the pixel matrix information is multiplied by a numerical value at a position corresponding to the corresponding coefficient matrix information shown in (2) (A ′) to (H ′) of FIG. For example, the multiplication result of the pixel matrix information of the contour c in FIG. 9 (3) and the coefficient matrix A ′ corresponding to the contour direction pattern A (0801) in FIG. 8 (1) is as shown in FIG. 10 (1). Become. Furthermore, by adding all the numerical values obtained in a matrix obtained by multiplication, it is possible to obtain a sensitivity result between the pixel matrix information (0901) of the contour c and the coefficient matrix A ′ of the contour direction pattern A (0801). it can. When all the multiplication results in FIG. 10A are added, a total of 140 is obtained, and the sensitivity result between the pixel matrix information (0901) and the coefficient matrix A ′ is calculated as 140. Similarly, the sensitivity results of the pixel matrix information (0901) and the coefficient matrix of all the contour direction pattern types (0801 to 0808) are acquired. This is shown in FIG. Since the coefficient matrix is set in correspondence with the contour direction pattern, the contour direction pattern having a higher numerical value of the obtained calculation result becomes a contour direction pattern having a higher sensitivity result. As shown in FIG. 10B, it can be seen that the pixel matrix information of the contour c in FIG. 9 has the highest sensitivity result with the coefficient matrix H ′ of the contour direction pattern H (0808).

  Similarly, for example, the sensitivity results of the pixel matrix information (0902) of the contour d shown in FIG. 9 (3) and the coefficient matrices A ′ to H ′ are calculated as shown in FIG. 11 (1). As shown in FIG. 11 (1), the contour direction patterns having the highest sensitivity results with the respective coefficient matrices are the contour direction pattern G and the contour direction pattern H, which have the same value. Therefore, it can be determined that the contour direction indicates the contour direction between the contour direction pattern G and the contour direction pattern H.

  The “contour direction information acquisition unit” (0706) corresponds to the contour direction information corresponding to the pixel matrix information acquired by the pixel matrix information acquisition unit (0701) in accordance with the sensitivity result acquired by the sensitivity result acquisition unit (0705). It is a means to acquire. For example, when the sensitivity result shown in FIG. 10 (2) is acquired by the sensitivity result acquisition means (0705), the coefficient matrix of the contour direction pattern H (0808) is the most among the eight types of contour direction patterns. It can be seen that the sensitivity results are high. Therefore, the contour direction information acquisition unit (0706) obtains the contour direction information indicating that the pixel matrix information (0901) of the contour c shown in FIG. 9 is a contour having the direction indicated by the contour direction pattern H (0808). Can be acquired. If the angles formed by the straight line contours of the contour direction patterns shown in FIGS. 8A to 8H with respect to the reference line are respectively determined, the contour pixels of the pixel matrix information are obtained by the same method as that described in the first embodiment. Can be corrected by the contour pixel correction unit (0703). Moreover, it is good also as correcting the outline pixel of pixel matrix information with the structure described below.

  The “vertical horizontal contour correction information acquisition means” (0707), based on the acquired contour direction information, obtains vertical contour correction information indicating the vertical contour correction amount and horizontal contour correction information indicating the horizontal contour correction amount. It is a means to acquire. The contour correction amount is an amplification amount (gain control amount) for emphasizing the pixels in the contour portion. The contour pixel correction unit (0703) of the present embodiment is characterized in that the contour pixels are corrected separately in the vertical direction and the horizontal direction according to the vertical contour correction information and the horizontal contour correction information. Below, the calculation method of vertical outline pixel correction information and horizontal outline pixel correction information is shown.

If the horizontal contour pixel correction information is G _h and the vertical contour pixel correction information is G _v , then G _h and G _v are calculated by the following calculation formulas (Formula 2) and (Formula 3).
(Formula 2)
G _h = α × k
(Formula 3)
G _v = β × (1−k)
(Where α and β are constants, and k is a value determined according to contour direction information)
As described above, the horizontal outline and the vertical outline are less visible in the horizontal outline. For this reason, in order to further emphasize the contour in the horizontal direction, it is desirable to satisfy α> β so that the gain control amount is increased. K is a value determined in accordance with the contour direction information in the range of 0 ≦ k ≦ 1. That is, the value is determined according to whether the contour is tilted in the horizontal direction or the vertical direction. The method of determining k is as follows.

For example, the eight contour direction patterns shown in FIG. 8 are further roughly divided into three patterns: a horizontal contour, a vertical contour, and an intermediate contour between the vertical contour and the horizontal contour (an oblique 45-degree contour). can do. Since the contour to be corrected needs to be emphasized and corrected as the contour is closer to the horizontal direction, the value of k is assigned to each of the three broadly divided patterns as follows.
Horizontal contour (contour direction patterns A and E)... K = 1
Vertical contour (contour direction pattern C, G)... K = 0
Diagonal 45 ° contour (contour direction patterns B, D, F, H)... K = 0.5

As described above, when the contour direction information indicated by the contour direction pattern H (0808) is acquired by the contour direction information acquisition unit, k = 0.5 can be determined. If α = 4 and β = 2, the horizontal contour pixel correction information G _h = 2 and the vertical contour pixel correction information G _v = 1 are obtained from (Equation 2) and (Equation 3). And correction information of the vertical outline can be acquired.

In the above example, α = 4 and β = 2 are set for clarity in order to clarify the amount of enhancement for the contour pixels. However, the horizontal contour pixel correction information _Gh and the vertical contour pixel correction information _Gv are added. It is desirable to set α and β to values such that the obtained values are 2% to 7%. In the above example, for example, by setting α = 0.04 and β = 0.02, it is possible to set the gain control values G _h and G _{v in} which the enhancement amount for the contour pixel satisfies the above range. is there.

  Further, the value of k can be similarly obtained for the contour d in FIG. In the pixel matrix information of the contour d, the sensitivity results of the contour direction pattern G and the contour direction pattern H are the same as shown in FIG. Therefore, it can be seen that the contour d is an intermediate contour between the contour direction pattern G and the contour direction pattern H. For this reason, it can be determined that k = 0.25 which is an average value of the value k in the case of the contour direction pattern G and the value k in the case of the contour direction pattern H.

Furthermore, depending on the value of the pixel matrix information, there may be two or more contour direction patterns that indicate a protruding sensitive result that is not the same value. For example, as shown in FIG. 11 (2), although the contour direction pattern G has the highest sensitivity result, there may be a case where the sensitivity result is also prominent with respect to the contour direction pattern H. Assuming that the respective sensitivity results of the contour direction pattern G and the contour direction pattern H at this time are M _G and _MH , the value of k can be calculated by the following (Equation 4).
(Formula 4)
k = Mg × K _G / (Mg + M _H ) + M _H × K _H / (Mg + M _H )
Here, _{K G} represents a value (= 0) k assigned to the contour direction pattern G, _{K H} represents the value of k assigned to the contour direction pattern H (= 0.5). If the sensitivity result 260 of the contour direction pattern G and the sensitivity result 240 of the contour direction pattern H are substituted into the above (Equation 4), k = 0.4 can be determined.

Furthermore, when three or more contour direction patterns with high sensitivity results appear, or when no contour direction pattern with sensitive results appears, it is difficult to specify the direction of the contour, and a clear contour is present. Since there is a high possibility that it does not exist, it is possible to prevent the outline from being corrected. In this case, G _h = G _v = 0 may be used regardless of the value of k. Whether or not the sensitivity result is a prominent value can be determined by setting a predetermined threshold value. For example, if the values of the sensitivity results are all less than 100, it is determined that there is no protruding contour direction pattern. Further, if the difference between the value of the contour direction pattern having the highest sensitivity result and the value of the third highest sensitivity result is within 50, it may be determined that there is no protruding contour direction pattern. These numerical values are merely examples, and need to be set as appropriate according to the contour to be corrected, the type of video, and the like.

The “corrected vertical contour pixel information acquisition unit” (0708) is a unit that acquires the corrected vertical contour pixel information by correcting the vertical contour pixel portion of the pixel matrix information using the vertical contour correction information. “Vertical pixels in the vertical direction” are contour pixels extracted in the vertical direction (column direction) from the pixel matrix. This is a vertical contour pixel. “Vertical contour pixels” is a value obtained by digitizing vertical contour pixels based on luminance information, color difference information, and the like. That is, the portion corresponding to the vertical contour pixel in the pixel matrix information corresponds to the vertical contour pixel. FIG. 12 (1) shows a case where the vertical contour pixels are extracted from the pixel matrix information (0901) of the contour c shown in FIG. As shown in FIG. 12A, three vertical contour pixels a, b, and c can be extracted from the 3 × 3 pixel matrix information. The corrected vertical contour pixel information acquisition unit corrects the vertical contour pixel by multiplying the vertical contour correction information acquired by the vertical horizontal contour correction information acquisition unit (0707) by each numerical value for the vertical contour pixel. . As described above, when the vertical contour correction information G _v = 1, the vertical contour pixels shown in FIG. 12 (1) are gain-controlled by the vertical contour correction information and the vertical contour pixel portions a, b, and c are As shown in FIG. 12 (2), the contour pixels are corrected to a ′, b ′, and c ′. In this case, since G _v = 1, the value is not changed numerically.

“Corrected horizontal contour pixel information acquisition means” (0709) is means for correcting horizontal contour pixel information in the horizontal direction of the pixel matrix information using horizontal contour correction information and acquiring corrected horizontal contour pixel information. “Horizontal pixel in the horizontal direction” is a contour pixel extracted in the horizontal direction (row direction) in the pixel matrix. This is a horizontal contour pixel. The “horizontal contour pixel portion” is a value obtained by digitizing a horizontal contour pixel based on luminance information, color difference information, and the like. That is, the portion corresponding to the horizontal contour pixel in the pixel matrix information corresponds to the horizontal contour pixel. FIG. 13 (1) shows a case where the horizontal contour pixels are extracted from the pixel matrix information (0901) of the contour c shown in FIG. As shown in FIG. 13 (1), three horizontal horizontal outline pixel portions d, e, and f can be extracted from the 3 × 3 pixel matrix information. The corrected horizontal contour pixel information acquisition unit corrects the horizontal contour pixel by multiplying the horizontal contour correction information acquired by the vertical horizontal contour correction information acquisition unit (0707) by each numerical value for the horizontal contour pixel. . As described above, when the horizontal contour correction information G _h = 2, the horizontal contour pixels shown in FIG. 13A are gain-controlled by the horizontal contour correction information, and the horizontal contour pixel portions d, e, and f are As shown in 13 (2), each value is doubled and corrected to contour pixel portions d ′, e ′, and f ′.

  The “corrected contour pixel information generating unit” (0710) is used to correct the corrected vertical contour pixel information acquired by the corrected vertical contour pixel information acquiring unit (0708) and the corrected horizontal contour pixel information acquiring unit (0709). And the corrected horizontal contour pixel information acquired in this way, and corrected contour pixel information that is the contour pixel of the new pixel matrix information. As shown in FIG. 14, the corrected contour pixel information generating means generates the corrected vertical contour pixel information a ′, b ′, c ′ values (1401) shown in FIG. 12 (2) and FIG. The corrected horizontal contour pixel information d ′, e ′, and f ′ values (1402) shown in FIG. 6) are added, and the corrected pixel component is the contour pixel of the pixel matrix information whose gain is controlled in the vertical and horizontal directions. Outline pixel information (1403) is generated. Then, by adding the corrected contour pixel information (1403) to the original pixel matrix information (1404), pixel matrix information (1405) in which the contour pixel is corrected can be acquired.

  (Specific Example of Embodiment 2) FIG. 15 shows a specific contour correction circuit (1500) for realizing the image processing apparatus of this embodiment. The contour correction circuit (1500) shown in FIG. 15 obtains n × n pixel matrix information from the output of the line memory (1501) and the output of the line memory (1501) by delaying the input image and outputting it to the pixels for each line. Pixel matrix information acquisition circuit (1502), a vertical contour extraction circuit (1503) that extracts only vertical contour pixels, a horizontal contour extraction circuit (1504) that extracts only horizontal contour pixels, and pixel matrix information The contour direction information acquisition circuit (1505) for acquiring the contour direction information from the contour, the contour control circuit (1506) for calculating the vertical contour correction information and the horizontal contour correction information from the contour direction information, and the vertical calculated by the contour control circuit A vertical contour correction circuit (1507) for correcting the vertical contour pixel of the pixel matrix information based on the contour correction information; A horizontal contour correction circuit (1508), a vertical contour correction circuit (1507), and a horizontal contour correction circuit (1508) for correcting the horizontal contour pixels of the pixel matrix information based on the horizontal contour correction information calculated by the control circuit. A contour addition circuit (1509) for adding the corrected vertical contour pixel and the horizontal contour pixel, a coring circuit (1510) for coring the contour corrected by the contour addition circuit (1509), and a pixel from the line memory A delay circuit (1511) for delaying the image and a summing circuit (1512) for adding the corrected contour pixel output from the coring circuit (1510) and the original image output from the delay circuit (1511). . The contour direction information acquisition circuit (1505) is shown in more detail in FIG. As shown in FIG. 15A, the contour direction information acquisition circuit (1505) includes a memory (1505a) for recording a coefficient matrix, a sensitivity result calculation circuit (1505b) for calculating pixel matrix information and a sensitivity result of the coefficient matrix, and the like. Consists of

  The functional blocks of the image processing apparatus (0700) shown in FIG. 7 correspond to the contour correction circuit (1500) as follows. The pixel matrix information acquisition unit (0701) corresponds to the line memory (1501) and the pixel matrix acquisition circuit (1502). The contour direction information acquisition unit (0702) corresponds to the contour direction information acquisition circuit (1505). The contour pixel correction unit (0703) includes a vertical contour extraction circuit (1503), a horizontal contour extraction circuit (1504), a contour control circuit (1506), a vertical contour correction circuit (1507), a horizontal contour correction circuit (1508), and a contour addition. It corresponds to a circuit (1509), a delay circuit (0511), and an adder circuit (0512). The contour pixel correction unit may include a coring circuit (0510). Note that these are merely forms for realizing the image processing apparatus of the present embodiment, and may be realized by other similar circuits as long as the configuration requirements of the present embodiment are provided. The operation of these circuits will be described below.

  When the video signal is input to the line memory (1501), the line memory (1501) delays and outputs the video signal for each line. The output from the line memory (1501) is output to a vertical contour extraction circuit (1503), a horizontal contour extraction circuit (1504), a pixel matrix information acquisition circuit (1502), and a delay circuit (1511). If the pixel matrix information is information represented by a 3 × 3 matrix, as shown in FIG. 15, the vertical matrix extraction circuit (1503) and the pixel matrix information acquisition circuit (0403) are delayed one line at a time. Pixel information for three lines is input. The pixel matrix information acquisition circuit (1502) acquires pixel matrix information from the luminance information and color difference information of the input video and outputs this. The vertical contour extraction circuit (1503) extracts only vertical contour components from the pixel matrix information, and the horizontal contour extraction circuit (1504) extracts only horizontal contour components from the pixel matrix information. The contour direction information acquisition circuit (1505) acquires contour direction information from the sensitivity results of the input pixel matrix information and the coefficient matrix. When the contour direction information is input from the contour direction information acquisition circuit (1505), the contour control circuit (1506) calculates and outputs vertical contour correction information and horizontal contour correction information for correcting the contour. The vertical contour correction circuit (1507) acquires corrected vertical contour pixel information by performing gain control on the vertical contour pixels based on the vertical contour correction information output from the contour control circuit (1506). Similarly, the horizontal contour correction circuit (1508) acquires corrected horizontal contour pixel information by performing gain control on the horizontal contour pixels based on the horizontal contour correction information output from the contour control circuit (1506). The contour addition circuit (1509) generates corrected contour pixel information from the corrected vertical contour pixel information and the corrected horizontal contour pixel information. The corrected contour pixel information is input to the coring circuit (1510). The following processing is the same as the specific example in the first embodiment.

  (Processing Flow of Embodiment 2) The basic processing flow of the contour correction method in this embodiment is shown in the processing flow of FIG. In the contour correction method according to the present embodiment, first, n × n pixel matrix information digitized according to luminance and color difference is acquired from a video signal (S1601: pixel matrix information acquisition step). Next, contour direction information indicating the direction of the contour is acquired from the acquired pixel matrix information (S1602: contour direction information acquisition step). In the contour direction information acquisition step in this embodiment (S1602), the contour direction information is acquired as follows. First, a sensitivity result of the coefficient matrix corresponding to the contour direction pattern type sensitive to the contour direction of the pixel matrix information and the acquired pixel matrix information is acquired (S1603: sensitivity result acquisition substep). Then, contour direction information corresponding to the pixel matrix information is acquired according to the sensitivity result in the sensitivity result acquisition substep (S1603) (S1604: contour direction information acquisition substep). When the contour direction information is acquired, the contour correction amount for correcting the contour pixel of the acquired pixel matrix information is calculated according to the contour direction information, and the contour pixel of the pixel matrix information acquired based on the calculated contour correction amount is calculated. Correction (S1605: contour pixel correction step). In the contour pixel correction step (S1605) in the present embodiment, the contour pixel is corrected as follows. First, based on the contour direction information acquired in the contour direction information acquisition substep (S1604), vertical contour correction information indicating the vertical contour correction amount and horizontal contour correction information indicating the horizontal contour correction amount are acquired. (S1606: Vertical and horizontal contour correction information acquisition sub-step). Then, the vertical contour correction information acquired in the vertical horizontal contour correction information acquisition sub-step (S1606) is used to correct the vertical contour pixels of the pixel matrix information to acquire corrected vertical contour pixel information ( S1607: Corrected vertical contour pixel information acquisition sub-step). Further, the corrected horizontal contour pixel information is acquired by correcting the horizontal pixel of the pixel matrix information using the acquired horizontal contour correction information (S1608: corrected horizontal contour pixel information acquisition sub-step). The context of the corrected vertical contour pixel information acquisition substep (S1607) and the corrected horizontal contour pixel information acquisition substep (S1608) may be reversed or simultaneous. When the corrected vertical contour pixel information and the corrected horizontal contour pixel information are acquired, corrected contour pixel information corresponding to the contour pixel of the new pixel matrix information is generated from these information (S1609: corrected). Outline pixel information generation sub-step). In this way, it is possible to generate an image modified so as to enhance the contour pixels.

  (Effect of Embodiment 2) In the image processing apparatus of the present embodiment having the above-described configuration, it is possible to acquire contour direction information from pixel matrix information and perform contour correction according to the contour direction information. In addition, by performing vertical and horizontal contour correction separately, it is possible to perform appropriate contour correction by increasing the gain control amount of the contour close to the horizontal, and the contour of the corrected video is unnatural. Can be prevented.

  << Embodiment 3 >> (Concept of Embodiment 3) This embodiment is based on the image processing apparatus of Embodiment 2 of the image processing, and further has a feature in that the degree of enhancement of the corrected contour pixel is controlled.

  (Configuration of Third Embodiment) FIG. 17 shows an example of a functional block diagram of an image processing apparatus according to the present embodiment. The “image processing apparatus” (1700) according to the present embodiment includes a “pixel matrix information acquisition unit” (1701), a “contour direction information acquisition unit” (1702), a “contour pixel correction unit” (1703), The contour direction information acquisition unit (1702) includes “coefficient matrix holding means” (1704), “sensitivity result acquisition means” (1705), and “contour direction information acquisition means” (1706). The pixel correction unit (1703) includes a “vertical horizontal contour correction information acquisition unit” (1707), a “corrected vertical contour pixel information acquisition unit” (1708), and a “corrected horizontal contour pixel information acquisition unit” (1709). And “corrected contour pixel information generating means” (1710), and the contour pixel correcting section (1703) further includes “contour emphasis coefficient acquisition means” (1711).Since the configuration other than the “contour enhancement coefficient acquisition unit” (1711) has already been described, the description thereof will be omitted.

  (Description of Configuration of Embodiment 3) The “contour emphasis coefficient acquisition means” (1711) in the new pixel matrix of the corrected contour pixel information according to the sensitivity intensity of the sensitivity result in the sensitivity result acquisition means (1705). This is a means for acquiring an edge enhancement coefficient, which is a coefficient that determines the degree of enhancement. The “sensitivity intensity” is an index indicated by the magnitude of the value in the sensitivity result between the pixel matrix information and the coefficient matrix. For example, as described above, the sensitivity result of the contour c of FIG. 9 with each contour direction pattern of FIG. 8 is shown in FIG. 10 (2), but the value of each sensitivity result at this time is the largest value. Is the sensitivity intensity of the contour c. As shown in FIG. 10 (2), the contour direction pattern H indicates the highest sensitivity result in each contour direction pattern of the contour c. Therefore, the contour calculated by the coefficient matrix shown in FIG. The sensitivity intensity of c is 240, which is the sensitivity result with the contour direction pattern H.

  The “enhancement degree” is a ratio for further adjusting the corrected contour. If the corrected contour has a high sensitivity intensity, the contour may be overemphasized. In addition, when the sensitivity results of three or more contour direction patterns are the same value, there is a possibility that the contour does not exist. In such a case, it is not necessary to correct the pixel matrix information. For this reason, it is desirable to adjust the degree of enhancement according to the sensitivity intensity.

  The “outline enhancement coefficient” is a coefficient that determines the degree of enhancement. Let this be γ. The degree of enhancement can be adjusted by multiplying the outline enhancement coefficient γ by the outline pixel of the corrected pixel matrix information. The contour enhancement coefficient can be calculated as follows. FIG. 20 shows the relationship between the contour strength and the contour enhancement coefficient γ. As shown in FIG. 20, the contour enhancement coefficient γ is constant until the contour strength reaches a predetermined value, and no special adjustment is made (for example, γ = 1). However, if the contour strength exceeds a predetermined value, the contour may be overemphasized, so the contour enhancement coefficient γ is reduced. In addition to the relationship with the contour strength shown in FIG. 20, when the sensitivity result is 3 or more, the contour may not exist, so γ = 0 may be set. Note that γ is preferably set to a value such that the enhancement amount for the contour pixel is 2% to 7% in accordance with the values of α and β described in the first and second embodiments.

  (Specific Example of Embodiment 3) FIG. 18 shows a specific contour correction circuit (1800) for realizing the image processing apparatus according to this embodiment. The contour correction circuit (1800) shown in FIG. 18 obtains n × n pixel matrix information from the output of the line memory (1801) and the output of the line memory (1801), which delays the input image and outputs it to the pixels for each line. A pixel matrix information acquisition circuit (1802), a vertical contour extraction circuit (1803) that extracts only vertical contour pixels, a horizontal contour extraction circuit (1804) that extracts only horizontal contour pixels, and pixel matrix information The contour direction information acquisition circuit (1805) for acquiring the contour direction information from the contour direction, the contour control circuit (1806) for calculating the vertical contour correction information and the horizontal contour correction information from the contour direction information, and the vertical calculated by the contour control circuit. A vertical contour correction circuit (1807) for acquiring corrected vertical contour pixel information of the pixel matrix information by the contour correction information; A horizontal contour correction circuit (1808), a vertical contour correction circuit (1807), and a horizontal contour correction circuit (1808) for acquiring corrected horizontal contour pixel information of the pixel matrix information based on the horizontal contour correction information calculated by the contour control circuit. The contour adding circuit (1809) for generating corrected contour pixel information from the corrected vertical contour pixel information and the corrected horizontal contour pixel information acquired in step 1 and coring the contour generated by the contour adding circuit (1809) A coring circuit (1810) that performs the contour emphasis coefficient acquisition circuit (1813) that calculates an edge emphasis coefficient, and a contour adjustment circuit (1814) that adjusts the coring contour output from the coring circuit using the contour emphasis coefficient. A delay circuit (1811) for delaying pixels from the line memory, and an adjusted wheel output from the contour adjustment circuit (1814) It comprises an adder circuit (1812) for adding the outline pixel and the pixel output from the delay circuit (1811).

  The functional blocks of the image processing apparatus (1700) shown in FIG. 17 correspond to the contour correction circuit (1800) as follows. The pixel matrix information acquisition unit (1701) corresponds to the line memory (1801) and the pixel matrix acquisition circuit (1802). The contour direction information acquisition unit (1702) corresponds to the contour direction information acquisition circuit (1805). The contour pixel correction unit (1703) includes a vertical contour extraction circuit (1803), a horizontal contour extraction circuit (1804), a contour control circuit (1806), a vertical contour correction circuit (1807), a horizontal contour correction circuit (1808), and contour addition. The circuit corresponds to a circuit (1809), a delay circuit (1811), an adder circuit (1812), an edge enhancement coefficient acquisition circuit (1813), and an adjustment circuit (1814). The contour pixel correction unit may include a coring circuit (1810). Note that these are merely forms for realizing the image processing apparatus of the present embodiment, and may be realized by other similar circuits as long as the configuration requirements of the present embodiment are provided.

  The operation of these circuits is substantially the same as that of the contour correction circuit described in the second embodiment, but the contour enhancement coefficient acquisition circuit (1813) in the present embodiment outputs the contour direction information acquisition circuit (1805). The edge enhancement coefficient corresponding to the sensitivity intensity calculated from the sensitivity results is acquired. The contour adjustment circuit (1814) adjusts the output from the coring circuit (1810) based on the acquired contour enhancement coefficient. In the contour adjustment circuit (1814), the correction value for the contour pixel of the pixel matrix information having a weak contour component can be set to 0, or the degree of enhancement of the contour whose contour component is too strong can be suppressed. The adjusted contour pixel is added to the original image delayed by the delay circuit (1811) by the adder circuit (1812), thereby outputting a video signal in which the contour component is emphasized.

  (Processing Flow of Embodiment 3) The basic processing flow of the contour correction method in this embodiment is shown in the processing flow of FIG. In the contour correction method according to the present embodiment, first, n × n pixel matrix information that is digitized according to luminance and color difference is acquired from a video signal (S1901: pixel matrix information acquisition step). Next, contour direction information indicating the direction of the contour is acquired from the acquired pixel matrix information (S1902: contour direction information acquisition step). In the contour direction information acquisition step (S1902), a sensitivity result of the coefficient matrix for the contour direction pattern type sensitive to the contour direction of the pixel matrix information and the acquired pixel matrix information is acquired (S1903: sensitivity result acquisition substep). ). Then, contour direction information corresponding to the pixel matrix information is acquired according to the sensitivity result in the sensitivity result acquisition substep (S1903) (S1904: contour direction information acquisition substep). When the contour direction information is acquired, the contour correction amount for correcting the contour pixel of the acquired pixel matrix information is calculated according to the contour direction information, and the contour pixel of the pixel matrix information acquired based on the calculated contour correction amount is calculated. Correction (S1905: contour pixel correction step). In the contour pixel correction step (S1905), based on the contour direction information acquired in the contour direction information acquisition substep (S1904), the vertical contour correction information indicating the vertical contour correction amount and the horizontal contour correction amount are indicated. Horizontal contour correction information is acquired (S1906: vertical horizontal contour correction information acquisition sub-step). Then, using the vertical contour correction information acquired in the vertical horizontal contour correction information acquisition sub-step (S1906), the vertical contour pixel information of the pixel matrix information is corrected to acquire corrected vertical contour pixel information ( S1907: Corrected vertical contour pixel information acquisition sub-step), using the acquired horizontal contour correction information, correct the horizontal contour pixels of the pixel matrix information to acquire corrected horizontal contour pixel information (S1908: Modified horizontal contour pixel information acquisition sub-step). When the corrected vertical contour pixel information and the corrected horizontal contour pixel information are acquired, corrected contour pixel information corresponding to the contour pixel of the new pixel matrix information is generated from these information (S1909: corrected). Outline pixel information generation sub-step). Furthermore, in the contour pixel correction step (S1905) of the present embodiment, the sensitivity intensity is acquired from the sensitivity result acquired in the sensitivity result acquisition substep (S1903), and new corrected contour pixel information is added according to the sensitivity intensity. An edge enhancement coefficient, which is a coefficient that determines the degree of enhancement in the pixel matrix information, is acquired, and the degree of edge enhancement is adjusted (outline adjustment substep: S1910). In this way, it is possible to generate an image modified so as to emphasize the contour pixels.

  (Effect of Embodiment 3) In the image processing apparatus of the present embodiment having the above-described configuration, it is possible to acquire contour direction information from pixel matrix information and perform contour correction according to the contour direction information. Further, by adjusting the corrected contour according to the contour strength, more appropriate contour correction can be performed, and it is possible to prevent the contour of the corrected image from becoming unnatural.

The figure explaining the outline of Embodiment 1 Embodiment 1 Functional Block Diagram The figure showing the outline pixel of pixel matrix information The figure showing the specific circuit structure of Embodiment 1. FIG. The figure explaining a coring circuit The figure explaining the flow of processing of Embodiment 1. Functional block diagram of Embodiment 2 Diagram explaining coefficient matrix The figure explaining the acquisition method of pixel matrix information from an outline The figure explaining the sensitivity result with each outline direction pattern (1) The figure explaining the sensitivity result with each outline direction pattern (2) The figure explaining the state which extracted the part for the vertical outline pixel from pixel matrix information The figure explaining the state which extracted the horizontal outline pixel part from pixel matrix information The figure explaining the method to acquire the corrected pixel matrix information from the corrected vertical outline pixel information and the corrected horizontal outline pixel information The figure showing the specific circuit structure of Embodiment 2. The figure explaining the flow of processing of Embodiment 2. Functional block diagram of Embodiment 3 The figure showing the specific circuit structure of Embodiment 3. Process flow of embodiment 3 Diagram showing the relationship between contour strength and contour enhancement coefficient

Explanation of symbols

0200 Image processing device 0201 Pixel matrix information acquisition unit 0202 Contour direction information acquisition unit 0203 Contour pixel correction unit 0704 Coefficient matrix holding unit 0705 Sensitive result acquisition unit 0706 Contour direction information acquisition unit 0707 Vertical horizontal contour correction information acquisition unit 0708 Corrected vertical contour Pixel acquisition means 0709 Modified horizontal outline pixel acquisition means 0710 Modified outline pixel information generation means 1711 Outline enhancement coefficient acquisition means S0601 Pixel matrix information acquisition step S0602 Outline direction information acquisition step S0603 Outline pixel correction step

Claims (6)

A pixel matrix information acquisition unit for acquiring pixel matrix information;
A contour direction information acquisition unit that acquires contour direction information that is information indicating a contour direction from the acquired pixel matrix information;
A contour pixel correction unit that corrects the contour pixel of the pixel matrix information using the acquired contour direction information and sets the new pixel matrix information by emphasizing the contour;
Have
The contour direction information acquisition unit
Coefficient matrix holding means for holding a coefficient matrix sensitive to the contour direction of the pixel matrix information for the contour direction pattern type;
A sensitivity result acquisition means for acquiring a sensitivity result of the held coefficient matrix with respect to the acquired pixel matrix information;
Contour direction information acquisition means for acquiring contour direction information corresponding to the pixel matrix information according to the sensitivity results;
An image processing apparatus. The contour pixel correction unit includes:
Vertical horizontal contour correction information acquisition means for acquiring vertical contour correction information indicating a vertical contour correction amount and horizontal contour correction information indicating a horizontal contour correction amount based on the acquired contour direction information;
Corrected vertical contour pixel information acquisition means for correcting vertical contour pixels in the vertical direction of the pixel matrix information using vertical contour correction information and acquiring corrected vertical contour pixel information;
Corrected horizontal contour pixel information acquisition means for correcting horizontal contour pixels in the horizontal direction of the pixel matrix information using horizontal contour correction information and acquiring corrected horizontal contour pixel information;
Corrected contour pixel information generating means for generating corrected contour pixel information that is a contour pixel of new pixel matrix information from the acquired corrected vertical contour pixel information and corrected horizontal contour pixel information;
The image processing apparatus according to claim 1 , comprising: The contour pixel correction unit includes:
3. The method according to claim 1, further comprising an edge enhancement coefficient acquisition unit configured to acquire an edge enhancement coefficient that is a coefficient for determining an enhancement degree in the new pixel matrix of the corrected edge pixel information in accordance with the sensitivity intensity of the acquired sensitivity result. the image processing apparatus according to one or. A pixel matrix information acquisition step for acquiring pixel matrix information;
A contour direction information acquisition step for acquiring contour direction information which is information indicating the contour direction from the acquired pixel matrix information;
A contour pixel correction step for correcting the contour pixel portion of the pixel matrix information using the acquired contour direction information to obtain new pixel matrix information with enhanced contours;
Consists of
The contour direction information acquisition step includes:
A sensitivity result acquisition substep for acquiring a sensitivity result of a coefficient matrix for a contour direction pattern type sensitive to a contour direction of pixel matrix information and the acquired pixel matrix information;
A contour direction information acquisition sub-step of acquiring contour direction information corresponding to pixel matrix information according to the sensitivity result;
A contour correction method comprising: The contour pixel correcting step includes:
A vertical horizontal contour correction information acquisition sub-step for acquiring vertical contour correction information indicating a vertical contour correction amount and horizontal contour correction information indicating a horizontal contour correction amount based on the acquired contour direction information;
A corrected vertical contour pixel information acquisition sub-step for acquiring corrected vertical contour pixel information by correcting the vertical contour pixels of the pixel matrix information using vertical contour correction information;
A corrected horizontal contour pixel information acquisition sub-step for acquiring corrected horizontal contour pixel information by correcting horizontal contour pixels of the pixel matrix information using horizontal contour correction information;
A corrected contour pixel information generation sub-step for generating corrected contour pixel information which is a contour pixel of new pixel matrix information from the acquired corrected vertical contour pixel information and the corrected horizontal contour pixel information;
The contour correction method according to claim 4 . The contour pixel correction step
A contour adjustment sub that adjusts the corrected contour pixel information by acquiring a contour emphasis coefficient that is a coefficient for determining the degree of emphasis in the new pixel matrix information of the corrected contour pixel information according to the sensitivity intensity of the acquired sensitivity result The contour correction method according to claim 4, further comprising steps.

Images