JP4049116B2 - Image processing device - Google Patents

Description

  The present invention relates to an image processing apparatus having a function of detecting a skin area of a person image, and relates to an image processing apparatus suitable for use in an apparatus for reproducing a natural image such as a television, a video camera, or a video printer. is there.

  In general, in an image processing apparatus, an object is extracted in order to analyze and recognize the contents of an image. For example, there are a method for obtaining an outline of an object by detecting an edge and a method for extracting an area occupied by the object by dividing an image into partial areas with uniform brightness.

  In addition, in an apparatus that reproduces a natural image such as a television, a video camera, or a video printer, processing such as contour enhancement and gradation conversion is performed on the object extracted by the image processing apparatus. This improves the sharpness and contrast of the image.

  Here, when a person image is processed by the image processing apparatus as described above, a skin area detection processing unit of the image processing apparatus detects a human skin area (hereinafter referred to as a skin area). In order to detect the skin area, the skin area detection processing unit, for example, a method of extracting the skin area based on color information for each pixel or the connectivity of the position of the skin color pixel extracted from the color information on the image. For example, a method of extracting a skin region by dividing the region and analyzing the characteristics of each region is used.

  However, in the method of extracting the skin area based on the color information for each pixel, all the skin color areas in the image are regarded as the skin area, and many erroneous detections have occurred. Alternatively, the method of extracting a skin region by region division is not suitable for real-time processing because a lot of calculation time is required for region division processing and feature analysis processing of each region.

  Further, in order to improve the sharpness and contrast of the image, the extracted object is subjected to contour enhancement processing or gradation conversion processing. The skin region, particularly the human face region ( Hereinafter, it is also referred to as a face area), and contour enhancement processing or gradation conversion processing has been performed. For this reason, for example, there is a problem that the face region is emphasized and an undesirable result is brought about.

  Accordingly, the present invention has been made in view of the above-described conventional situation and has the following objects.

  That is, an object of the present invention is to provide an image processing apparatus capable of improving the sharpness of an image by enhancing the outline of another area without enhancing the outline of the face area of a person.

In order to solve the above-described problems, the present invention converts input human image data into digitized luminance data and color difference data, and detects a skin area from the digitized luminance data and color difference data. An image processing apparatus comprising: means for converting the luminance data and color difference data into three attribute data of luminance data, hue data, and chroma data; and a vertical in a skin color region from the luminance data obtained by the converting means. A vertical luminance change detecting means for detecting a luminance change in a direction and generating a cumulative distribution for each position in the horizontal direction and a cumulative distribution for each position in the vertical direction by detecting the luminance change in the vertical direction, and obtained by the converting means. A flesh color pixel extracting means for extracting flesh color pixels from the three attribute data, and a spatial flesh color pixel distribution of the flesh color pixels extracted by the flesh color pixel extracting means. A flesh color pixel distribution generating unit, a flesh color pixel distribution generated by the flesh color pixel distribution generating unit, a cumulative distribution for each position in the horizontal direction of a vertical luminance change generated by the vertical luminance change detecting unit, and The skin area detecting means comprising a face area determining means for determining a face area of a person image from the cumulative distribution at each position in the vertical direction, and an outline for performing edge enhancement processing on the luminance data of the input person image data A color difference synchronized with luminance data that has been given a predetermined delay amount corresponding to the time required for the contour emphasizing process to the emphasizing means and the color difference data of the input human image data, Delay means for outputting data, and controlling the contour enhancement means based on the skin area information detected by the skin area detection means.

The image processing apparatus according to the present invention further stores the human image data converted into the luminance data and the color difference data, and a predetermined delay corresponding to a time required for detecting the skin area information by the skin area detecting means. A storage means for giving an amount and outputting; a conversion means for converting the person image data given the predetermined delay amount by the storage means into three-attribute data of luminance data, hue data and chroma data; and the conversion means Detecting means for detecting skin color data from the three attribute data obtained, and the delay means takes time required for the contour enhancement processing on the color difference data of the person image data given the predetermined delay amount by the storage means. gives a predetermined delay amount corresponding to said contour enhancement means, the control had groups Dzu skin region information detected by the skin area detection unit, to the detection means And performing edge enhancement processing on the luminance data of the person image data corresponding to the detected detection result.

The image processing apparatus according to the present invention, furthermore, the scene change detecting means for detecting a scene change from the person image data that has been the input, the scene when the same scene have groups Dzu on the detection result of the change detection means is continuous only Smoothing means for temporally smoothing the spatial position of the skin area information detected by the skin area detection means by the spatial position of the skin area information detected in the previous image by the skin area detection means; comprising a said edge enhancement means that performs edge enhancement processing on the luminance data of the person image data corresponding to the detection result detected by the detection means by controlling who were based Dzu skin region information from said smoothing means Features.

In the image processing apparatus according to the present invention, the skin area detection means converts the digitized human image data into luminance data, color difference data, and three attribute data of luminance data, hue data, and chroma data by the conversion means , and performs the above conversion. From the luminance data obtained by the means, the vertical luminance change detecting means detects the vertical luminance change in the skin color region, and the cumulative distribution of the vertical luminance changes for each position in the horizontal direction and the vertical direction for each position. A cumulative distribution is generated, and a skin color pixel is extracted by the skin color pixel extraction unit from the three attribute data obtained by the conversion unit, and a spatial skin color pixel distribution of the skin color pixel extracted by the skin color pixel extraction unit is generated. The flesh color pixel distribution generated by the flesh color pixel distribution generating unit and the vertical luminance generated by the vertical luminance change detecting unit. The human skin color area is analyzed by determining the face area of the person image by the face area determining means from the cumulative distribution at each position in the horizontal direction and the cumulative distribution at each position in the vertical direction. Skin area detection that detects human face area more accurately, converts input human image data into digitized luminance data and color difference data, and detects skin area from digitized luminance data and color difference data Based on the skin region information detected by the means, the contour emphasizing means is controlled to perform the contour emphasizing process on the luminance data of the person image data. Further, the delay unit gives a predetermined delay amount corresponding to the time required for the contour enhancement process to the color difference data of the input human image data, and synchronizes with the luminance data subjected to the contour enhancement process by the contour enhancement unit. By outputting the color difference data, it is possible to enhance the sharpness of the image by enhancing the outline of other areas without enhancing the outline of the face area of the person.

Further, in the image processing apparatus according to the present invention, the human image data converted into the luminance data and the color difference data is stored in a storage unit, and the time required for detecting the skin region information by the skin region detection unit. The corresponding predetermined delay amount is given. The person image data given the predetermined delay amount by the storage means is converted into three attribute data of luminance data, hue data and chroma data by the converting means, and the skin color data is detected by the detecting means from the three attribute data. The delay means gives a predetermined delay amount corresponding to the time required for the contour emphasizing process to the color difference data of the person image data given the predetermined delay amount by the storage means. the control had groups Dzu the detected skin region information by the skin area detection unit performs edge enhancement processing on the luminance data of the person image data corresponding to the detection result detected by the detection means. Thereby, the sharpness of the image can be further improved.

Further, in the image processing apparatus according to the present invention, the detection result of the scene change detection means for detecting a scene change from the person image data the input to be had based Dzu, by smoothing means, if the same scene is continuous only the The spatial position of the skin area information detected by the skin area detection means is temporally smoothed by the spatial position of the skin area information detected in the previous image by the skin area detection means . Then, the edge enhancement means performs edge enhancement processing on the luminance data of the person image data corresponding to the detection result detected by the detection means to the skin region information by control had groups Dzu from said smoothing means. As a result, the temporal variation of the skin area in the same scene can be reduced, so that the sharpness of the image can be further improved.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  First, an image processing apparatus according to the first embodiment of the present invention will be described.

  For example, the image processing apparatus includes a skin region detection unit 1 that detects a human skin region (hereinafter referred to as a skin region) from input image data. As shown in FIG. 1, a coordinate converter 11 that converts input image data into color three-attribute data, and a spatial skin-color pixel distribution of skin-color pixels from the three-attribute data obtained by the coordinate converter 11 are stored in a memory. 13, a histogram generator 12 generated on the memory 13, the memory 13, and a face area determiner 14 that determines a human face area (hereinafter referred to as a face area) from the skin color pixel distribution generated on the memory 13. And.

First, as shown in FIG. 2, an analog luminance signal Y _a and color difference signals U _a and V _a obtained by scanning the image P in the horizontal direction are converted into an analog / digital converter (hereinafter referred to as an A / D converter). referred to.) by 101 digitally converted, digitally converted luminance signal Y _d and the color difference signals U _d, V _d is input to the face region detector 1.

  Hereinafter, the face area detection unit 1 will be described in detail.

The coordinate converter 11 converts the luminance signal Y _d and the color difference signals U _d and V _d digitally converted by the A / D converter 101 into luminance data Y, hue data H, and chroma data, which are three color attribute data. Convert to C. There are various definitions of the three attribute data of color. For example, luminance data Y, hue data H, and chroma data are calculated from the luminance signal Y _d and the color difference signals U _d and V _d by the calculation shown in Formula 1. Find C.

  As shown in FIG. 3, the histogram generator 12 generates on the memory 13 a horizontal skin color pixel histogram SH in which the number of skin color pixels arranged in the vertical direction is accumulated for each horizontal position i of the image P.

That is, with respect to the luminance data Y, hue data H, and chroma data C obtained by the coordinate converter 11, luminance constants Y _min and Y _max set in advance to define the skin color range, and the hue Having constants H _min , H _max and chroma constants C _min , C _max ,
Y _min <Y <Y _max
H _min <H <H _max
C _min <C <C _max
It is determined whether or not the following skin color condition is satisfied. For example, when the position on the image P corresponding to the luminance data Y, hue data H, and chroma data C is indicated by a pixel S (i, j), the pixel S (i, j) satisfies the skin color condition. If it is determined that the skin color is satisfied, 1 is added to the i-th value SH [i] of the skin color pixel histogram SH on the memory 13 (SH [i] = SH [i] +1). By performing such processing for each horizontal position, the number of skin color pixels arranged in the vertical direction is accumulated, and a skin color pixel histogram SH [i] is generated on the memory 13.

  The face area determination unit 14 analyzes the skin color pixel histogram SH [i] when the skin color pixel histogram SH [i] of one image P is generated on the memory 13 by the histogram generator 12. The face area F of the person M in the image P shown in FIG. 3 is detected.

That is, as shown in FIG. 4, first, the horizontal position i having the highest frequency is detected from the skin color pixel histogram SH [i], and this is set as the maximum value i _max .

Then, the maximum value _i minimum point _max first appear towards both the front and rear about the _i m, detects a _{i p.}

Then, towards both back and forth about it said maximum value _{i max,} the maximum value _{i max} th frequency SH _{[i max]} first point i.alpha _m corresponding to alpha% of the frequency of detecting the i.alpha _p. Here, “α” is a constant satisfying 0 ≦ α ≦ 1.0 for defining the width of the skin region. However, when the point iα _m does not exist in the range of i _m ≦ iα _m ≦ i _max , or when the point iα _p does not exist in the range of i _max ≦ iα _p ≦ i _p Respectively, iα _m = i _m and iα _p = i _p .

In this way, the detected minimum point _i m, _{i p,} and the point i.alpha _m, i.alpha _p is the threshold _{T asp1} the lower limit of the ratio of the length of the vertical and horizontal, upper threshold _{T asp2} and, It is determined whether or not the facial area conditional expression shown in Expression 2 is satisfied with the lower threshold value _Tnum of the number of pixels.

If it meets the indicated condition with the number 2 considers minimum point i _m, the area S _mp sandwiched i _p and the face region F, the output minimum point i _m, a i _p as the information of the face area To do. Alternatively, when the conditional expression shown in the above equation 2 is not satisfied, SH [k] = 0 (im <k <ip) is repeated and the above-described analysis processing is repeated.

Note that the constant α and the threshold values T _asp1 , T _asp2 , and T _num are set in advance.

  The operation of the skin area detection unit 1 configured as described above will be described.

The luminance signal Y _d and the color difference signals U _d and V _d of the image P digitally converted by the A / D converter 101 are input to the coordinate converter 11.

The coordinate converter 11 converts the input luminance signal Y _d and color difference signals U _d and V _d into luminance data Y, hue data H, and chroma data C, and converts the converted luminance data Y, hue data H, The chroma data C is supplied to the histogram generator 12.

  The histogram generator 12 determines whether or not the luminance data Y, hue data H, and chroma data C from the coordinate converter 11 satisfy the skin color condition, and stores the skin color pixel histogram SH on the memory 13. [I] is generated.

The face area determination unit 14 detects an area where skin color pixels are dense from the skin color pixel histogram SH [i] for one image generated on the memory 13 by the histogram generator 12, and detects the detected skin area. It is determined whether it is a face area based on the ratio of the size of the face and the length and width. If the determination result is a face area, the face area information i _m and _ip is output.

  As described above, in this embodiment, the skin color pixel histogram SH [i] in the horizontal direction is generated, and an area where skin color pixels are concentrated is detected as a skin area based on the skin color pixel histogram SH [i]. Since the face area is determined based on the ratio of the size of the skin area and the length and width, the analysis of the human skin area and the detection of the human face area can be performed quickly and accurately. .

  Next, an image processing apparatus according to a second embodiment of the present invention will be described.

  The image processing apparatus includes the skin region detection unit 1 having the same configuration as the skin region detection unit 1 of the image processing device according to the first embodiment described above, but the histogram of the skin region detection unit 1 in the present embodiment. As shown in FIG. 5, the generator 12 generates a skin color pixel histogram SH [i] in the horizontal direction and a skin color pixel histogram SV [j] in the vertical direction on the memory 13.

  That is, when it is determined that the position S (i, j) on the image P is a skin color, 1 is added to the i-th value SH [i] of the horizontal skin color pixel histogram SH on the memory 13. (SH [i] = SH [i] +1) and 1 is added to the jth value SV [j] of the skin color pixel histogram SV in the vertical direction (SV [j] = SV [j] +1).

In this way, the horizontal flesh color pixel histogram SH [i] and the vertical flesh color pixel histogram SV [j] are generated on the memory 13, and the face area determination unit 14 uses the horizontal flesh color pixel histogram SH [ i] and the skin color pixel histogram SV [j] in the vertical direction, analysis is performed in the same manner as in the first embodiment described above, and face area information i _m , i _p , and j _m , j _p is output.

  As described above, in this embodiment, in addition to the horizontal skin color pixel histogram SH [i], the vertical skin color pixel histogram SV [j] is also generated. The region can be detected more accurately.

  Next, an image processing apparatus according to the third embodiment of the present invention will be described.

  As shown in FIG. 6, the image processing apparatus includes a vertical luminance for detecting a luminance change in the vertical direction in the skin area in addition to the configuration of the skin area detecting unit 1 of the first and second embodiments described above. A skin region detection unit 2 having a change detector 21 is provided.

  In addition, the same code | symbol is attached | subjected to the location which shows the same operation | movement as the image processing apparatus shown in the said FIG. 1, and the detailed description is abbreviate | omitted.

  That is, the vertical luminance change detector 21 calculates the differential value in the vertical direction of the luminance data Y among the luminance data Y, hue data H, and chroma data C converted by the coordinate converter 11 in each horizontal position. A horizontal differential histogram DH [i] accumulated every time is generated on the memory 13. In addition, a vertical differential histogram DV [j] obtained by accumulating the vertical differential values of the luminance data Y for each position in the vertical direction is also generated on the memory 13.

For example, a vertical differential value v _ij of luminance data Y at an arbitrary position (i, j) on the image has a constant d _j for differential calculation,
v _ij = | 2Y (i, j) −Y (i, j−d _j ) −Y (i, j + d _j ) |
Is obtained by the following calculation. This arithmetic expression represents the absolute value of the secondary differential value in the vertical direction.

Accordingly, the differential value v _ij is added to the i-th value DH [i] of the horizontal differential histogram DH on the memory 13 (DH [i] = DH [i] + v _ij ). Further, the differential value v _ij is added to the j-th value DV [j] of the skin color pixel histogram DV in the vertical direction (DV [j] = DV [j] + v _ij ).

  In this way, the horizontal differential histogram DH [i] and the vertical differential histogram DV [j] are generated on the memory 13.

  On the other hand, for example, the histogram generator 12 generates a horizontal skin color pixel histogram SH [i] and a vertical skin color pixel histogram SV [j] on the memory 13 in the same manner as in the second embodiment described above. .

The face area determination unit 14 uses face area information i _m , i as in the second embodiment based on the horizontal skin color pixel histogram SH [i] and the vertical skin color pixel histogram SV [j]. _{Find p 1} and j _m , j _p . Here, the face area determination unit 14 further performs the calculation shown in Formula 3 or Formula 4 based on the differential histogram DH [i] in the horizontal direction and the differential histogram DV [j] in the vertical direction.

Equation 3 or, the value d obtained by the calculation shown in Equation 4 is the case were predetermined threshold T _dev above, information i _m of the detected face _region, i _p, and _outputs a j m, _{j p,} if the value d is less than or equal to the threshold value _{T dev} is, SH [k] _{= 0} and repeats the analysis process as _{(i m <k <i p} ).

  The operation of the skin area detection unit 2 configured as described above will be described.

The luminance signal Y _d and the color difference signals U _d and V _d of the image P digitally converted by the A / D converter 101 are input to the coordinate converter 11.

The coordinate converter 11 converts the input luminance signal Y _d and color difference signals U _d and V _d into luminance data Y, hue data H, and chroma data C, and converts the converted luminance data Y, hue data H, The chroma data C is supplied to the histogram generator 12 and the luminance data Y is supplied to the vertical luminance change detector 21.

  The histogram generator 12 determines whether the luminance data Y, the hue data H, and the chroma data C from the coordinate converter 11 satisfy the skin color condition, thereby causing the horizontal skin color on the memory 13. The pixel histogram SH [i] and the skin color pixel histogram SV [j] in the vertical direction are generated.

On the other hand, the vertical luminance change detector 21 includes a horizontal differential histogram DH [i] obtained by accumulating vertical differential values of the luminance data Y from the coordinate converter 11 for each position in the horizontal direction, and the luminance. A vertical differential histogram DV [j] obtained by accumulating the vertical differential values of the data Y at each position in the vertical direction is generated on the memory 13. The face area determination unit 14 detects, as a skin area, an area where skin color pixels are concentrated from the skin color pixel histogram SH [i] and the skin color pixel histogram SV [j] generated on the memory 13 by the histogram generator 12. Whether or not it is a face region is determined based on the ratio of the size of the skin region and the length and width. If the determination result is a face area, the face area is determined based on the differential histogram DH [i] and the differential histogram DV [j] generated on the memory 13 by the vertical luminance change detector 21. Information i _m , i _p , j _m , and j _p are output.

The detection of face area information i _m , i _p and j _m , j _p based on the distribution of luminance change in the vertical direction as described above, the human face includes horizontal edges such as eyes and mouth. However, by detecting the face area based on the distribution of the luminance change in the vertical direction, for example, avoiding false detection of a uniform area with a small luminance change such as a wall. Can do. Therefore, the analysis of the person's skin color area and the detection of the person's face area can be performed more accurately.

  In the present embodiment, the histogram generator 12 generates the skin color pixel histogram SH [i] in the horizontal direction and the skin color pixel histogram SV [j] in the vertical direction on the memory 13. Similarly to the first embodiment, only the skin color pixel histogram SH [i] in the horizontal direction may be generated on the memory 13.

  Next, an image processing apparatus according to a fourth embodiment of the present invention is described.

As shown in FIG. 7, the image processing apparatus includes luminance data Y, hue data H, chroma data C, and luminance in the vertical direction in addition to the configuration of the skin area detection unit 2 of the third embodiment described above. A skin region detection unit 3 having an in-block feature amount calculator 31 that averages the change v _ij in units of blocks is provided.

  In addition, the same code | symbol is attached | subjected to the location which shows the same operation | movement as the image processing apparatus shown in the said FIG. 1 and the said FIG. 6, and the detailed description is abbreviate | omitted.

That is, the intra-block feature quantity calculator 31 converts the luminance data Y, hue data H, and chroma data C of the image P converted by the coordinate converter 11 into, for example, the block B as shown in FIG. The block is divided into blocks B each having a constant number m _i and m _j representing the size and having a size indicated by the number of pixels in the horizontal direction (2m _i +1) and the number of pixels in the vertical direction (2m _j +1). Therefore, the number of pixels M in each block is
M = (2m _i +1) (2m _j +1)
Can be expressed as

Then, with the number of pixels M, the number 5, 6, the calculation shown in Equation 7, the average value H _ave of the average value Y _ave, hue data H of the luminance data Y for each _block, and the average of the chroma data _{Find the} value C _ave .

Therefore, the values Y _ave , H _ave , C _ave , and v _ijbmax calculated as described above are supplied to the histogram generator 12 as representative values of each block. The strogram generator 12 adds 1 to the i-th block SH [i _b ] of the horizontal flesh color block histogram SH on the memory 13 as shown in FIG. 8 (SH [i _b ] = SH Along with [i _b ] +1), 1 is added to the j-th block SV [j _b ] of the skin color block histogram SV in the vertical direction (SV [j _b ] = SV [j _b ] +1). In this manner, a horizontal flesh color block histogram SV [i _b ] and a vertical flesh color block histogram SV [j _b ] are generated on the memory 13.

The intra-block feature quantity calculator 31 obtains the maximum value v _ijbmax in the block by the calculation shown in _Expression 8 for the vertical differential value v _ij obtained by the vertical luminance change detector 21.

Accordingly, the maximum value v _ijbmax is also output to the memory 13 as a representative value of each block, and the differential value v _ibjb is _{stored in} the i-th block DH [i _b ] of the horizontal differential histogram DH on the memory 13. Addition is made (DH [i _b ] = DH [i _b ] + v _ibjb ), and the differential value v _ibjb is added to the j th block DV [j _b ] of the skin color pixel histogram DV in the vertical direction (DV [j _b ] = DV [j _b ] + v _ibjb ). In this way, the horizontal differential histogram DH [i _b ] and the vertical differential histogram DV [j _b ] in units of blocks are also generated on the memory 13.

The face area determination unit 14 includes a flesh color block histogram SV [i _b ], a flesh color block histogram SV [j _b ], a differential histogram DH [i _b ], and a differential histogram DV [j generated on the memory 13. _b ] is detected in the same manner as in the second embodiment described above, and a skin color block is detected, and it is determined whether or not it is a face region based on the ratio of the detected skin color block size and vertical and horizontal lengths. Then, the block number _i bm skin color blocks determined as the face _region, i _bp, j _bm, the center position of the _{j bp i m, i p,} j m, the _{j p i m = i bm (} 2m i +1) + m _i
i _p = i _bp (2m _i +1) + m _i
j _m = j _bm (2m _j +1) + m _j
j _p = j _bp (2m _j +1) + m _j
The obtained center positions i _m , i _p , j _m , and j _p are used as coordinates indicating the boundary of the face area, and are output as face area information.

  The operation of the skin area detection unit 3 configured as described above will be described.

The luminance signal Y _d and the color difference signals U _d and V _d of the image P digitally converted by the A / D converter 101 are input to the coordinate converter 11.

The coordinate converter 11 converts the input luminance signal Y _d and color difference signals U _d and V _d into luminance data Y, hue data H, and chroma data C, and converts the converted luminance data Y, hue data H, The chroma data C is supplied to the in-block feature quantity calculator 31 and the luminance data Y is supplied to the vertical luminance change detector 21.

The vertical luminance change detector 21 calculates a differential value v _ij in the vertical direction of the luminance data Y from the coordinate converter 11 and supplies the calculated differential value v _ij to the in-block feature value calculator 31.

The intra-block feature quantity calculator 31 divides the luminance data Y, hue data H, and chroma data C from the coordinate converter 11 into blocks, and the average value Y _ave of the luminance data Y for each block, the hue The average value H _ave of the data H and the average value C _ave of the chroma data are obtained, and the obtained average value H _ave of the hue data H and the average value C _ave of the chroma data H are supplied to the histogram generator 12.

The histogram generator 12 determines whether the average values Y _ave , H _ave , and C _ave of the luminance data Y, hue data H, and chroma data C from the coordinate converter 11 satisfy the skin color condition. As a result, the skin color block histogram SH [i] in the horizontal direction and the skin color block histogram SV [j] in the vertical direction are generated on the memory 13.

The intra-block feature quantity calculator 31 _obtains the maximum value v _ibjbmax in the block for the differential value v _ij from the vertical luminance change detector 21, and the obtained maximum value v _ibjbmax in the block is obtained. Output to the memory 13. As a result, a horizontal differential histogram DH [i _b ] and a vertical differential histogram DV [j _b ] in units of blocks are also generated on the memory 13.

The face area determination unit 14 detects, as a skin area, an area where skin colors are dense from the skin color block histogram SH [i _b ] and the skin color block histogram SV [j _b ] generated on the memory 13 in units of blocks. Whether or not it is a face area is determined based on the ratio of the size of the area and the length and width. If the determination result is a face area, the face area information i _m is generated based on the differential histogram DH [i _b ] and the differential histogram DV [j _b ] in block units generated on the memory 13. , I _p , j _m , j _p are output.

  As described above, in this embodiment, the image is divided into blocks, the skin color block is detected based on the average color distribution in the block, and the size of the detected skin color block and the ratio of the vertical and horizontal lengths are determined. Since the face area is detected, analysis of the person's skin color area and detection of the person's face area can be performed quickly and accurately.

In the present embodiment, the histogram generator 12 generates the skin color block histogram SH [i _b ] in the horizontal direction and the skin color block histogram SV [j _b ] in the vertical direction on the memory 13. Only the flesh color block histogram SH [i _b ] in the direction may be generated on the memory 13.

In the present embodiment, the vertical luminance change detector 21 is provided. However, for example, in the configuration of the skin region detection unit 1 shown in FIG. An in-block feature quantity calculator 31 may be provided after the converter 11. In this case, a face area is detected only from the horizontal flesh color block histogram SH [i _b ] and the vertical flesh color block histogram SV [j _b ], or from the horizontal flesh color block histogram SH [i _b ]. Become.

  Next, an image processing apparatus according to a fifth embodiment of the present invention is described.

  The image processing apparatus applies the skin area detection unit 2 of the image processing apparatus according to the third embodiment described above. For example, as shown in FIG. 9, the skin area detection unit shown in FIG. 2, a contour enhancer 4 that performs a contour enhancement process by control based on information on the face region obtained by the skin region detection unit 2, and synchronizes data output from the contour enhancer 4 and input data. A delay unit 5 for taking the signal, an A / D converter 101, and a digital / analog converter (hereinafter referred to as a D / A converter) 102 are provided.

  In addition, the same code | symbol is attached | subjected to the location which shows the same operation | movement as the skin region detection part 2 shown in the said FIG. 6, and the detailed description is abbreviate | omitted.

The skin area detection unit 2 receives the luminance signal Y _d and the color difference signals U _d and V _d digitally converted by the A / D converter 101, and the skin area detection unit 2 receives the input as described above. luminance signal _{Y d} and the color difference signals _U d, detects a face area from _{V d.} The detected face area information i _m , i _p , j _m , and j _p , that is, four coordinates indicating the coordinates of the boundary of the face area are supplied to the contour enhancer 4.

Here, when the skin region detection unit 2 determines that the pixel corresponding to the input luminance signal Y _d and the color difference signals U _d and V _d is a skin color, the histogram generator of the skin region detection unit 2 12 supplies the contour enhancer 4 with a skin color detection signal cntl ₁ indicating whether or not a skin color has been detected.

The aforementioned contour enhancement circuit 4, the luminance signal Y _d which is digitally converted by the A / D converter 101 is inputted. The contour enhancer 4 performs contour enhancement processing on the input luminance signal Y _d and supplies the luminance signal Y _dd subjected to contour enhancement processing to the D / A converter 102.

At this time, face area information i _m , i _p , j _m , and j _p are supplied from the skin area detection unit 2, and a skin color detection signal cntl ₁ is supplied, and corresponds to the input luminance signal Y _d . When the pixel is present in the area indicated by the information i _m , i _p , j _m , and j _p of the face area, the edge enhancer 4 performs the edge enhancement process on the luminance signal Y _d. Reduce the degree of emphasis.

Specifically, first, information i _m of the face region supplied from the skin area detection section 2 to the edge enhancement unit _{4, i p, j m,} j p , rather than an image temporally image currently being input It is obtained from one past image. On the other hand, the skin color detection signal cntl ₁ supplied from the skin region detection unit 2 to the contour enhancer ₄ is information relating to the color of each pixel of the currently input image. In other words, the contour enhancer 4 adjusts the degree of enhancement based on face area information obtained from past images and color information in units of pixels of the currently input image.

  For example, the contour enhancer 4 includes a coefficient calculator 41, a coefficient smoother 42, an enhancement signal generator 43, and an adder 45, as shown in FIG. A conventional contour enhancer for adding the differential is applied.

The coefficient calculator 41 is supplied with the face area information i _m , i _p , j _m , j _p and the skin color detection signal cntl ₁ from the skin area detection unit 2. The coefficient calculator 41 obtains a coefficient c _ij corresponding to the position (i, j) on the image from these pieces of information.

That is, when the position (i, j) on the image that is the current processing target is included in the face area (i _m <i <i _p , j _m <j <j _p ), “1 "Is not included in the face area, the function f ₀ (i, j) whose output value is" 0 "and the skin color detection signal cntl ₁ indicate that the current pixel is skin color (Cntl ₁ = 1) has a function f ₁ (cntl ₁ ) whose output value is “1”, and when it indicates that it is not a skin color (cntl ₁ = 0), the facial region A coefficient c _ij corresponding to the position (i, j) on the image from the information i _m , i _p , j _m , j _p and the skin color detection signal cntl ₁ ,
c _ij = 1−f ₀ (i, j) f ₁ (cntl ₁ )
Is obtained by the following calculation.

Accordingly, the coefficient c _ij is “0” only when the position (i, j) on the current processing target image is included in the face area and the color is skin color, In this case, “1” is set.

The coefficient c _ij obtained in this way is supplied to the coefficient smoother 42. The coefficient smoother 42, in order to alleviate the discontinuity of the function _f 0 (i, j) and the function _{f 1} (cntl 1), the average value _{c dij} in the vicinity of the coefficients _{c ij.}

That is, the number N of pixels in the neighborhood system includes constants n _i and n _j representing the size of the neighborhood system for _obtaining the average value c _dij , the number of pixels in the horizontal direction (2n _i +1) constituting the neighborhood system, and With the number of pixels in the vertical direction (2n _j +1)
N = (2n _i +1) (2n _j +1)
It can be expressed.

Therefore, the coefficient smoother 42 _{obtains the} average value c _{dij in} the vicinity of the coefficient c _ij by the calculation shown in _Expression 9 with the number N of pixels in the vicinity system.

Then, the coefficient smoother 42 _{supplies the} average value c _{dij in} the vicinity of the obtained coefficient c _{ij to} the enhancement signal generator 43 as a smoothing correction coefficient c _dij .

The enhancement signal generator 43 is supplied with the smoothing correction coefficient c _dij from the coefficient smoother 42 and the luminance signal Y _d digitally converted by the A / D converter 101.

The enhancement signal generator 43, by multiplying the luminance signal Y smoothing correction coefficient from the coefficient smoother 42 in the second derivative of the _d c _dij from the A / D converter 101, generating an enhanced signal dY To do.

That is, when the luminance signal Y _d corresponding to the smoothing correction coefficient c _dj is represented by the luminance signal Y _d (i, j), the enhancement signal dY has a constant di,
dY = c _dij (2Y _d (i, j) −Y _d (i + di, j)
-Y _d (i-di, j))
Is obtained by the following calculation. At this time, an appropriate delay is applied to synchronize the input data. Then, the enhancement signal generator 43 supplies the generated enhancement signal dY to the adder 45.

The aforementioned adder 45, along with enhanced signal dY is supplied from the enhancement signal generator 43, the luminance signal Y _d from the A / D converter 101 is supplied.

The adder 45 adds the enhancement signal dY from the enhancement signal generator 43 and the luminance signal Y _d (i, j) corresponding to the smoothing correction coefficient c _dij from the A / D converter 101, The addition result is output as a luminance signal Y _dd that has undergone contour enhancement processing.

That is, the luminance signal Y _dd (i, j) obtained by performing the edge enhancement process on the luminance signal Y _d (i, j) at the position (i, j) on the current processing target image is
Y _dd (i, j) = Y _d (i, j) + dY
Is obtained by the following calculation.

As described above, the contour enhancer ₄ controls the degree of enhancement for each pixel based on the face area information i _m , i _p , j _m , j _p from the skin area detection unit 2 and the skin color detection signal cntl ₁ , performing edge enhancement processing to the input luminance signal Y _d. Then, the luminance signal Y _dd subjected to the contour enhancement process is supplied to the D / A converter 102. Here, in the above-described edge enhancement processing, an appropriate delay is applied in order to synchronize input data.

On the other hand, the color difference signals U _d and V _d digitally converted by the A / D converter 101 are input to the delay unit 5. The delay unit 5 delays the input color difference signals U _d and V _d and synchronizes with the luminance signal Y _dd subjected to the contour enhancement process by the contour enhancer 4. The color difference signals U _d and V _d delayed by the delay unit 5 are supplied to the D / A converter 102.

Accordingly, the D / A converter 102 includes the luminance signal Y _dd that has been subjected to the contour enhancement process by the contour enhancer 4 and the color difference signal U _d that is synchronized with the luminance signal Y _dd by the delay device 5. the _{V d} to analog conversion, and outputs the luminance signal _{Y da,} the color difference signals _U a, as _{V a.}

  The operation of the image processing apparatus configured as described above will be described.

The A / D converter 101 digitally converts the input luminance signal Y _a and the color difference signals U _a and V _a, and converts the digitally converted luminance signal Y _d and the color difference signals U _d and V _d to the skin region detection unit 2. At the same time, the luminance signal Y _d is supplied to the contour enhancer 4, and the color difference signals U _d and V _d are supplied to the delay unit 5.

The skin area detection unit 2 detects a skin area from the luminance signal Y _d and the color difference signals U _d and V _d from the A / D converter 101 and outputs a skin color detection signal cntl ₁ indicating the detection result to the contour enhancer 4. To supply. Further, the face area is detected by analyzing the detected skin area, and information i _m , i _p , j _m , j _p of the detected face area is supplied to the contour enhancer 4.

The edge enhancement unit 4 is subjected to edge enhancement processing on the luminance signal Y _d from the A / D converter 101, and supplies the luminance signal Y _da subjected to edge enhancement processing to the D / A converter 102.

At this time, the area corresponding to the face area information i _m , i _p , j _m , and j _p from the skin area detection unit 2 includes a pixel corresponding to the luminance signal Y _d that is the current processing target. If the skin color detection signal cntl ₁ from the skin region detection unit 2 indicates that the pixel to be processed is a skin color, the contour enhancer 4 determines the degree of enhancement. the by reducing subjected to edge enhancement processing to the luminance signal Y _d, and supplies the luminance signal Y _da subjected to edge enhancement processing to the D / a converter 102.

The delay unit 5 delays the color difference signals U _a and V _a from the A / D converter 101 to synchronize with the luminance signal Y _da subjected to the contour enhancement process by the contour enhancer 4 and delays it. The color difference signals U _a and V _a are supplied to the D / A converter 102.

The D / A converter 102 converts the luminance signal Y _dd from the contour enhancer 4 and the color difference signals U _d and V _d from the delay unit 5 into analog signals, thereby _obtaining the luminance signal Y _ad and the color difference signals U _a , and outputs it as the V _a.

  As described above, in this embodiment, the skin area detection unit 2 shown in FIG. 6 is applied, and information on the face area with respect to the past image obtained by the skin area detection unit 2 and the currently input image are displayed. Since the contour enhancement process is performed by adjusting the degree of enhancement using the skin color detection signal in pixel units, it is possible to enhance other regions without enhancing the human face region. Therefore, the sharpness of the image can be improved.

  Next, an image processing apparatus according to the sixth embodiment of the present invention is described.

As shown in FIG. 11, the image processing apparatus has a luminance signal Y _d and a color difference signal U digitally converted by the A / D converter 101 in addition to the configuration of the image processing apparatus according to the fifth embodiment described above. _d, an image memory 6 for storing the V _d, the coordinate converter 7 for converting the luminance signals stored in the image memory 6 Y _d and the color difference signals U _d, the V _d to 3 attribute data of the color, the coordinate transformation And a flesh color pixel detector 8 for detecting whether or not the three attribute data converted by the device 7 satisfies the flesh color condition.

Further, in the image processing apparatus, a skin color detection signal cntl ₁ from the skin area detection unit 2 is not output, and the skin color detection signal cntl ₁ from the skin color pixel detector 8 is output to the edge enhancement unit 4 Become. Further, the luminance enhancer Y _d stored in the image memory 6 is supplied to the contour enhancer 4, and the color difference signals U _d and V _d stored in the image memory 6 are also supplied to the delay unit 5. It becomes.

  In addition, the same code | symbol is attached | subjected to the location which shows the same operation | movement as the image processing apparatus shown in the said FIG. 9, and the detailed description is abbreviate | omitted.

That is, in this embodiment, by using the image memory 6, the luminance signal inputted _{Y d} and the color difference signals _U d, the _{V d,} information _i m in the skin area detection unit 2 face _area, i _{p, j} m, delay amount corresponding to the time required to detect the j _p, for example, provides a delay of one pixel. The luminance signal Y _d and the color difference signals U _d , V _d delayed by the image memory 6 are supplied to the coordinate converter 7, and the luminance signal Y _d is supplied to the contour enhancer 4, and the color difference signal U _d and V _d are supplied to the delay unit 5.

The coordinate converter 7 converts the luminance signal Y _d and the color difference signals U _d and V _d from the image memory 6 into luminance data Y, hue data H, and chroma data C, and converts the converted luminance data Y and hue. Data H and chroma data C are supplied to the flesh color pixel detector 8. The skin color pixel detector 8 is used when the luminance data Y, the hue data H, and the chroma data C from the coordinate converter 7 are skin color conditions, that is, when the histogram generator 12 detects a skin region. Conditions Y _min <Y <Y _max
H _min <H <H _max
C _min <C <C _max
It is judged whether or not Therefore, the determination result is supplied to the contour enhancer 4 as the skin color detection signal cntl ₁ .

On the other hand, the face area information i _m , i _p , j _m , and j _p detected by the skin area detection unit 2 is supplied to the contour enhancer 4.

  The operation of the image processing apparatus configured as described above will be described.

The A / D converter 101 digitally converts the input luminance signal Y _a and the color difference signals U _a and V _a, and converts the digitally converted luminance signal Y _d and the color difference signals U _d and V _d to the skin region detection unit 2. At the same time, it is supplied to the image memory 6.

The image memory 6, the luminance signal from the A / D converter 101 Y _d and the color difference signals U _d, giving a delay amount of one pixel in V _d, luminance signal to give a delay amount Y _d and the color difference signals U _d and V _d are supplied to the coordinate converter 7. The image memory 6 supplies the luminance signal Y _d to the contour enhancer 4 and supplies the color difference signals U _d and V _d to the delay unit 5.

The coordinate converter 7 converts the luminance signal Y _d and the color difference signals U _d and V _d from the image memory 6 into luminance data Y, hue data H, and chroma data C, and converts the converted luminance data Y and hue. Data H and chroma data C are supplied to the flesh color pixel detector 8. The skin color pixel detector 8 determines whether the luminance data Y, hue data H, and chroma data C from the coordinate converter 7 satisfy the skin color condition, and the determination result is used as the skin color detection signal cntl. ₁ is supplied to the contour enhancer 4.

On the other hand, the skin area detection unit 2 detects a skin area from the luminance signal Y _d and the color difference signals U _d and V _d from the A / D converter 101, and detects the face area by analyzing the detected skin area. Then, the detected face area information i _m , i _p , j _m , and j _p are supplied to the contour enhancer 4. The contour enhancer 4 performs a contour enhancement process on the luminance signal Y _d from the image memory 6 and supplies the D / A converter 102 with the luminance signal Y _da subjected to the contour enhancement process.

At this time, the area corresponding to the face area information i _m , i _p , j _m , and j _p from the skin area detection unit 2 includes a pixel corresponding to the luminance signal Y _d that is the current processing target. If the skin color detection signal cntl ₁ from the skin region detection unit 2 indicates that the pixel to be processed is a skin color, the contour enhancer 4 determines the degree of enhancement. the by reducing subjected to edge enhancement processing to the luminance signal Y _d, and supplies the luminance signal Y _da subjected to edge enhancement processing to the D / a converter 102.

Further, the delay unit 5 delays the color difference signals U _a and V _a from the image memory 6 to synchronize with the luminance signal Y _da subjected to the contour emphasis processing by the contour enhancer 4 and delays it. The color difference signals U _a and V _a are supplied to the D / A converter 102.

The D / A converter 102 converts the luminance signal Y _dd from the contour enhancer 4 and the color difference signals U _d and V _d from the delay unit 5 into analog signals, thereby _obtaining the luminance signal Y _ad and the color difference signals U _a , and outputs it as the V _a.

As described above, the face area information i _m , i _p , j _m , and j _p obtained by the skin area detection unit 2 is obtained from an image that is actually subjected to the contour emphasis process by the contour enhancer 4. is there. Therefore, since the more accurate position of the face area can be reflected in the contour enhancement process, the sharpness of the image can be further improved.

  Next, an image processing apparatus according to the seventh embodiment of the present invention is described.

As shown in FIG. 12, the image processing apparatus has a luminance signal Y _d and a color difference signal U digitally converted by the A / D converter 101 in addition to the configuration of the image processing apparatus according to the seventh embodiment. _d, from the V _d and the scene change detector 9 for detecting a scene change, the information of the face region obtained by the skin area detection unit 2 based on the detection result of the scene change detector 9 i _m, i _p, j _{and a} face area smoother 10 that performs temporal smoothing on _{m 1} and j _p .

In the image processing apparatus, the face area information i _m , i _p , j _m , and j _p obtained by the skin area detection unit 2 is supplied to the contour enhancer 4 via the face area smoother 10. The Further, in the above image processing apparatus, the delay amount in the image memory 6 is set to a delay amount that is twice the time required for the skin region detection unit 2 to detect the facial region information, and is obtained by the skin region detection unit 2. The face area information i _m , i _p , j _m , and j _p are completely associated with the image subjected to the contour emphasis processing by the contour enhancer 4.

  In addition, the same code | symbol is attached | subjected to the location which shows the same operation | movement as the image processing apparatus shown in the said FIG. 11, and the detailed description is abbreviate | omitted.

That is, the scene change detector 9 includes the image corresponding to the input luminance signal Y _d and the color difference signals U _d and V _d , and the face area information i _m , i _p , j obtained by the skin area detection unit 2. _m, to detect whether a scene change occurs between the image corresponding to the j _p.

  More specifically, for example, when the time required for detecting information on the face area by the skin area detection unit 2 is one image, the scene change detector 9 uses the edge enhancer 4 to perform edge enhancement processing. It is detected whether or not a scene change has occurred between an image subjected to the above and an image input one image earlier in time.

  For detection of such a scene change, a conventionally used detection method can be applied.

Therefore, the detection result of the scene change detector 9 is supplied to the face area smoother 10 as the control signal cntl ₂ .

The face area smoother 10 temporally smoothes the spatial position of the face area information i _m , i _p , j _m , and j _p obtained by the skin area detection unit 2 to obtain the face area information i. _dm , i _dp , j _dm , and j _dp are supplied to the contour _enhancer 4.

For example, as the temporal smoothing process, the face region boundary coordinates l _n (= i _m , i _p , j _m , j _p ) calculated from the image at time n and the contour enhancement process for the image at time n are used. With boundary coordinates l _dn (= i _dm , i _dp , j _dm , j _dp ) and the contribution rate p of the boundary coordinates obtained at the current time,
l _dn = pl _n + (1-p) l _{dn −1}
Use recursive filtering.

At this time, when the control signal cntl ₂ from the scene change detector 9 indicates the occurrence of a scene change, the face region smoother 10 does not perform the temporal smoothing process as described above, and detects the skin region. The face area information i _m , i _p , j _m , and j _p obtained by the unit 2 are supplied to the contour enhancer 4 as they are.

  The operation of the image processing apparatus configured as described above will be described.

The A / D converter 101 digitally converts the input luminance signal Y _a and the color difference signals U _a and V _a, and converts the digitally converted luminance signal Y _d and the color difference signals U _d and V _d into the skin region detection unit 2. This is supplied to the image memory 6 and the scene change detector 9, respectively.

The image memory 6, the luminance signal from the A / D converter 101 Y _d and the color difference signals U _d, giving a delay amount of one pixel in V _d, luminance signal to give a delay amount Y _d and the color difference signals U _d and V _d are supplied to the coordinate converter 7. The image memory 6 supplies the luminance signal Y _d to the contour enhancer 4 and supplies the color difference signals U _d and V _d to the delay unit 5.

The coordinate converter 7 converts the luminance signal Y _d and the color difference signals U _d and V _d from the image memory 6 into luminance data Y, hue data H, and chroma data C, and converts the converted luminance data Y and hue. Data H and chroma data C are supplied to the flesh color pixel detector 8. The skin color pixel detector 8 determines whether the luminance data Y, hue data H, and chroma data C from the coordinate converter 7 satisfy the skin color condition, and the determination result is used as the skin color detection signal cntl. ₁ is supplied to the contour enhancer 4.

On the other hand, the skin area detection unit 2 detects a skin area from the luminance signal Y _d and the color difference signals U _d and V _d from the A / D converter 101, and detects the face area by analyzing the detected skin area. Then, the detected face area information i _m , i _p , j _m , and j _p are supplied to the face area smoother 10.

The scene change detector 9 also includes an image corresponding to the luminance signal Y _d and the color difference signals U _d and V _d from the A / D converter 101, and information on the face area obtained by the skin area detection unit 2. It is detected whether or not a scene change has occurred between images corresponding to i _m , i _p , j _m , and j _p , and the detection result is supplied to the face region smoother 10 as a control signal cntl ₂ .

Based on the control signal cntl 2 from the scene change detector 9, the face area smoother 10 spatially stores the face area information i _m , i _p , j _m , and j _p from the skin area detector 2. The smooth positions are smoothed in time, and the smoothed face area information i _dm , i _dp , j _dm , j _dp is supplied to the contour _enhancer 4.

The contour enhancer 4 performs a contour enhancement process on the luminance signal Y _d from the image memory 6 and supplies the D / A converter 102 with the luminance signal Y _da subjected to the contour enhancement process.

At this time, a pixel corresponding to the luminance signal Y _d that is the current processing target is present in the area indicated by the face area information i _dm , i _dp , j _dm , and j _dp from the face area smoother 10. If it is included and the skin color detection signal cntl ₁ from the skin region detection unit 2 indicates that the pixel that is the current processing target is skin color, the contour enhancer 4 performing edge enhancement processing to the luminance signal Y _d by reducing the degree and supplies the luminance signal Y _da subjected to edge enhancement processing to the D / a converter 102.

Further, the delay unit 5 delays the color difference signals U _a and V _a from the image memory 6 to synchronize with the luminance signal Y _da subjected to the contour emphasis processing by the contour enhancer 4 and delays it. The color difference signals U _a and V _a are supplied to the D / A converter 102.

The D / A converter 102 converts the luminance signal Y _dd from the contour enhancer 4 and the color difference signals U _d and V _d from the delay unit 5 into analog signals, thereby _obtaining the luminance signal Y _ad and the color difference signals U _a , and outputs it as the V _a.

As described above, in the present embodiment, when the scenes are the same, temporal smoothing is performed on the face area information i _m , i _p , j _m , and j _p obtained by the skin area detection unit 2. Since no processing is performed, temporal variation of the face area in the same scene can be reduced. Therefore, the sharpness of the image can be improved.

  In the above-described embodiment, the scene change detector 9 and the face area smoother 10 are added to the configuration of the image processing apparatus according to the above-described seventh embodiment, but the above-described fifth embodiment. The scene change detector 9 and the face area smoother 10 may be added to the configuration of the image processing apparatus according to FIG.

  Next, an image processing apparatus according to an eighth embodiment of the present invention is described.

  In the image processing apparatus, the contour enhancer 4 of the image processing apparatus according to the fifth embodiment described above is replaced with a gradation converter.

  As shown in FIG. 13, the gradation converter includes a coefficient calculator 41, a coefficient smoother 42, a level converter 403, and a weighted adder 404.

  In addition, the same code | symbol is attached | subjected to the location which shows the same operation | movement as the outline enhancer shown in the said FIG. 10, and the detailed description is abbreviate | omitted.

The coefficient calculator 41 and the coefficient smoother 42 are the same as those provided in the contour _enhancer 4 described above, and the smoothing correction coefficient c _dij is _calculated in the same manner as in the case of the contour _{enhancer 4} . The calculated smoothing correction coefficient c _dij is supplied to the level converter 403.

The level converter 403 performs a gradation conversion processing on the input luminance signal Y _d, and generates a luminance signal Y _ce subjected to the gradation conversion processing.

The gradation conversion has a conversion function g and Y _ce = g (Y _d )
The conversion function g is, for example, a non-linear monotonically increasing function as shown in FIG. The shape depends on the properties of the image to be converted. As a method of constructing the conversion function g from an image in this manner, a conventionally used method can be applied.

The luminance signal Y _{ce that} has been subjected to the gradation conversion processing by the level converter 403 is supplied to the weighted adder 404. The weighted adder 404 is supplied with the luminance signal Y _ce from the level converter 403 and the input luminance signal Y _d . Also, the smoothing correction coefficient c _dij obtained by the coefficient smoother 42 is supplied to the weighted adder 404 via the level converter 403.

The weighted adder 404 performs a luminance signal Y _d input, the weighted addition of luminance signal Y _ce from the level converter 403. As the weight used at this time, the smoothing correction coefficient c _dij obtained by the coefficient smoother 42 is used.

That is, the luminance signal Y _{dd obtained by performing} gradation conversion processing on the input luminance signal Y _d is
Y _dd = (1−c _dij ) Y _d + c _dij Y _ce
Is obtained by the following calculation.

This is because when the current pixel is located in the face area and is a skin color, that is, when the smoothing correction coefficient c _dj is small, the weight of the input luminance signal Y _d increases. In other cases, the weight of the luminance signal Y _ce that has been subjected to the gradation conversion processing by the level converter 403 is increased.

As described above, in this embodiment, the tone modulation processing is performed by adjusting the weight by the smoothing correction coefficient c _dj based on the skin color detection signal cntl ₁ of the pixel of the image being input. By converting the gradation of other areas without converting the gradation of the face area, the contrast of the image can be improved.

In the above-described embodiment, the contour enhancer 4 of the image processing apparatus according to the fifth embodiment described above is replaced with a gradation converter, but the image processing apparatus according to the sixth embodiment described above is replaced. The contour enhancer 4 may be replaced with a gradation converter. In this case, the face area information i _m , i _p , j _m , and j _p obtained by the skin area detection unit 2 is calculated from an image that is actually subjected to gradation conversion processing, and thus more accurately The position of the area can be reflected in the gradation conversion process. Accordingly, the contrast of the image can be further improved.

  Further, the contour enhancer 4 of the image processing apparatus according to the seventh embodiment described above may be replaced with a gradation converter. In this case, the temporal variation of the face area in the same scene can be mitigated, so that the contrast of the image can be further improved.

  In the fifth to eighth embodiments described above, the skin region detection unit of the image processing apparatus according to the third embodiment is applied. However, the first, second, and fourth embodiments are applied. The skin region detection unit of the image processing apparatus according to the above may be applied.

It is a block diagram which shows the structure of the skin area | region detection part of the image processing apparatus which concerns on 1st Example of this invention. It is a figure showing the scanning direction of an image. It is a figure for demonstrating the production | generation process of the skin color pixel histogram of a horizontal direction. It is a figure for demonstrating the analysis process of the said skin color pixel histogram. It is a figure for demonstrating the production | generation process of the skin color pixel histogram of a horizontal direction and a vertical direction. It is a block diagram which shows the structure of the skin area | region detection part of the image processing apparatus which concerns on 3rd Example of this invention. It is a block diagram which shows the structure of the skin area | region detection part of the image processing apparatus which concerns on the 4th Example of this invention. It is a figure for demonstrating the production | generation process of the skin color block histogram of a horizontal direction and a vertical direction. It is a block diagram which shows the structure of the image processing apparatus which concerns on the 5th Example of this invention. It is a block diagram which shows the structure of the outline enhancer of the said image processing apparatus. It is a block diagram which shows the structure of the image processing apparatus which concerns on the 6th Example of this invention. It is a block diagram which shows the structure of the image processing apparatus which concerns on the 7th Example of this invention. It is a block diagram which shows the structure of the gradation converter of the said image processing apparatus. It is a figure which shows a gradation conversion function.

Explanation of symbols

  1, 2, 3 Skin region detection unit, 4 Outline enhancer, 5 Delay device, 6 Image memory, 7 Coordinate converter, 8 Skin color pixel detector, 9 Scene change detector, 10 Face region smoother, 11 Coordinate transformation , 12 Histogram generator, 13 Memory, 14 Face area determiner, 21 Vertical luminance change detector, 31 In-block feature quantity calculator

Claims (3)

An image processing apparatus comprising skin area detecting means for converting input human image data into digitized luminance data and color difference data, and detecting a skin area from the digitized luminance data and color difference data,
Conversion means for converting the luminance data and the color difference data into three attribute data of luminance data, hue data, and chroma data; and a vertical luminance change in the skin color region is detected from the luminance data obtained by the conversion means; Vertical luminance change detection means for generating a cumulative distribution for each position in the horizontal direction and a cumulative distribution for each position in the vertical direction, and skin color pixels are extracted from the three attribute data obtained by the conversion means. A flesh color pixel extracting means; a flesh color pixel distribution generating means for generating a spatial flesh color pixel distribution of the flesh color pixels extracted by the flesh color pixel extracting means; a flesh color pixel distribution generated by the flesh color pixel distribution generating means; Cumulative distribution of vertical luminance changes generated by the vertical luminance change detecting means for each position in the horizontal direction and cumulative distribution for each position in the vertical direction And the skin area detection means and a determining face area determination means a face area of a person image from,
Contour emphasizing means for performing contour emphasis processing on the luminance data of the input human image data;
A predetermined delay amount corresponding to the time required for the contour enhancement process is given to the color difference data of the input human image data, and the color difference data synchronized with the luminance data subjected to the contour enhancement process by the contour enhancement means is output. A delay means,
An image processing apparatus for controlling the contour emphasizing means based on skin area information detected by the skin area detecting means. Storage means for storing the human image data converted into the luminance data and the color difference data, and giving and outputting a predetermined delay amount corresponding to the time required for detecting the skin area information by the skin area detection means;
Conversion means for converting the human image data given the predetermined delay amount by the storage means into three attribute data of luminance data, hue data, and chroma data;
Detection means for detecting skin color data from the three attribute data obtained by the conversion means,
The delay means gives a predetermined delay amount corresponding to the time required for the contour emphasizing process to the color difference data of the person image data given the predetermined delay amount by the storage means,
The contour emphasizing unit performs contour emphasis processing on the luminance data of the person image data corresponding to the detection result detected by the detecting unit by control based on the skin region information detected by the skin region detecting unit. The image processing apparatus according to claim 1, wherein: Scene change detection means for detecting a scene change from the input person image data;
Only when the same scene continues based on the detection result of the scene change detection means, the skin where the spatial position of the skin area information detected by the skin area detection means is detected in the previous image by the skin area detection means Smoothing means for smoothing temporally according to the spatial position of the region information,
The contour emphasizing unit performs contour emphasis processing on luminance data of human image data corresponding to a detection result detected by the detecting unit by control based on skin area information from the smoothing unit. Item 6. The image processing apparatus according to Item 1.

Images