JP5205230B2 - Image processing device - Google Patents

Description

  The present invention relates to an image processing apparatus. For example, the present invention relates to an image processing apparatus capable of correcting a color in a video signal in an imaging device such as a digital camera or a display device such as a TV or a display.

  Japanese Laid-Open Patent Publication No. 2004-297698 (Patent Document 1) states that “[Problem] When correcting a specific color, the color correction amount depends on the shooting situation and the video to be shot without affecting the colors other than the correction target. [Solution Means] An imaging mode selection means, position data indicating the position of a predetermined color on the color difference plane, and a predetermined range centered on the position of the predetermined color are provided. Color convergence parameter storing color convergence parameter values including correction range setting data for setting as a correction range and convergence coefficient data for converging a specific color corresponding to the correction range to a position indicating a predetermined color Based on the shooting mode information selected by the storage means and the shooting mode selection means, the color convergence parameter for selecting and setting the color convergence parameter value of the corresponding specific color from the color convergence parameter storage means. And color convergence correction processing means for correcting a specific color in the video signal to a predetermined color by a correction amount calculated based on the color convergence parameter value set by the color convergence parameter setting means. Is described.

JP 2004-297698 A

  Conventionally, the color of an image cannot be completely reproduced due to the influence of the characteristics of a display such as a sensor such as a CCD or CMOS of a photographed camera or a liquid crystal panel or a plasma panel.

  In addition, since humans have a common image with respect to a specific target such as human skin or blue sky, a color (memory color) that is preferred may be different from an actual color.

  Such a function for correcting the color of an image is mainly performed by a signal processing circuit, and there is a method of correcting a specific color to a target color by changing the gain of RGB or rotating the hue. It has been done.

  However, with these methods, all colors are corrected, and even colors that are not intended to be corrected are corrected. In addition, since the stored color is caused by an image stored by a person, the correction level changes depending on the subject even when only a specific color is corrected.

  Therefore, there is a problem that a specific color in a specific subject has to be corrected.

  Incidentally, as an invention for correcting only a specific color of hue, there is the above-mentioned Japanese Patent Application Laid-Open No. 2004-297698 (Patent Document 1), but this known example describes that a subject is recognized and color correction is performed. Absent.

  In the present invention, for example, in order to correct a specific color of a specific subject, the subject is recognized by an object recognition unit capable of recognizing a specific object such as a person or the sky, and the specific color is obtained using information obtained from the subject. Correction only for An object is defined as one that can be recognized from the characteristics of the target. The details are as described in the claims.

  For example, the image processing apparatus of the present invention can recognize a specific object and can correct only a specific color. Therefore, by designating a target subject as a specific object, other colors are not affected. Only a specific color can be corrected for a target subject.

  The best mode for carrying out the invention is shown in the following examples.

FIG. 1 is a diagram showing a basic configuration in the present invention. Details will be described below.
As shown in FIG. 1, an object recognition unit 100 for recognizing a specific object from a video signal and a color correction unit 200 for correcting the color of the video using the recognized object information are configured.

The object recognition unit 100 detects an object detection unit 101 for detecting a specific object, an object region setting unit 102 for setting a range in the image of the detected object, and detects color information of the detected object. And object color detection means 103 for the purpose.
The object detection means 101 compares the input video with the predetermined feature information of the specific object and the image to detect the target subject, and outputs area information indicating the position and size of the specific object.

  As a method for detecting a subject, a method is generally used in which image data indicating the outline characteristics of an object is stored as feature information, and detection is performed by pattern matching with an input image. Detailed description is omitted here. To do.

  Further, as the area information, for example, a reference position coordinate that allows the position of the detected object in the video to be known and a difference value from the reference position to the area boundary so that the size of the area can be known are used.

By using such area information, it is possible to easily transmit the range in which the object exists irrespective of the area setting method of the object area setting means 102.
The object area setting means 102 sets area coordinate information so as to include an object detected based on the output information.

The area setting method takes into account the correlation information between the detected object and the surrounding pixels, and if it is set wider than the detected size, it can also detect the surrounding pixel information. The color correction area can be set in detail.
The object color detection means 103 detects correlation information between a color and an object using a pixel signal such as a luminance signal or a color difference signal from a set pixel area.

  The correlation information includes, for example, a simple average value of pixel signals in a region, a signal level difference that is weighted and added according to a position in the region, and an index that represents a spread such as dispersion. By using such correlation information, it is possible to set an optimum color correction region and target color for each subject even for the same object.

  The color correction unit 200 includes a correction color region control unit 201 for controlling a range of a color region to be corrected, a correction color region setting unit 202 for setting a region, and whether or not a pixel is a correction target. Correction color region determining means 203 for determining, target color setting means 204 for setting the target value and correction condition of the color to be corrected, and color signal correcting means for correcting the pixels determined to be corrected 205.

  The correction color area control unit 201 represents the reference position coordinates and size of the correction color area on the currently set color difference plane based on the correlation information between the object and the color detected by the object color detection unit 103. Control difference information.

  By adjusting the correction color area to the optimum setting in accordance with the object detection timing, when taking a still image with a still camera, it is detected at the timing when the shutter is pressed, and the input video such as a movie camera or TV is displayed. In the case of a moving image, it is possible to perform optimum color correction according to the subject and the scene by detecting the frame unit.

  The correction color area setting means 202 sets the boundary of the correction color area on the color difference plane according to the control information of the position and size of the correction color area set by the correction color area control means 201.

FIG. 2 shows a color difference plane, in which the horizontal axis represents Cb, which is one of the equation difference signals, and the vertical axis represents Cr, which is one of the equation difference signals. The symbol in the figure represents the color that represents the broken line axis drawn from the origin to the symbol, and the meaning of the symbol is as shown in the figure. Further, the correction color region in the figure represents the range of colors to be corrected, and the target color represents the target color after correction.
As shown in FIG. 2, it is possible to correct only a specific color having a close phase by setting the correction color region to a rhombus region or an ellipse region.

  The correction color area determination unit 203 determines whether or not the pixel color is included in the correction color area from the coordinate information of the correction color area set by the correction color area setting unit 202.

  When the correction color region setting means 202 uses a region as a rhombus and boundary information as vertex coordinates, the coordinates on the color difference plane of the discrimination target pixel and the one coordinate on the region are the same for a straight line using a specific vertex coordinate. By repeating the determination of being on the opposite side or on the opposite side, processing by a simple calculation becomes possible. At this time, if the above-described discrimination is performed using two straight lines that can be drawn on two points that do not form a rhombus side with respect to the discrimination target pixel and the remaining vertex, the side of the rhombus closest to the discrimination target pixel Therefore, it is possible to easily determine whether or not the determination target pixel is included in the region by performing the same determination using the last determined side and the point on the region.

  The target color setting unit 204 sets target value coordinates and a correction rate in order to bring the color to be corrected close to the target color according to the information output from the object recognition unit 100. For setting the target value coordinates and the correction rate, a method of preliminarily holding them as set values or a method of automatically setting them at a fixed ratio according to the correction color region can be considered. At that time, by setting the target value coordinates in the correction color area, it is possible to prevent the color reversal phenomenon between the uncorrected color and the corrected color, and thus correct the color appropriately without any adverse effects. I can do it.

  The color signal correction unit 205 approximates the target color by weighting and adding the correction factor to the target value coordinates set by the target color setting unit 204 and the color signal of the pixel to be processed. Correct as follows.

  At this time, ripples generated by performing a correction that suddenly changes in a part where the color changes smoothly, such as gradation, by calculating the correction rate so that the color farther from the target value does not work. Since the phenomenon can be suppressed, color correction can be appropriately performed without any harmful effects.

  In order to perform color correction only for a specific subject, when correcting only the detected area in the screen, the color is adjusted on the boundary of the detection area by nonlinearly controlling so that the correction factor becomes weaker as it approaches the boundary. It is possible to prevent sudden changes and unnatural appearance.

  As shown in FIG. 4, a lens 301 for imaging a subject on the image sensor surface, an image sensor 302 for converting the imaged light into an electrical signal, and converting the input electrical signal into an image. The signal processing unit 303 for performing processing, the recognition color correction unit 000 described in the first embodiment, and a recording unit 304 for recording a captured image.

  When photographing is started so that a target subject is reflected, light that has passed through the lens 301 such as a focus lens or a zoom lens is imaged on an image sensor surface such as a CCD or CMOS. The image sensor 302 performs photoelectric conversion on the imaged light so that the signal processing unit 303 can process it. The signal processing unit 303 converts the converted electrical signal into an image and performs various processes such as white balance and gamma correction.

  Next, in order to correct the specific color of the specific subject, the recognition color correction means sets in advance the characteristic information of the subject to be corrected and the target color to be corrected, and the subject color is set on the processed image. Detect whether or not it is reflected.

  When used with a surveillance camera, a human or moving object is set as a specific object, and a conspicuous color is intentionally set as the target color, so that it can be displayed in an easy-to-understand manner when an abnormality occurs.

  The video signal corrected as described above is compressed by the recording unit 304 and recorded on a recording medium such as a DVD or HDD.

  In addition, it is possible to display and record a video that is recorded as it is with conventional color reproduction, and color-corrected for a specific subject as necessary.

  The white balance of the entire screen may be disrupted depending on the shooting conditions, such as when two lights with different color temperatures, such as sunlight and fluorescent lights, are present in the screen at the same time or there is a portion where the signal is saturated in the screen. Thus, even when the target subject color is not preferable, the color correction can be automatically performed without concern for the photographer.

  FIG. 5 shows a configuration example when the present invention is applied to a TV as one of display devices. Details will be described below.

  As shown in FIG. 5, an antenna unit 401 for receiving broadcast waves, a tuner unit 402 for modulating broadcast waves, a decoder unit 403 for converting video signals and audio signals, and correcting video signals A video signal processing unit 404 for correcting the audio signal, an audio signal processing unit 405 for correcting the audio signal, the recognition color correcting unit 000 described in the first embodiment, a display unit 406 for displaying the video, an audio And a speaker unit 407 for outputting.

  First, the antenna unit 401 receives a broadcast wave transmitted from a broadcast station. The received broadcast wave is modulated by the tuner unit 402 and converted into a digital signal. The decoder unit 403 separates the modulated digital signal into a video signal and an audio signal and decodes them.

  The video signal processing unit 404 and the audio signal processing unit 405 perform correction according to the characteristics of the output device for each decoded signal.

  The recognition color correction means performs color correction when a target subject is recognized for the corrected video.

  By enabling the user to change the color correction setting, it is possible to view with an optimal color such as skin color or sky color according to the preference of each viewer.

  Finally, the corrected video is displayed on the display unit 406, and the corrected audio is reproduced on the speaker unit 407.

  As mentioned above, although preferred embodiment concerning this invention was described referring an accompanying drawing, it cannot be overemphasized that this invention is not limited to this example. It is obvious for those skilled in the art that various changes or modifications can be conceived within the scope of the technical idea described in the claims. It is understood that it belongs to. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. In addition, each of the above-described configurations may be configured such that some or all of them are configured by hardware, or are implemented by executing a program by a processor. Further, the control lines and information lines indicate what is considered necessary for the explanation, and not all the control lines and information lines on the product are necessarily shown. Actually, it may be considered that almost all the components are connected to each other.

It is a figure showing a basic composition figure. It is a figure showing the area | region of the color to correct | amend. It is a figure showing the effect which correct | amended the color with respect to the recognized object. It is a figure showing the example of a structure which applied this invention to the camera. It is a figure showing the example of a structure which applied this invention to TV.

Explanation of symbols

100 ・ ・ ・ Object recognition means
101 ... Object detection means
102 ... Object area setting means
103 ... Object color detection means
200 ... Color correction means
201 ... corrected color area control means
202 ... Correction color area setting means
203 ... Correction color area discrimination means
204 ... Target color setting means
205 ... Color signal correction means

Claims (6)

An image processing apparatus comprising: an object recognition unit that can recognize one or more specific objects; and a color correction unit that corrects the color of a video.
The object recognition means includes object detection means for detecting a specific object, and object area setting means for setting an area for an input image.
The color correction means includes a correction color area control means for controlling the position and width of a specific color area to be corrected, a correction color area setting means for setting a specific color area to be corrected, and a pixel color as a specific color. Correction color determination means for determining whether or not the image is included in the set area, target color setting means for setting a target color, and color signal correction means for correcting the color signal,
The object area setting means sets an area including the detected object according to the position of the object detected by the object detection means,
The color correction means is a line Umono nonlinear processing to the color information based on the correlation between the object and the color detected from the set area,
The correction color area control means changes the setting of the correction color area based on correlation information between the object and color information calculated from the set area,
The correction color area setting means sets a correction color area based on the setting determined by the correction color area control means,
The color determining unit determines whether or not the color of the input pixel is included in the correction color region calculated by the correction color region setting unit;
The color signal correcting unit corrects the color signal of the pixel so as to approach the target color set by the target color setting unit when it is determined that the color signal is included in the correction color region. An image processing apparatus. In claim 1 ,
The image processing apparatus according to claim 1, wherein the correction color area control unit adjusts a reference position and size of the correction color area in accordance with information from the object color detection unit. In claim 1 ,
The correction color area setting means sets the boundary of the correction color area on the color difference plane in accordance with the position and size of the correction color area adjusted by the correction color area control means. . In claim 1 ,
The image processing apparatus according to claim 1, wherein the correction color determination unit determines whether or not a pixel color is included in the correction color area from the coordinate information of the set correction color area. In claim 1 ,
The target color setting means sets a target value coordinate and a correction rate in order to bring the color to be corrected closer to the target color, and at this time, the target value coordinate is set in a correction color region. An image processing apparatus. In claim 1 ,
The color signal correcting means corrects the pixel so as to approach the target color by weighting and adding the correction rate to the set target value coordinates and the color signal, and the correction rate is calculated from the target value. An image processing apparatus that performs calculation so that a distant color does not work.

Images