WO2014080613A1 - 色補正装置、色補正方法および色補正用プログラム - Google Patents
色補正装置、色補正方法および色補正用プログラム Download PDFInfo
- Publication number
- WO2014080613A1 WO2014080613A1 PCT/JP2013/006776 JP2013006776W WO2014080613A1 WO 2014080613 A1 WO2014080613 A1 WO 2014080613A1 JP 2013006776 W JP2013006776 W JP 2013006776W WO 2014080613 A1 WO2014080613 A1 WO 2014080613A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- color
- moving object
- image
- region
- stable
- Prior art date
Links
- 238000012937 correction Methods 0.000 title claims abstract description 269
- 238000000034 method Methods 0.000 title claims description 66
- 230000008859 change Effects 0.000 claims abstract description 158
- 238000006243 chemical reaction Methods 0.000 claims abstract description 101
- 239000003086 colorant Substances 0.000 claims abstract description 23
- 238000001514 detection method Methods 0.000 claims abstract description 12
- 238000004364 calculation method Methods 0.000 claims description 94
- 238000005286 illumination Methods 0.000 claims description 34
- 230000008569 process Effects 0.000 claims description 30
- 238000003384 imaging method Methods 0.000 claims description 28
- 238000011161 development Methods 0.000 claims description 22
- 238000012545 processing Methods 0.000 claims description 20
- 230000036962 time dependent Effects 0.000 claims description 5
- 230000015572 biosynthetic process Effects 0.000 abstract 1
- 238000000605 extraction Methods 0.000 description 39
- 238000010586 diagram Methods 0.000 description 17
- 239000000284 extract Substances 0.000 description 13
- 238000005259 measurement Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 230000002776 aggregation Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000010365 information processing Effects 0.000 description 2
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/46—Colour picture communication systems
- H04N1/56—Processing of colour picture signals
- H04N1/60—Colour correction or control
- H04N1/6027—Correction or control of colour gradation or colour contrast
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/14—Picture signal circuitry for video frequency region
- H04N5/144—Movement detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/46—Colour picture communication systems
- H04N1/56—Processing of colour picture signals
- H04N1/60—Colour correction or control
- H04N1/6083—Colour correction or control controlled by factors external to the apparatus
- H04N1/6086—Colour correction or control controlled by factors external to the apparatus by scene illuminant, i.e. conditions at the time of picture capture, e.g. flash, optical filter used, evening, cloud, daylight, artificial lighting, white point measurement, colour temperature
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/46—Colour picture communication systems
- H04N1/56—Processing of colour picture signals
- H04N1/60—Colour correction or control
- H04N1/62—Retouching, i.e. modification of isolated colours only or in isolated picture areas only
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/10—Cameras or camera modules comprising electronic image sensors; Control thereof for generating image signals from different wavelengths
- H04N23/12—Cameras or camera modules comprising electronic image sensors; Control thereof for generating image signals from different wavelengths with one sensor only
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/80—Camera processing pipelines; Components thereof
- H04N23/84—Camera processing pipelines; Components thereof for processing colour signals
- H04N23/88—Camera processing pipelines; Components thereof for processing colour signals for colour balance, e.g. white-balance circuits or colour temperature control
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
Definitions
- the present invention relates to a color correction apparatus, a color correction method, and a color correction program that correct colors in an image.
- the methods for correcting the color of an image there is a method of performing color correction using a color chart in which a plurality of colors such as natural color, chromatic color, primary color, and gray scale are arranged.
- This method can correct the color of the image by obtaining a color correction parameter for the captured image based on the imaging result of the color chart.
- Non-Patent Document 1 describes an example of a luminance transfer function for associating objects between cameras having different colors. For example, in the image taken by the camera A, Mr. A who is wearing an outfit with a blue appearance may have a light blue outfit for the camera B due to a change in color. In this method, image features of blue and light blue clothes are associated with each other so that the same object can be recognized even if the colors are different between the cameras.
- Non-Patent Document 2 describes a luminance transfer function that takes into account fluctuations in the light source.
- Non-Patent Document 1 aims at associating objects between cameras with different colors, and assumes a light source change at each position in an image captured by one camera. Not. That is, Non-Patent Document 1 has no description about associating the same person in an image photographed by one camera in order to detect a light source change in the camera.
- Non-Patent Document 1 uses the timing when the color of the object changes when the camera that captures the object changes, so that it is not possible to detect a color change on a pixel in the same camera. Can not. This also applies to the technique described in Non-Patent Document 2.
- the present invention provides a color correction apparatus, a color correction method, and a color correction device that can easily correct a color changing for each area in an image without preparing a color chart for each area whose color changes due to a light source change or the like.
- An object is to provide a color correction program.
- the color correction apparatus includes a moving object detection unit that detects a moving object from a moving image whose shooting range is the same as the shooting range of the correction target image or includes the entire shooting range of the correction target image, and a moving object detection unit.
- Position-specific color information acquisition means for determining the identity of the moving object detected by the moving image in the moving image and acquiring the position of the moving object accompanying the movement and information indicating the color of the moving object in the image;
- Position-specific color information storage means for storing the position of the moving object associated with the movement of the moving object acquired by the position-specific color information acquisition means and information indicating the color of the moving object in the image in association with each other;
- a color stable area estimating means for estimating a color stable area, which is an area where the color is stable in the shooting range, based on information indicating the relationship between the position of the moving object and the color stored in the separate color information storage means; Stable This is a parameter for converting colors between color stable areas based on the estimation result by the area estimation means and the information indicating the relationship between the position and color of the moving object stored in the position-specific color information storage means.
- Color conversion parameter calculation means for calculating a color conversion parameter, and color correction means for correcting the color of the correction target image using the color conversion parameter
- the color correction method detects a moving object from a moving image whose shooting range is the same as the shooting range of the correction target image or includes the entire shooting range of the correction target image, and detects the moving object in the moving image of the detected moving object.
- the identity of the moving object is determined, the position of the moving object that accompanies the movement and the information indicating the color of the moving object in the image are acquired, stored in association with each other, and the moving object stored in the storage means Based on the information indicating the relationship between the position of the image and the color, the color stable region, which is a region where the color is stable in the shooting range, is estimated, the estimation result of the color stable region, and the position of the moving object stored in the storage unit Based on the information indicating the relationship between the color and the color, a color conversion parameter that is a parameter for converting the color between the color stable regions is calculated, and correction is performed using the calculated color conversion parameter between the color stable regions. Correct the color of the target image And wherein the Rukoto.
- the color correction program is a computer program for detecting a moving object from a moving image whose shooting range is the same as the shooting range of the correction target image or which includes the entire shooting range of the correction target image. Processing for determining identity of an object in a moving image, obtaining information indicating the position of the moving object associated with movement and information indicating the color of the moving object in the image, and storing them in association with each other; Based on the information indicating the relationship between the position of the moving object and the color stored in the image, the process of estimating the color stable region, which is the region where the color is stable in the shooting range, the estimation result of the color stable region, and the storage means Based on the stored information indicating the relationship between the position of the moving object and the color, processing for calculating a color conversion parameter, which is a parameter for converting the color between the color stable regions, and the calculated color reduction Using the color conversion parameters between the regions, characterized in that to execute a process for correcting the color of the correction target image.
- the color changing for each region in the image can be easily corrected.
- FIG. 1 is a block diagram illustrating a configuration example of a color correction apparatus according to a first embodiment of the present invention.
- the color correction apparatus shown in FIG. 1 includes a position-specific color change feature extraction unit 1, a color stable region estimation unit 2, a region-specific color correction parameter calculation unit 3, and a color correction unit 4.
- the position-specific color change feature extraction unit 1 extracts the color change feature of the image at each position in the shooting range.
- the position-specific color change feature extraction unit 1 extracts, for example, an area in which a moving object is captured (hereinafter referred to as a moving object area) from a moving image captured by a fixed camera, and tracks (tracks) the moving object.
- Information indicating the relationship between the position of the region and the color and the relationship between the position of the moving object region and the color change may be extracted. Further, the position-specific color change feature extraction unit 1 holds the extracted information in a storage unit (not shown).
- the information indicating the relationship between the position and color of the moving object area is, for example, information indicating the position in the shooting range and the color in the image at the position (for example, image feature amount) for the corresponding moving object area. May be associated with each other.
- the information indicating the relationship between the position in the shooting range and the color change is the color change amount in the image between the position in the shooting range and the adjacent position at the position for the corresponding moving object region. The information which matched with the information which shows may be sufficient.
- the position-specific color change feature extraction unit 1 may extract a color histogram as information indicating a color, for example. Further, the distance between the histograms may be calculated as information indicating the amount of color change. The distance between the histograms can be calculated using, for example, a histogram intersection. The calculation of the color change amount preferably includes direction information (that is, information indicating from which direction the moving object has entered a certain position). The histogram intersection described above can relate the difference of the histogram intersection to its position and direction.
- the position-specific color change feature extraction unit 1 performs tracking for a plurality of moving objects, and the relationship between the position and color of the moving object region in the shooting range, and the relationship between the position and color change of the moving object region in the shooting range.
- the information indicating is extracted.
- the color change characteristics of the image (information indicating what kind of color change has occurred between the adjacent areas) are extracted at the main positions in the shooting range.
- color correction for the image frame can be performed with only the tracking result for one moving object, but it is preferable to have tracking results for many moving objects in order to improve accuracy. This is because the accuracy is reduced if one moving object is relied on for the shadow, the noise on the image, or the change of the object area due to some factor.
- the color stable region estimation unit 2 Based on the information extracted by the position-specific color change feature extraction unit 1, the color stable region estimation unit 2 generates an image feature amount change model indicating a color feature change amount (a change amount of an image feature amount related to a color) for each position. Create and estimate the color stability region.
- the color stable region refers to a region where the color is stable in the photographing range, that is, a region where color change hardly occurs.
- the color stable region estimation unit 2 creates, for example, an image feature amount change model indicating a color feature change amount for each position, and sets a region having a large color feature change amount as a boundary region, and the colors divided by the boundary regions Each region having a small feature change amount may be estimated as one color stable region. In this case, the range of the region having a small color feature change amount divided by the boundary region is estimated as a color stable region under the same illumination condition.
- the area-specific color correction parameter calculation unit 3 calculates a color correction parameter for each color stable region estimated by the color stable region estimation unit 2. For example, when a reference color such as a color chart is photographed under one illumination condition, the color correction parameter calculation unit 3 for each region has the region (reference color) based on the photographing result of the reference color. A color correction parameter for the color stable region where the object is located) may be calculated. If the color correction parameter is calculated in one color stable region, based on the information indicating the relationship between the position and the color in the shooting range extracted by the position-specific color change feature extracting unit 1, Color correction parameters can also be calculated.
- the color correction parameter calculation unit 3 for each region uses another color stable region based on the color in one color stable region in the correction target image.
- the color correction parameter in each color stable region may be calculated so that the inner color tone approaches the reference color tone.
- the color correction unit 4 corrects the color of the target image based on the color correction parameter for each color stable region calculated by the region-specific color correction parameter calculation unit 3.
- the color correction unit 4 may correct the color in the target image for each color stable region using the color correction parameter calculated for each color stable region.
- the position-specific color change feature extraction unit 1, the color stable region estimation unit 2, the region-specific color correction parameter calculation unit 3, and the color correction unit 4 are realized by an information processing apparatus that operates according to a program such as a CPU, for example. Is done.
- the color change characteristics of the image at each position extracted by the position-specific color change feature extraction unit 1, information about the color stability region estimated by the color stability region estimation unit 2, and color correction for each region You may provide the memory
- FIG. 2 is a flowchart showing an operation example of this embodiment.
- the color stable region estimation unit 2 estimates the color stable region in the shooting range of the correction target image or the input moving image based on the extraction result by the position-specific color change feature extraction unit 1 (step S2).
- the color stable region estimation unit 2 is configured for each position of the shooting range of the correction target image based on the information indicating the relationship between the position in the shooting range extracted by the position-specific color change feature extraction unit 1 and the color change.
- An image feature amount change model indicating a color feature change amount is created, and a color stable region is estimated based on the magnitude of the color feature change amount for each position indicated by the created image feature amount change model.
- the shooting range of the correction target image and the shooting range of the input moving image are the same or the correspondence relationship is known, if the color stable region is obtained in one of the shooting ranges, the other shooting is performed. The color stability region in the range is inevitably obtained.
- the color correction parameter calculation unit 3 for each region calculates a color correction parameter for each color stable region (step S3).
- the area-specific color correction parameter calculation unit 3 includes, for example, information on the color stability area estimated by the color stability area estimation unit 2 and the relationship between the position and color in the shooting range extracted by the position-specific color change feature extraction unit 1. And a color correction parameter for each color stable region.
- the color correction unit 4 calculates the correction target image based on the color correction parameter calculated by the color correction parameter calculation unit 3 for each region.
- the color is corrected (step S4).
- the boundary region can be corrected based on position-specific color information.
- a color correction parameter is prepared for each color stable region, and the color correction parameter is used to perform color correction. Color correction may be performed for each stable region.
- the color change for each position of the moving object is grasped, the color stable region is estimated based on the state of the color change for each position, and for each position. Since the color correction parameter is calculated for each color stable region based on the color, when light enters from a plurality of types of light sources into the shooting range of a fixed camera such as a surveillance camera, the influence of the light in one image Even when the color changes for each region, the color changing for each region can be easily corrected.
- the association of people between cameras was performed in a part of the area, but in order to obtain the color stable area extraction and the color change for each color stable area, No tracking of a moving object and associating the color of the object with position information is performed.
- a moving object within the shooting range of the camera is tracked using an image feature amount or the like, and information that associates the color of the object with position information is acquired.
- the presence / absence (or degree) of color change between regions is estimated.
- a color stable region is estimated from the presence or absence of a color change between the estimated regions.
- a color correction parameter is calculated from the state of color change of the moving object for each estimated color stable region.
- the color change characteristics associated with the movement of the moving object are used to estimate the color stable area and calculate the color correction parameter for each color stable area.
- Colors that change from region to region can be easily corrected. That is, even if color charts are not arranged at multiple locations, the colors in the image are corrected so that the color of the same object does not change for each position of the shooting area, or the colors in the image are close to the original colors
- the color in the image can be corrected so as to be tint.
- the position-specific color change feature extraction unit 1 and the color stable region estimation unit 2 of the present embodiment extract the color change feature of the image for each position and time, and the color for each position and time based on the extraction result.
- an image feature amount change model indicating the amount of feature change may be created, and the color stable region for each time may be estimated.
- FIG. FIG. 3 is a block diagram illustrating a configuration example of the color correction apparatus according to the second embodiment of the present invention.
- the color correction apparatus according to the present embodiment is an example of a more specific configuration of the color correction apparatus according to the first embodiment illustrated in FIG. 3 includes a moving object region extraction unit 21, a moving object identification unit 22, a position-specific image feature amount holding unit 23, a boundary information calculation unit 24, and an inter-region color conversion parameter calculation unit 25.
- the specific position color correction parameter calculation unit 26, the area-specific color correction parameter calculation unit 27, and the color correction unit 28 are provided.
- a moving image captured by a fixed camera is used as an input image.
- the correction target image is an image taken with the same fixed camera. It may be one or a plurality of frames included in the input image. Note that the input image does not have to be a moving image shot by a fixed camera as long as the shooting range is the same as the shooting range of the correction target image or includes the entire shooting range of the correction target image.
- the input image may be a plurality of types of moving images captured by a plurality of fixed cameras as long as the relationship between the capturing range and the actual position is known.
- the moving object area extraction unit 21 extracts the moving object from the input moving image. More specifically, the moving object area extraction unit 21 extracts an area where the moving object is located (hereinafter referred to as an object area) from the input image.
- object area extraction may be performed by finding pixel changes between images. For example, a background difference or an inter-frame difference may be calculated, and a pixel area having the difference may be detected as an object area.
- the moving object region extraction unit 21 may perform an object detection process using a classifier (identification model) generated by learning or template matching when extracting an object region. For example, in order to determine whether or not what is shown in the extracted area as a difference is a moving object, a discriminator that learned the feature amount of a general moving object image was generated and generated The object determination process may be performed by inputting the image information of the extracted region to the classifier.
- a support vector machine (SVM), AdaBoost, or the like can be used as the discriminator.
- the moving object region extraction unit 21 performs an image feature amount expressing color from the extracted object region and an image feature amount for tracking for the identification process and the image feature amount calculation process for each position performed in the subsequent stage. To extract.
- the image feature amount expressing a color include a color histogram in a color space such as RGB, HSV, L * a * b, and L * u * v.
- the image feature amount for tracking depends on the algorithm used, but in addition to the image feature amount expressing the color described above, the edge feature amount and the feature amount indicating a local region (Haar-like, HOG, SIFT) , SURF, etc.). Note that the image feature amount expressing color and the image feature amount for tracking are not limited to the above-described examples.
- the moving object region extraction unit 21 may extract only the image feature amount at a specific position of the object when performing object identification or the like. For example, when person identification is performed, only the image feature amount of the upper body may be extracted.
- the moving object identification unit 22 tracks the moving object in the input moving image based on the image feature amount of the object region extracted by the moving object region extraction unit 21. More specifically, the moving object identification unit 22 performs identification processing of the object region extracted by the moving object region extraction unit 21 between each frame in the input moving image. For example, the moving object identification unit 22 may determine the identity of a moving object whose position changes by tracking, and may assign an ID indicating the same if the moving object is extracted from each frame. . The moving object identification unit 22 may also record a change in the angle of view and a change in magnification due to the PTZ when a person is tracked with a pan / tilde / zoom (PTZ) camera.
- PTZ pan / tilde / zoom
- Object tracking methods include, for example, the movement of an object, such as the color or edge of an object, with respect to an object extracted by an object discriminator such as pattern matching, SVM, or AdaBoost. Tracking technology that estimates by image feature can be used. Specifically, mean-shift tracking, particle filter, KLT method, Kalman filter, or the like can be used.
- the position of the moving object in each frame is specified by such identification processing.
- position information such as the moving path of the moving object in the shooting region and the position of each movement is specified.
- the conversion from the position in the frame image to the position in the shooting area may be performed using a conversion expression registered in advance or a conversion expression calculated from camera parameters.
- the position-specific image feature value holding unit 23 holds the object position and the image feature value indicating the color in association with each other.
- the object position for example, the lower end portion of the object area extracted by the moving object area extraction unit 21 may be set. Further, for example, when it is determined that the person is a person by a discriminator, the foot position of the person may be set.
- the positional relationship of the changing shooting range is registered together with the image feature amount indicating the object position and color. Note that it is also possible to set the position information in the three-dimensional region, or simply set the position information (pixel position, etc.) on the image, assuming that the real-world shooting range is three-dimensional.
- the moving object identification unit 22 may extract the same position based on the information and register it in the position-specific image feature amount holding unit 23. Good. It is also possible to extract the same position of the object using the image feature amount. In this case, the same position may be determined using SURF (Speeded Up Up Robust Features), SIFT (Scale Invariant Feature Up Transformation) features, template matching, or the like.
- SURF Speeded Up Up Robust Features
- SIFT Scale Invariant Feature Up Transformation
- the boundary information calculation unit 24 changes the color in the shooting region (real space) based on the information that associates the object position held in the position-specific image feature quantity holding unit 23 with the image feature quantity indicating the color.
- a position (boundary region) where the occurrence of the color is calculated to estimate a color stable region.
- the boundary information calculation unit 24 converts the imaging region into a plurality of partial regions based on information that associates the object position held in the position-specific image feature amount holding unit 23 with the image feature amount indicating the color. For each divided partial region, an image feature amount change model indicating a change amount (representative value) of the image feature amount between adjacent regions in the partial region may be generated.
- each partial region includes a region having a large color change with respect to the adjacent region, or the boundary between each partial region and the adjacent region has a large color change
- the real space shooting area to which the area belongs belongs to a boundary area where the color is not stable May be determined. You may estimate that the range of the area
- a region group with a small amount of change in the image feature amount in which a predetermined size or more is secured as a continuous region without being divided by the boundary region may be extracted.
- the change amount between which adjacent region is included in the total result of the change amount between the adjacent regions is extracted. It is possible to extract a position in the photographing area of the real space corresponding to the boundary with the adjacent area indicated by the direction as a position where the color change occurs. Even if the extracted position where the color change occurs is defined as a boundary position, the range of the area where the color change does not occur divided by the boundary position is estimated as a color stable area under the same illumination condition. Good. When the boundary position is not continuous, interpolation processing or the like may be performed so that the boundary position is continuous.
- FIG. 4 is an explanatory diagram showing an example of the boundary area and the color stability area.
- irradiation with two different illuminations 303 and 304 is performed within the imaging range 300 of the fixed camera 301, and the illumination range 305 of the illumination 303 and the illumination range 306 of the illumination 304 partially overlap.
- Yes That is, in the imaging range 300, there are an area included in the irradiation range 305, an area included in the irradiation range 306, and an area included in the irradiation range 305 and the irradiation range 306.
- FIG. 4 it is shown that three color stable regions divided into two boundary regions are estimated as a result of tracking the person 307 and the like under such a situation.
- a color stable region under illumination condition A corresponds to a region irradiated only by the illumination 304.
- the color stable region under the illumination condition B corresponds to a region irradiated with the illumination 303 and the illumination 304.
- the color stable region under the illumination condition C corresponds to a region irradiated only by the illumination 303.
- the boundary information calculation unit 24 uses the difference in the image feature amount of the color of the moving object extracted sequentially in time series based on the tracking result as the change amount of the image feature amount (for example, the distance between histograms, etc.). Is calculated as an image feature amount change amount. For example, the boundary information calculation unit 24 calculates the difference between the image feature amount of the color of the moving object at time t and the image feature amount of the color of the moving object at time t + 1. At this time, the position information and the moving direction information are simultaneously held.
- the boundary information calculation unit 24 aggregates the image feature amount change amount calculated for each position, and calculates an image feature amount change amount represented by each position.
- an area where the object can move for example, the bottom surface
- the image feature amount change amount may be calculated and aggregated between adjacent partial areas.
- a method for dividing the area is not particularly limited, but the area may be divided into rectangular partial areas.
- the boundary information calculation unit 24 may set a plurality of virtual measurement reference points instead of setting the partial areas. In such a case, the boundary information calculation unit 24 may associate the position information closest to each measurement reference point and calculate and aggregate the image feature amount change amount between adjacent measurement reference points. .
- the boundary information calculation unit 24 may store the calculated image feature amount change amount in association with each of the adjacent partial regions or measurement reference points used for the calculation together with the comparison destination information.
- the image feature amount change amount is calculated for each movement of a plurality of objects, and these are totaled and modeled to obtain a representative value of the image feature amount change amount at each position.
- the model for calculating the representative value of the image feature amount change amount may be, for example, a model for calculating the average or variance value of the image feature amount change amount for each aggregated position. At this time, since the total number varies depending on each region / point, the amount of change may be normalized. In addition, in order to make the calculated representative value more closely match the actual change in the light source, a value that seems to be a shadow or noise is removed according to a predetermined rule, or interpolation processing is performed so that the boundary region is continuous. Also good.
- FIG. 5 to FIG. 8 are explanatory diagrams for explaining a totaling method of the image feature amount change amount for each position.
- FIG. 5 is an explanatory diagram illustrating an example of a camera viewpoint video.
- 401 indicates a frame range.
- Reference numeral 402 denotes a wall, and 403 denotes a floor surface.
- the camera viewpoint video in the frame range 401 shows that the person 404 has moved in the illustrated direction (tilt leftward as viewed from the person 404) from time t to time t + 1. Is shown.
- FIG. 6 is a top view showing the real-world positional relationship of the subject included in the frame range shown in FIG.
- the shooting range of the camera 405 is indicated by a broken line 406.
- the alternate long and short dash line 406 ′ indicates a floor area included in the imaging range.
- FIG. 7 is an explanatory diagram showing an example of dividing the floor area 406 ′ of the photographing range into rectangular partial areas.
- the boundary information calculation unit 24 divides the floor area 406 ′ of the imaging range into a plurality of partial areas 501, and image features between adjacent partial areas (see shaded display). The amount of change in the amount may be calculated sequentially, and then summed up.
- the partial area may not be rectangular.
- FIG. 8 is an explanatory diagram showing an example in which a plurality of measurement reference points are provided in the floor area 406 ′ of the imaging range.
- the boundary information calculation unit 24 provides a plurality of measurement reference points 601 in the floor surface area 406 ′ of the imaging range, and associates with the position information closest to each measurement reference point 601.
- the amount of change in the image feature amount between adjacent measurement reference points 601 may be calculated sequentially, and these may be aggregated. Further, when the amount of change is obtained for each arbitrarily determined measurement reference point, the amount of change between the measurement reference points may be calculated by interpolation.
- the boundary information calculation unit 24 performs region classification based on the representative value of the image feature amount change amount for each position obtained in this way. For example, the boundary information calculation unit 24 performs threshold processing on the representative value of the image feature amount change amount for each obtained position, and sets a single region where the change amount is determined to be small as one color stable region. Also good. In addition, for example, threshold processing may be performed, and a boundary of a region with a large amount of change may be defined as a boundary region, and a region of a certain size or more divided by the boundary region may be defined as a color stable region.
- FIG. 9 is an explanatory diagram showing an example of the total result of the image feature amount change amount.
- FIG. 9 as a result of classifying regions based on the representative value of the image feature amount change amount for each partial region, adjacent regions where the image feature amount change amount is equal to or less than a predetermined value are shown by the same shading. Yes.
- a boundary with an adjacent region where the image feature amount change amount is equal to or greater than a predetermined value is indicated by a solid line as a boundary position. That is, in FIG. 9, the area surrounded by the solid line corresponds to one color stable area.
- a total of 10 color stable regions are detected as a result of each region not shaded due to the range of the photographing region being regarded as one color stable region.
- FIG. 10 is an explanatory diagram showing another example of the total result of the image feature amount change amount.
- FIG. 10 is an explanatory diagram illustrating an example of a result of aggregating the change amount of the image feature amount obtained for each measurement reference point arranged at equal intervals.
- the image feature amount change amount is equal to or greater than a predetermined value.
- a boundary with an adjacent area is indicated by a solid line as a boundary position, and an area divided by the boundary position is indicated as one tint stable area. That is, in the example shown in FIG. 10, the region surrounded by the solid line corresponds to one color stable region.
- a total of six color stable regions are detected as a result of each region not shaded due to the range of the photographing region being regarded as one color stable region.
- the inter-region color conversion parameter calculation unit 25 calculates a color conversion parameter that is a parameter for converting a color between color stable regions.
- the color conversion parameter is a parameter for converting the color in the image.
- the color conversion parameter absorbs the difference in the illumination condition between the target areas, and the color of the same object in the image is changed between the areas. It is a parameter for expressing with the same color.
- the color conversion parameter can be said to be a kind of color correction parameter in that the color of one region is corrected based on the color of one region.
- the inter-region color conversion parameter calculation unit 25 calculates a color conversion parameter between the color stable regions based on the position-specific image feature amount held in the position-specific image feature amount holding unit 23.
- the inter-region color conversion parameter calculation unit 25 may perform calculation so that the error is minimized within the color stable region.
- a least square method, robust estimation, RANSAC, or the like can be used for the calculation of the color conversion parameter.
- the inter-region color conversion parameter calculating unit 25 uses the two color stable regions 602 and the color stable region 604.
- a conversion formula for converting the image feature amount in the region 604 into the image feature amount in the color stable region 602 may be calculated.
- FIG. 11 is an explanatory diagram for explaining a method of calculating a color conversion parameter between regions.
- the color correction unit 28 uses the color conversion parameter for each region calculated by the inter-region color conversion parameter calculation unit 25 or the color correction parameter for each region calculated by the region-specific color correction parameter calculation unit 27 to be described later.
- the colors in the image are corrected for each region so that the color of the same object does not change between them, or the color in the color stable region becomes a color close to the reference color.
- the color correction unit 28 performs inter-region color conversion parameter calculation unit for each color stable region along the movement path of the object. What is necessary is just to convert the color of an object using the color conversion parameter between each color stable area
- FIG. 12 is an explanatory diagram for explaining a method of calculating a color conversion parameter between color stable regions.
- the color correction unit 28 calculates a color correction parameter for correcting the color under the reference light source by, for example, a specific position color correction parameter calculation unit 26 and a region-specific color correction parameter calculation unit 27 described later. If so, color conversion may be performed using the calculated area-specific color correction parameters. Thereby, the color of each color stable area in the image can be converted to the color under the reference light source.
- the specific position color correction parameter calculation unit 26 determines a color (hereinafter referred to as a reference color) in which a color development state (for example, image feature amount) is grasped under a reference illumination condition (for example, under sunlight).
- a color development state for example, image feature amount
- a reference illumination condition for example, under sunlight.
- the specific position color correction parameter calculating unit 26 grasps the color development state of the object B in the color stable region A in the image. Therefore, the color (image feature amount) of the object B in the color stable region A is calculated from the input image. With respect to the color of the object B, the color in the color stable region A is obtained by calculating the amount of change between the image feature amount under the reference illumination condition and the calculated image feature amount in the color stable region A. Correction parameters can be calculated.
- the object B here may be a color chart.
- the color correction parameter calculation unit 27 for each region is based on the inter-region color conversion parameter calculated by the inter-region color conversion parameter calculation unit 25 and the color correction parameter at the specific position calculated by the specific position color correction parameter calculation unit 26.
- a color correction parameter for converting the color development state in the color stability region into a color development state under a reference illumination condition is calculated. For example, as shown by an arrow in FIG. 13, when the color development state of the reference color is grasped in the color stable region 603, the color correction parameter between the color stable region 603 and the adjacent color stable region 604 is set.
- the color correction parameter of the color stable region 603 calculated by grasping the color development state of the reference color and the color conversion parameter between the regions, that is, the color conversion parameter between the color stable regions 603 to 604 may be calculated.
- the color correction parameter between the color stable region 604 for which the color correction parameter is calculated and the adjacent color stable region 602 is the same as the calculated color correction parameter of the color stable region 604 and the color conversion parameter between the regions, that is, It may be calculated based on the color conversion parameters between 604 and 602.
- the color correction parameter between the color stable region 604 for which the color correction parameter is calculated and the adjacent color stable region 605 is the same as the calculated color correction parameter of the color stable region 604 and the color conversion parameter between the regions.
- the color correction parameter between the color stable region 605 for which the color correction parameter is calculated and the adjacent color stable region 606 is the same as the calculated color correction parameter of the color stable region 605 and the color conversion parameter between the regions. It may be calculated based on the color conversion parameter between 605 and 606.
- the color correction parameter between the color stable region 605 for which the color correction parameter is calculated and the adjacent color stable region 607 is the same as the calculated color correction parameter for the color stable region 605 and the color conversion parameter between the regions. It may be calculated based on a color conversion parameter between 605 and 607.
- the color correction parameter of each color stable region can be obtained.
- the color correction parameter calculation method for each color stable region may be the same as the color conversion parameter calculation method.
- the moving object region extraction unit 21, the moving object identification unit 22, the boundary information calculation unit 24, the inter-region color conversion parameter calculation unit 25, the specific position color correction parameter calculation unit 26, and the region-specific color correction parameter calculation unit 27 The color correction unit 28 is realized by, for example, an information processing apparatus that operates according to a program such as a CPU. Further, the position-specific image feature value holding unit 23 is realized by a storage device.
- the color correction apparatus includes a display control unit (not shown) that displays the color stable area on the display device in such a manner that the boundary of each color stable area can be determined. Also good.
- the display control unit displays on the display device as shown in FIG. 10, it accepts a user input such as a click to the color stability region 605 and the color stability region 604 (see FIG. 12), and the color stability with the user input.
- Color correction processing may be executed between regions.
- FIG. 14 is a flowchart illustrating an operation example of the present embodiment.
- an imaging region is set in accordance with camera settings (step S ⁇ b> 21).
- camera parameters may be input.
- step S22 an initialization process for detecting a background difference is performed (step S22).
- a moving image in which only the background is photographed is input together with the fact so that the background difference can be detected in the tracking process.
- step S23 the moving object region extraction unit 21 extracts an object region from the input moving image.
- the extraction of the object region may be performed for each frame or may be performed at regular time intervals in the moving image.
- the moving object region extraction unit 21 extracts an image feature amount for displaying a color from the extracted object region and an edge feature amount as necessary (step S24).
- the extracted image feature amount is held in the position-specific image feature amount holding unit 23 in step S26 together with information for specifying the position of the object region in the shooting frame.
- the moving object identification unit 22 tracks the moving object in the input moving image based on the image feature amount of the object region extracted in step S24 (step S25).
- the moving object identification unit 22 assigns an ID indicating that the moving object extracted from each frame is the same to the moving object whose position changes by the identification process.
- a position-specific image is created by associating an ID (moving body identifier), information specifying the position of the object region in the shooting frame, an image feature amount at the position, and information indicating the shooting time or time in the moving image. It holds in the feature amount holding unit 23 (step S26). Note that information specifying the position of the moving object in the imaging region may be held instead of the information specifying the position of the object region in the imaging frame, or together with the information specifying the position of the object region in the imaging frame. .
- step S23 to step S26 The processing from step S23 to step S26 is performed for a plurality of moving objects.
- the boundary information calculation unit 24 extracts a color stable region (step S27). For example, the boundary information calculation unit 24 causes a color change in the imaging region based on information that associates the object position and the image feature amount indicating the color that are held in the position-specific image feature amount holding unit 23.
- the boundary region may be calculated to estimate the color stable region.
- the inter-region color conversion parameter calculation unit 25 performs color conversion between the color stable regions based on the position-specific image feature amount held in the position-specific image feature amount holding unit 23.
- a parameter is calculated (step S28).
- the color correction unit 28 corrects the colors in the image for each region based on the calculated color conversion parameters between the color stable regions (step S29).
- the color correction unit 28 and the color chart image input to the specific position color correction parameter calculation unit 26 and the area-specific color correction parameter calculation unit 27 Using the calculated color conversion parameters between the color stable areas, calculate the color correction parameters for each color stable area, and correct the colors in the image for each area based on the calculated color correction parameters for each color stable area May be.
- the color chart image may be, for example, an image taken in the initial shooting for taking a background difference.
- an image obtained by photographing a color chart including a reference color together with the background may be first input as an initial image.
- the color change for each position of the moving object is grasped, the color stable region is estimated based on the change state for each position, and the color for each position is determined. Based on the calculated color conversion parameters between the color stability areas, color correction is performed for each area based on the calculated color conversion parameters between the color stability areas. Even when light is incident from the light source, it is possible to easily correct the color changing in one image due to the influence of the light.
- the moving object region extraction unit 21 extracts the color change feature of the image for each position and time, and the boundary information calculation unit 24 determines for each position based on the extraction result by the moving object region extraction unit 21. By estimating the color stable region for each time, it is possible to cope with the change of the light source due to the time change.
- the color correction apparatus includes a moving object detection unit 101, a position-specific color information acquisition unit 102, a position-specific color information storage unit 103, a color stable region estimation unit 104, and a color conversion.
- Parameter calculation means 105 and color correction means 106 are provided.
- the moving object detection means 101 (for example, the position-specific color change feature extraction unit 1 or the moving object region extraction unit 21) has a shooting range that is the same as the shooting range of the correction target image or a moving image that includes the entire shooting range of the correction target image. A moving object is detected from the image.
- the position-specific color information acquisition unit 102 determines the identity of the moving object detected by the moving object detection unit 101 in the moving image, and moves Acquired is the position of the moving object and information indicating the color of the moving object in the image.
- the position-specific color information storage unit 103 (for example, the position-specific image feature amount holding unit 23) is a position of the moving object that accompanies the movement of the moving object acquired by the position-specific color information acquisition unit 102 and the movement in the image.
- the information indicating the color of the object is stored in association with it.
- the color stable region estimation unit 104 (for example, the color stable region estimation unit 2 or the boundary information calculation unit 24) is configured to detect the position of the moving object and the inside of the image as the moving object stored in the position-specific color information storage unit 103 moves. Based on the information associated with the information indicating the color of the moving object at, a color stable region that is a region where the color is stable in the photographing range is estimated.
- the color conversion parameter calculation unit 105 (for example, the region-specific color correction parameter calculation unit 3 or the inter-region color conversion parameter calculation unit 25) is stored in the estimation result by the color stable region estimation unit 104 and the position-specific color information storage unit 103. Based on the information indicating the relationship between the position of the moving object and the color, a color conversion parameter that is a parameter for converting the color between the color stable regions is calculated.
- the color conversion parameter calculation unit 105 refers to information indicating the color of the moving object in each color stable region from the position-specific color information storage unit 103 based on the estimation result by the color stable region estimation unit 104, for example, A color conversion parameter, which is a parameter for converting colors between them, is calculated.
- the color correction unit 106 (for example, the color correction unit 4 or the color correction unit 28) corrects the color of the correction target image using the color conversion parameter between the color stable regions calculated by the color conversion parameter calculation unit 105.
- the color correction apparatus shown in FIG. 16 includes a specific area color correction parameter calculation unit 107 and a color correction parameter calculation unit 108 in addition to the configuration shown in FIG.
- the specific area color correction parameter calculation means 107 (for example, the specific position color correction parameter calculation unit 26) is an image whose shooting range is the same as or partially overlapped with the shooting range of the moving image used for detecting the moving object. From an image in which an object having a reference color is arranged in the color stable region, for at least one color stable region, the color development state in the region is brought close to the color development state under a reference illumination condition. The color correction parameter that is the parameter of is calculated.
- the color correction parameter calculation unit 108 uses the color correction parameter for one color stable region calculated by the specific region color correction parameter calculation unit 107 and the color conversion parameter calculation unit 105. Based on the calculated color conversion parameter between the color stable regions, a color correction parameter is calculated for each color stable region.
- the color correcting unit 106 may correct the color of the correction target image for each color stable region using the color correction parameter for each color stable region calculated by the color correction parameter calculating unit 108. .
- the color stable region estimation unit 104 indicates a color between adjacent positions for each position based on the information indicating the relationship between the position of the moving object and the color stored in the position-specific color information storage unit 103.
- the amount of change in information may be aggregated, and the color stable region may be estimated by extracting a position where a color change occurs in the imaging region based on the aggregation result.
- the color stable region estimation means 104 is an image feature that indicates the amount of color feature change for each position based on the information indicating the relationship between the position and color of the moving object stored in the position-specific color information storage means 103.
- a quantity change model may be created, and the color stable region may be estimated based on the color feature change quantity for each position indicated by the created image feature quantity change model.
- the position-specific color information acquisition unit 102 may acquire information indicating the color of the specific position of the moving object as information indicating the color of the moving object.
- the position-specific color information acquisition unit 102 acquires the position of the moving object accompanying the movement, information indicating the color of the moving object in the image, and the time at that time, and the color stable region estimation unit 104 Based on information indicating the relationship between the position and color of the moving object acquired by the different color information acquisition unit 102 and stored in the position-specific color information storage unit 103, the time-dependent color stable region may be estimated. Further, the estimation of the color stable region for each time may be automatically performed at an appropriate time as the actual time elapses. For example, the color correction device may automatically perform color correction processing by estimating the color stable region every hour until noon, 15:00, 17:00, and after that until midnight. With this configuration, the color correction apparatus of the present invention can automatically correct color changes on an image due to outdoor weather and time zones.
- the position-specific color information acquisition unit 102 may acquire position information of the lower end portion of the object area as information indicating the position of the moving object.
- a moving object detecting unit that detects a moving object from a moving image in which the shooting range is the same as or including the entire shooting range of the correction target image, and is detected by the moving object detection unit.
- Position-specific color information acquisition means for determining the identity of the moving object in the moving image and acquiring the position of the moving object accompanying the movement and information indicating the color of the moving object in the image;
- Position-specific color information storage means for storing the position of the moving object associated with the movement of the moving object acquired by the different color information acquisition means and information indicating the color of the moving object in the image in association with each other, and the position Based on information indicating the relationship between the position of the moving object and the color stored in the separate color information storage unit, a color stable region estimation unit that estimates a color stable region that is a region where the color is stable in the photographing range; color Parameters for converting colors between color stable regions based on estimation results by the constant region estimation means and information indicating the relationship between the position and color of the moving object stored in the
- the said color stable area estimation means is a color between the adjacent positions for every position based on the information which shows the relationship between the position and color of the moving object memorize
- the color correction apparatus according to Supplementary Note 1 or Supplementary Note 2, wherein the color stable region is estimated by summing the amount of change in information indicating the color and extracting a position where a color change occurs in the imaging region based on the summary result.
- the said color stable area estimation means shows the color feature variation
- the color correction apparatus according to appendix 1 or appendix 2, wherein an image feature amount change model is created and a color stable region is estimated based on a color feature change amount for each position indicated by the created image feature amount change model.
- the position-specific color information acquisition unit acquires the position of the moving object accompanying the movement, the information indicating the color of the moving object in the image, and the time at that time, and the color stable region estimation unit 6.
- the information according to any one of appendix 1 to appendix 5, in which a time-dependent color stable region is estimated based on information indicating a relationship between a position, a color, and time of a moving object acquired by the position-specific color information acquisition unit.
- Color correction device is any one of appendix 1 to appendix 5, in which a time-dependent color stable region is estimated based on information indicating a relationship between a position, a color, and time of a moving object acquired by the position-specific color information acquisition unit.
- a moving object is detected from a moving image whose shooting range is the same as that of the correction target image or includes the entire shooting range of the correction target image, and the identity of the detected moving object in the moving image is determined. Determining, acquiring the position of the moving object accompanying the movement and the information indicating the color of the moving object in the image, storing them in association with each other, storing them in the storage unit, and storing the position of the moving object stored in the storage unit Based on the information indicating the relationship between the color and the color, the color stable region which is a region where the color is stable in the photographing range is estimated, the estimation result of the color stable region, and the position of the moving object stored in the storage unit And a color conversion parameter that is a parameter for converting the color between the color stable regions based on the information indicating the relationship between the color and the color, and using the calculated color conversion parameter between the color stable regions Of the image to be corrected Color correction method characterized by correcting the.
- a color correction parameter which is a parameter for approximating the color development state under illumination conditions based on the color development state in the region, is calculated, and the calculated at least one color
- the color correction parameter is calculated for each color stable region, and correction is performed using the calculated color correction parameter for each color stable region.
- the computer acquires the position of the moving object accompanying the movement, the information indicating the color of the moving object in the image and the time at that time, and the position and color of the acquired moving object.
- the color correction program according to any one of supplementary note 15 to supplementary note 19, which estimates a time-dependent color stable region based on information indicating a relationship with time.
- a moving object detecting unit that detects a moving object from a moving image having a shooting range that is the same as or including the entire shooting range of the correction target image, and that is detected by the moving object detection unit.
- a position-specific color information acquisition unit that determines the identity of the moving object in the moving image and acquires information indicating the position of the moving object and the color of the moving object in the image,
- Color correction comprising color stable region estimation means for estimating a plurality of color stable regions, which are regions where the color is stable in the photographing range, based on information indicating the relationship between the position and color of the moving object apparatus.
- the color correction device according to claim 22, further comprising display control means for displaying the color stable region on the display device in a manner in which the estimated boundary for each color stable region can be determined.
- a moving object is detected from a moving image whose shooting range is the same as that of the correction target image or includes the entire shooting range of the correction target image, and the identity of the detected moving object in the moving image. Based on the information indicating the relationship between the position of the moving object and the color of the moving object in the image, and the information indicating the relationship between the position and color of the acquired moving object.
- Processing for detecting a moving object from a moving image in which the shooting range is the same as that of the correction target image or includes the entire shooting range of the correction target image in the computer, and the moving image of the detected moving object A process for acquiring the position of the moving object accompanying the movement and information indicating the color of the moving object in the image, and the relationship between the acquired position and color of the moving object.
- a color correction program for executing a process of estimating a plurality of color stable regions, which are regions where colors are stable in an imaging range, based on the information shown.
- the color correction program according to supplementary note 26 which causes a computer to execute a process of displaying a color stable region on a display device in a manner in which a boundary for each estimated color stable region can be determined.
- the present invention can be suitably applied to camera video analysis applications, such as when it is desired to determine the color of clothes, which is one of the characteristics of a person, from a video shot by a surveillance camera.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Color Television Image Signal Generators (AREA)
- Processing Of Color Television Signals (AREA)
Abstract
Description
以下、本発明の実施形態を図面を参照して説明する。図1は、本発明の第1の実施形態の色補正装置の構成例を示すブロック図である。図1に示す色補正装置は、位置別色変化特徴抽出部1と、色安定領域推定部2と、領域別色補正パラメタ算出部3と、色補正部4とを備える。
図3は、本発明の第2の実施形態の色補正装置の構成例を示すブロック図である。本実施形態の色補正装置は、図1で示した第1の実施形態の色補正装置のより具体的な構成の一例である。図3に示す色補正装置は、移動物体領域抽出部21と、移動物体同定部22と、位置別画像特徴量保持部23と、境界情報算出部24と、領域間色変換パラメタ算出部25と、特定位置色補正パラメタ算出部26と、領域別色補正パラメタ算出部27と、色補正部28とを備える。
2 色安定領域推定部
3 領域別色補正パラメタ算出部
4 色補正部
21 移動物体領域抽出部
22 移動物体同定部
23 位置別画像特徴量保持部
24 境界情報算出部
25 領域間色変換パラメタ算出部
26 特定位置色補正パラメタ算出部
27 領域別色補正パラメタ算出部
28 色補正部
101 移動物体検出手段
102 位置別色情報取得手段
103 位置別色情報記憶手段
104 色安定領域推定手段
105 色変換パラメタ算出手段
106 色補正手段
107 特定領域色補正パラメタ算出手段
108 色補正パラメタ算出手段
Claims (14)
- 撮影範囲が補正対象画像の撮影範囲と同じまたは補正対象画像の撮影範囲を全て含んでいる動画像から、移動物体を検出する移動物体検出手段と、
前記移動物体検出手段により検出された移動物体の前記動画像内における同一性を判定し、移動に伴う当該移動物体の位置と画像内での当該移動物体の色を示す情報とを取得する位置別色情報取得手段と、
前記位置別色情報取得手段によって取得された移動物体の移動に伴う当該移動物体の位置と画像内での当該移動物体の色を示す情報とを対応づけて記憶する位置別色情報記憶手段と、
前記位置別色情報記憶手段に記憶されている移動物体の位置と色との関係を示す情報に基づいて、撮影範囲において色が安定する領域である色安定領域を推定する色安定領域推定手段と、
前記色安定領域推定手段による推定結果と、前記位置別色情報記憶手段に記憶されている移動物体の位置と色との関係を示す情報とに基づいて、色安定領域間で色を変換するためのパラメタである色変換パラメタを算出する色変換パラメタ算出手段と、
前記色変換パラメタ算出手段によって算出された色安定領域間の色変換パラメタを用いて、補正対象画像の色を補正する色補正手段とを備えた
ことを特徴とする色補正装置。 - 撮影範囲が移動物体の検出に用いた動画像の撮影範囲と同じまたは一部重複している画像であって、レファレンス色を有する物体が色安定領域に配置されている画像から、少なくとも1つの色安定領域に対して、当該領域での発色状態を基準とされる照明条件下での発色状態に近づけるためのパラメタである色補正パラメタを算出する特定領域色補正パラメタ算出手段と、
前記特定領域色補正パラメタ算出手段によって算出された1つの色安定領域に対する色補正パラメタと、色変換パラメタ算出手段によって算出された色安定領域間の色変換パラメタとに基づいて、色安定領域ごとに色補正パラメタを算出する色補正パラメタ算出手段とを備え、
色補正手段は、前記色補正パラメタ算出手段によって算出された色安定領域ごとの色補正パラメタを用いて、補正対象画像の色を補正する
請求項1に記載の色補正装置。 - 色安定領域推定手段は、位置別色情報記憶手段に記憶されている移動物体の位置と色との関係を示す情報に基づいて、位置ごとの隣接位置との間の色を示す情報の変化量を集計し、集計結果に基づいて撮影領域内で色変化が起きる位置を抽出することにより、色安定領域を推定する
請求項1または請求項2に記載の色補正装置。 - 色安定領域推定手段は、位置別色情報記憶手段に記憶されている移動物体の位置と色との関係を示す情報に基づいて、位置ごとの色の特徴変化量を示す画像特徴量変化モデルを作成し、作成された画像特徴量変化モデルによって示される位置ごとの色の特徴変化量に基づいて、色安定領域を推定する
請求項1または請求項2に記載の色補正装置。 - 位置別色情報取得手段は、移動物体の色を示す情報として、当該移動物体の特定位置の色を示す情報を取得する
請求項1から請求項4のうちのいずれか1項に記載の色補正装置。 - 位置別色情報取得手段は、移動に伴う当該移動物体の位置と画像内での当該移動物体の色を示す情報とその時の時刻とを取得し、
色安定領域推定手段は、位置別色情報取得手段によって取得された移動物体についての位置と色と時刻との関係を示す情報に基づいて、時間別の色安定領域を推定する
請求項1から請求項5のうちのいずれか1項に記載の色補正装置。 - 位置別色情報取得手段は、移動物体の位置を示す情報として、物体領域の下端部分の位置情報を取得する
請求項1から請求項6のうちのいずれか1項に記載の色補正装置。 - 撮影範囲が補正対象画像の撮影範囲と同じまたは補正対象画像の撮影範囲を全て含んでいる動画像から移動物体を検出し、
検出された移動物体の前記動画像内における同一性を判定し、移動に伴う当該移動物体の位置と画像内での当該移動物体の色を示す情報とを取得し、対応づけて記憶手段に記憶させ、
前記記憶手段に記憶されている移動物体の位置と色との関係を示す情報に基づいて、撮影範囲において色が安定する領域である色安定領域を推定し、
前記色安定領域の推定結果と、前記記憶手段に記憶されている移動物体の位置と色との関係を示す情報とに基づいて、前記色安定領域間で色を変換するためのパラメタである色変換パラメタを算出し、
算出された前記色安定領域間の色変換パラメタを用いて、補正対象画像の色を補正する
ことを特徴とする色補正方法。 - 撮影範囲が移動物体の検出に用いた動画像の撮影範囲と同じまたは一部重複している画像であって、レファレンス色を有する物体が色安定領域に配置されている画像から、少なくとも1つの色安定領域に対して、当該領域での発色状態を基準とされる照明条件下での発色状態に近づけるためのパラメタである色補正パラメタを算出し、
算出された前記少なくとも1つの色安定領域に対する色補正パラメタと、色安定領域間の色変換パラメタとに基づいて、色安定領域ごとに色補正パラメタを算出し、
算出された前記色安定領域ごとの色補正パラメタを用いて、補正対象画像の色を補正する
請求項8に記載の色補正方法。 - コンピュータに、
撮影範囲が補正対象画像の撮影範囲と同じまたは補正対象画像の撮影範囲を全て含んでいる動画像から、移動物体を検出する処理、
検出された移動物体の前記動画像内における同一性を判定し、移動に伴う当該移動物体の位置と画像内での当該移動物体の色を示す情報とを取得し、対応づけて記憶手段に記憶させる処理、
前記記憶手段に記憶されている移動物体の位置と色との関係を示す情報に基づいて、撮影範囲において色が安定する領域である色安定領域を推定する処理、
前記色安定領域の推定結果と、前記記憶手段に記憶されている移動物体の位置と色との関係を示す情報とに基づいて、前記色安定領域間で色を変換するためのパラメタである色変換パラメタを算出する処理、および
算出された前記色安定領域間の色変換パラメタを用いて、補正対象画像の色を補正する処理
を実行させるための色補正用プログラム。 - コンピュータに、
撮影範囲が移動物体の検出に用いた動画像の撮影範囲と同じまたは一部重複している画像であって、レファレンス色を有する物体が色安定領域に配置されている画像から、少なくとも1つの色安定領域に対して、当該領域での発色状態を基準とされる照明条件下での発色状態に近づけるためのパラメタである色補正パラメタを算出する処理、および
前記少なくとも1つの色安定領域における色補正パラメタと、前記色安定領域間の色変換パラメタとに基づいて、色安定領域ごとに色補正パラメタを算出する処理を実行させ、
色を補正する処理で、算出された前記色安定領域ごとの色補正パラメタを用いて、補正対象画像の色を補正させる
請求項10に記載の色補正用プログラム。 - 撮影範囲が補正対象画像の撮影範囲と同じまたは補正対象画像の撮影範囲を全て含んでいる動画像から、移動物体を検出する移動物体検出手段と、
前記移動物体検出手段により検出された移動物体の前記動画像内における同一性を判定し、移動に伴う当該移動物体の位置と画像内での当該移動物体の色を示す情報とを取得する位置別色情報取得手段と、
取得した前記移動物体の位置と色との関係を示す情報に基づいて、撮影範囲において色が安定する領域である色安定領域を複数推定する色安定領域推定手段とを備えた
ことを特徴とする色補正装置。 - 撮影範囲が補正対象画像の撮影範囲と同じまたは補正対象画像の撮影範囲を全て含んでいる動画像から、移動物体を検出し、
検出された移動物体の前記動画像内における同一性を判定し、移動に伴う当該移動物体の位置と画像内での当該移動物体の色を示す情報とを取得し、
取得した前記移動物体の位置と色との関係を示す情報に基づいて、撮影範囲において色が安定する領域である色安定領域を複数推定する
ことを特徴とする色補正方法。 - コンピュータに、
撮影範囲が補正対象画像の撮影範囲と同じまたは補正対象画像の撮影範囲を全て含んでいる動画像から、移動物体を検出する処理、
検出された移動物体の前記動画像内における同一性を判定し、移動に伴う当該移動物体の位置と画像内での当該移動物体の色を示す情報とを取得する処理、および、
取得した前記移動物体の位置と色との関係を示す情報に基づいて、撮影範囲において色が安定する領域である色安定領域を複数推定する処理
を実行させるための色補正プログラム。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/646,536 US9462160B2 (en) | 2012-11-22 | 2013-11-19 | Color correction device, method, and program |
JP2014548454A JP6428266B2 (ja) | 2012-11-22 | 2013-11-19 | 色補正装置、色補正方法および色補正用プログラム |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2012-255981 | 2012-11-22 | ||
JP2012255981 | 2012-11-22 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2014080613A1 true WO2014080613A1 (ja) | 2014-05-30 |
Family
ID=50775808
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2013/006776 WO2014080613A1 (ja) | 2012-11-22 | 2013-11-19 | 色補正装置、色補正方法および色補正用プログラム |
Country Status (3)
Country | Link |
---|---|
US (1) | US9462160B2 (ja) |
JP (1) | JP6428266B2 (ja) |
WO (1) | WO2014080613A1 (ja) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2016099941A (ja) * | 2014-11-26 | 2016-05-30 | 日本放送協会 | オブジェクト位置推定システム、及びそのプログラム |
JP2017076288A (ja) * | 2015-10-15 | 2017-04-20 | キヤノン株式会社 | 情報処理装置、情報処理方法及びプログラム |
JP2017191544A (ja) * | 2016-04-15 | 2017-10-19 | ホーチキ株式会社 | 火災検知装置及び火災検知方法 |
JP2018182593A (ja) * | 2017-04-17 | 2018-11-15 | キヤノン株式会社 | 画像処理装置、画像処理方法 |
JP2019103128A (ja) * | 2017-12-01 | 2019-06-24 | キヤノン株式会社 | 画像処理装置、制御方法及びプログラム |
JP2020509499A (ja) * | 2017-07-14 | 2020-03-26 | テンセント・テクノロジー・(シェンジェン)・カンパニー・リミテッド | ターゲット追跡方法、装置、電子機器及び記憶媒体 |
JP2020107349A (ja) * | 2014-09-26 | 2020-07-09 | 日本電気株式会社 | 物体追跡システム、物体追跡方法、プログラム |
US11113538B2 (en) | 2014-09-26 | 2021-09-07 | Nec Corporation | Object tracking apparatus, object tracking system, object tracking method, display control device, object detection device, and computer-readable medium |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP6335498B2 (ja) * | 2013-03-19 | 2018-05-30 | キヤノン株式会社 | 画像処理装置及びその制御方法 |
JP2016134803A (ja) * | 2015-01-20 | 2016-07-25 | キヤノン株式会社 | 画像処理装置及び画像処理方法 |
JP2017034444A (ja) * | 2015-07-31 | 2017-02-09 | オリンパス株式会社 | 撮像装置および撮像方法 |
US9858686B2 (en) | 2016-02-10 | 2018-01-02 | Google Llc | Dynamic color determination for user interface components of a video player |
CN105976398A (zh) * | 2016-04-28 | 2016-09-28 | 天津大学 | 一种白天火灾视频检测方法 |
US10134154B2 (en) | 2016-12-30 | 2018-11-20 | Google Llc | Selective dynamic color management for user interface components of a media player |
EP3493539B1 (en) * | 2017-12-01 | 2021-03-24 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method, and storage medium |
WO2020027233A1 (ja) | 2018-07-31 | 2020-02-06 | ソニーセミコンダクタソリューションズ株式会社 | 撮像装置及び車両制御システム |
JP6725733B2 (ja) * | 2018-07-31 | 2020-07-22 | ソニーセミコンダクタソリューションズ株式会社 | 固体撮像装置および電子機器 |
DE102021205096A1 (de) * | 2021-05-19 | 2022-11-24 | Robert Bosch Gesellschaft mit beschränkter Haftung | System und Verfahren, insbesondere computerimplementiertes Verfahren zum Bestimmen einer Störung zum Angreifen und/oder Validieren eines Objektverfolgers |
CN113467678A (zh) * | 2021-06-10 | 2021-10-01 | 北京达佳互联信息技术有限公司 | 目标对象显示方法、装置、电子设备及计算机可读存储介质 |
JP2022191030A (ja) * | 2021-06-15 | 2022-12-27 | 株式会社リコー | 画像形成装置、画像形成システム、及び、画像形成方法 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009065290A (ja) * | 2007-09-04 | 2009-03-26 | Canon Inc | 画像処理装置及び方法、及び撮像装置 |
JP2009212787A (ja) * | 2008-03-04 | 2009-09-17 | Ricoh Co Ltd | 撮像装置及び撮像方法 |
JP2011181014A (ja) * | 2010-03-03 | 2011-09-15 | Secom Co Ltd | 移動物体追跡装置 |
WO2012099013A1 (ja) * | 2011-01-20 | 2012-07-26 | シャープ株式会社 | 画像補正装置、画像補正表示装置、画像補正方法、プログラム、及び、記録媒体 |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4742542B2 (ja) * | 2004-09-02 | 2011-08-10 | 株式会社ニコン | 撮像装置 |
JP4813456B2 (ja) * | 2007-12-21 | 2011-11-09 | 日本放送協会 | 映像編集装置及び映像編集プログラム |
-
2013
- 2013-11-19 WO PCT/JP2013/006776 patent/WO2014080613A1/ja active Application Filing
- 2013-11-19 JP JP2014548454A patent/JP6428266B2/ja active Active
- 2013-11-19 US US14/646,536 patent/US9462160B2/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009065290A (ja) * | 2007-09-04 | 2009-03-26 | Canon Inc | 画像処理装置及び方法、及び撮像装置 |
JP2009212787A (ja) * | 2008-03-04 | 2009-09-17 | Ricoh Co Ltd | 撮像装置及び撮像方法 |
JP2011181014A (ja) * | 2010-03-03 | 2011-09-15 | Secom Co Ltd | 移動物体追跡装置 |
WO2012099013A1 (ja) * | 2011-01-20 | 2012-07-26 | シャープ株式会社 | 画像補正装置、画像補正表示装置、画像補正方法、プログラム、及び、記録媒体 |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2020107349A (ja) * | 2014-09-26 | 2020-07-09 | 日本電気株式会社 | 物体追跡システム、物体追跡方法、プログラム |
US11676388B2 (en) | 2014-09-26 | 2023-06-13 | Nec Corporation | Object tracking apparatus, object tracking system, object tracking method, display control device, object detection device, and computer-readable medium |
JP7004017B2 (ja) | 2014-09-26 | 2022-01-21 | 日本電気株式会社 | 物体追跡システム、物体追跡方法、プログラム |
US11113538B2 (en) | 2014-09-26 | 2021-09-07 | Nec Corporation | Object tracking apparatus, object tracking system, object tracking method, display control device, object detection device, and computer-readable medium |
JP2016099941A (ja) * | 2014-11-26 | 2016-05-30 | 日本放送協会 | オブジェクト位置推定システム、及びそのプログラム |
JP2017076288A (ja) * | 2015-10-15 | 2017-04-20 | キヤノン株式会社 | 情報処理装置、情報処理方法及びプログラム |
CN106603968A (zh) * | 2015-10-15 | 2017-04-26 | 佳能株式会社 | 信息处理装置及信息处理方法 |
JP2017191544A (ja) * | 2016-04-15 | 2017-10-19 | ホーチキ株式会社 | 火災検知装置及び火災検知方法 |
US11037014B2 (en) | 2017-04-17 | 2021-06-15 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method, and non-transitory computer-readable storage medium |
JP7058471B2 (ja) | 2017-04-17 | 2022-04-22 | キヤノン株式会社 | 画像処理装置、画像処理方法 |
JP2018182593A (ja) * | 2017-04-17 | 2018-11-15 | キヤノン株式会社 | 画像処理装置、画像処理方法 |
JP2020509499A (ja) * | 2017-07-14 | 2020-03-26 | テンセント・テクノロジー・(シェンジェン)・カンパニー・リミテッド | ターゲット追跡方法、装置、電子機器及び記憶媒体 |
US11145069B2 (en) | 2017-07-14 | 2021-10-12 | Tencent Technology (Shenzhen) Company Limited | Target tracking method and apparatus, electronic device, and storage medium |
JP2019103128A (ja) * | 2017-12-01 | 2019-06-24 | キヤノン株式会社 | 画像処理装置、制御方法及びプログラム |
JP7191633B2 (ja) | 2017-12-01 | 2022-12-19 | キヤノン株式会社 | 画像処理装置、制御方法及びプログラム |
Also Published As
Publication number | Publication date |
---|---|
JPWO2014080613A1 (ja) | 2017-01-05 |
JP6428266B2 (ja) | 2018-11-28 |
US20150334267A1 (en) | 2015-11-19 |
US9462160B2 (en) | 2016-10-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6428266B2 (ja) | 色補正装置、色補正方法および色補正用プログラム | |
TWI750498B (zh) | 視訊流的處理方法和裝置 | |
CN109684924B (zh) | 人脸活体检测方法及设备 | |
JP6482195B2 (ja) | 画像認識装置、画像認識方法及びプログラム | |
JP6560480B2 (ja) | 画像処理システム、画像処理方法、及びプログラム | |
CN104202547B (zh) | 投影画面中提取目标物体的方法、投影互动方法及其系统 | |
JP5567853B2 (ja) | 画像認識装置および方法 | |
JP6438403B2 (ja) | 結合された深度キューに基づく平面視画像からの深度マップの生成 | |
KR102462818B1 (ko) | 모션벡터 및 특징벡터 기반 위조 얼굴 검출 방법 및 장치 | |
JP5873442B2 (ja) | 物体検出装置および物体検出方法 | |
EP2549759B1 (en) | Method and system for facilitating color balance synchronization between a plurality of video cameras as well as method and system for obtaining object tracking between two or more video cameras | |
WO2011121688A1 (ja) | 顔認識装置及び顔認識方法 | |
JP6036824B2 (ja) | 画角変動検知装置、画角変動検知方法および画角変動検知プログラム | |
JP2014178957A (ja) | 学習データ生成装置、学習データ作成システム、方法およびプログラム | |
CN111862296A (zh) | 三维重建方法及装置、系统、模型训练方法、存储介质 | |
JP7334432B2 (ja) | 物体追跡装置、監視システムおよび物体追跡方法 | |
JP6521626B2 (ja) | 被写体追跡装置、方法およびプログラム | |
JP2018120283A (ja) | 情報処理装置、情報処理方法及びプログラム | |
US20220366570A1 (en) | Object tracking device and object tracking method | |
Kim | Pedestrian detection and distance estimation using thermal camera in night time | |
CN112073640B (zh) | 全景信息采集位姿获取方法及装置、系统 | |
CN110111368B (zh) | 一种基于人体姿态识别的相似移动目标的检测跟踪方法 | |
KR20120091970A (ko) | 객체 이미지 정합 시스템, 정합 장치 및 정합 방법과 상기 방법을 기록한 기록 매체 | |
JP5539565B2 (ja) | 撮像装置及び被写体追跡方法 | |
WO2019229855A1 (ja) | 物体検出装置、物体検出システム、物体検出方法、およびプログラムを記録する記録媒体 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 13857358 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2014548454 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14646536 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 13857358 Country of ref document: EP Kind code of ref document: A1 |