WO2006090449A1 - 画像処理方法、画像処理装置、画像処理システム及びコンピュータプログラム - Google Patents
画像処理方法、画像処理装置、画像処理システム及びコンピュータプログラム Download PDFInfo
- Publication number
- WO2006090449A1 WO2006090449A1 PCT/JP2005/002928 JP2005002928W WO2006090449A1 WO 2006090449 A1 WO2006090449 A1 WO 2006090449A1 JP 2005002928 W JP2005002928 W JP 2005002928W WO 2006090449 A1 WO2006090449 A1 WO 2006090449A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image processing
- luminance
- detection target
- pixels
- pixel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
- G06V10/507—Summing image-intensity values; Histogram projection analysis
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/59—Context or environment of the image inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
- G06V40/165—Detection; Localisation; Normalisation using facial parts and geometric relationships
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
- G06T2207/30201—Face
Definitions
- Image processing method image processing apparatus, image processing system, and computer program
- the present invention relates to an image processing method for detecting a specific detection target from a two-dimensional image in which pixels are arranged in different first and second directions, an image processing apparatus to which the image processing method is applied, and the image processing More particularly, the present invention relates to an image processing method, an image processing apparatus, an image processing system, and a computer program for improving detection accuracy of a detection target.
- Patent Document 1 JP 2004-234367 A
- the auto gain function that adjusts the brightness of the entire image cannot cope with local changes in illuminance. For example, if there is a partial change in which sunlight hits only the left half of the face, a dark part that is not exposed to sunlight cannot be recognized as a face and There is a problem in that erroneous detection in which only a large portion is detected as a face outline occurs. The rays of light that strike the driver's face in a vehicle running in this way are constantly changing, so it is sufficient to detect each part of the face or face using various methods and do not make a comprehensive judgment. There is a problem that the accuracy is not obtained.
- the present invention has been made in view of such circumstances, and in the case where a detection target such as a human nose is detected from an image obtained by processing such as imaging, the present invention is arranged in the vertical direction at the time of imaging. A plurality of positions are detected as candidates, a range of candidate detection targets is detected based on the luminance of the pixel for each column of pixels arranged in the horizontal direction corresponding to each detected position, and Image processing method capable of diversifying detection methods and improving detection accuracy by specifying detection targets from detection target candidates based on length, and image processing apparatus to which the image processing method is applied
- An object of the present invention is to provide an image processing system including the image processing apparatus and a computer program for realizing the image processing apparatus.
- the luminance of one pixel is added based on the luminance of each other adjacent pixel, and at a position away from the one pixel by a predetermined distance in the horizontal and vertical directions at the time of imaging.
- the present invention provides a horizontal range based on a result of accumulating changes in luminance of pixels arranged in the horizontal direction in the vertical direction during imaging, and changes in luminance of pixels arranged in the vertical direction in the horizontal direction.
- An image processing method, an image processing apparatus, an image processing system, and a computer program capable of diversifying detection methods and improving detection accuracy by detecting a detection target as a vertical range based on the integrated result. The provision is for another purpose.
- the present invention selects an effective detection method according to the situation by determining the priority order of the detection method based on the average value and the variance value of the luminance, thereby improving the detection accuracy.
- Still another object is to provide an image processing method, an image processing apparatus, an image processing system, and a computer program capable of performing the above.
- the image processing method is an image processing method for detecting a specific detection target from a two-dimensional image in which pixels are arranged in different first and second directions, respectively.
- the luminance is integrated to derive the change in the integrated value in the second direction, and based on the calculated change in the integrated value, multiple positions in the second direction are detected and detected as positions corresponding to the detection target candidates.
- a range in the first direction based on the luminance of the pixel is detected as a candidate for detection, and based on the length of the detected range, A detection target is specified from detection target candidates.
- An image processing method is the image processing method for detecting a specific detection target from a two-dimensional image in which pixels are arranged in different first and second directions, respectively. Addition based on the luminance of each other adjacent pixel, and subtraction based on the luminance of the pixel located at a predetermined distance in the first direction from the one pixel and the luminance of the pixel located at a predetermined distance in the second direction. The conversion is based on the result, and the detection target is detected based on the converted result.
- the image processing method is an image processing method for detecting a specific detection target from a two-dimensional image in which pixels are arranged in different first and second directions, respectively.
- the numerical value based on the change in brightness is integrated in the second direction to derive the change in the integrated value in the first direction, and the numerical value based on the change in luminance of the pixels arranged in the second direction is integrated in the first direction.
- Based on the range in the first direction based on the derived change in the integrated value in the first direction and the range in the second direction based on the derived change in the integrated value in the second direction.
- a detection target is detected.
- An image processing method is an image processing method for detecting a specific detection target from a two-dimensional image including a plurality of pixels by a plurality of detection methods, and calculating an average value of the luminances of the pixels.
- the dispersion value of the luminance of the pixel is calculated, and the priority of the detection method is determined based on the calculated average value and dispersion value.
- An image processing apparatus is an image processing apparatus for detecting a specific detection target from a two-dimensional image in which pixels are arranged in different first and second directions, and the pixels arranged in the first direction.
- Derivation means for deriving the change in the integrated value in the second direction by integrating the luminance, and based on the derived change in the integrated value, the position corresponding to the detection target candidate in the second direction
- Candidate detection means that detects multiple positions and the range of the first direction based on the luminance of the pixels as a candidate for detection for each row of pixels aligned in the first direction corresponding to each detected position
- the candidate detecting means detects a plurality of positions indicating minimum values of change in the integrated value of luminance of the pixels integrated in the first direction. It is configured as described above.
- An image processing apparatus is the image processing apparatus according to the sixth invention, wherein the means for secondarily differentiating the change in the integrated value derived by the derivation means, and the candidate detection means include a plurality of positions indicating minimum values. In the apparatus, a predetermined number of positions are detected based on the result of the second derivative.
- the range detecting means detects the range based on a change in luminance of pixels arranged in the first direction. It is comprised by these.
- An image processing apparatus is the image processing apparatus according to any one of the fifth to eighth inventions, wherein the first direction range of the detection region includes the detection target and the first direction range is larger than the detection target.
- Detecting means wherein the specifying means detects the length of the detection target candidate detected by the range detection means in the first direction and the length of the detection area including the detection target in the first direction. It is characterized in that it is configured to identify the detection target based on the result of comparing the lengths.
- An image processing apparatus is the integrated value in the first direction by integrating the luminance of the pixels arranged in the second direction related to the specified detection object in any of the fifth to ninth inventions.
- a means for deriving a change in the value a minimum value detecting means for detecting a minimum value from a change in the integrated value in the first direction, a means for counting the number of detected minimum values, and a case where the counted number is less than a predetermined number.
- the apparatus further comprises means for determining that the specified detection target is false.
- An image processing apparatus is the image processing apparatus according to the tenth aspect of the invention, comprising pixels corresponding to the minimum value detected by the minimum value detecting means, and having the same luminance as the pixel and continuing in the second direction. If the number of primes exceeds a predetermined number, a means for determining that the identified detection target is false is added. It prepares for.
- An image processing apparatus is an image processing apparatus for detecting a specific detection target from two-dimensional images in which pixels are arranged in different first and second directions, wherein the luminance of one pixel is set. , Addition based on the brightness of each other adjacent pixel, and brightness of a pixel located a predetermined distance away from the one pixel in the first direction and brightness of a pixel located a predetermined distance away in the second direction. It comprises a means for converting based on the result of subtraction based on, and a detecting means for detecting a detection target based on the result of conversion.
- An image processing apparatus is characterized in that, in the twelfth invention, the detection means is configured to detect a pixel having a minimum value after conversion as a detection target.
- An image processing device is the image processing device for detecting a specific detection target from two-dimensional images in which the pixels are arranged in different first and second directions, and the pixels arranged in the first direction.
- the first derivation means for deriving the change in the integrated value in the first direction by integrating the numerical values based on the change in luminance in the second direction, and the numerical value based on the change in luminance of the pixels arranged in the second direction in the first direction
- Second derivation means for deriving a change in the integrated value in the second direction by integrating to the first direction, a range in the first direction based on a change in the integrated value in the first direction derived by the first derivation means, and the second derivation And detecting means for detecting a detection object based on a range in the second direction based on a change in the integrated value in the second direction derived by the means.
- the first derivation means includes a numerical value based on a luminance difference from pixels adjacent in the first direction and pixels adjacent in the first direction.
- An indicator based on a numerical value indicating low brightness is integrated in the second direction to derive a change in the integrated value in the first direction
- the second derivation means is adjacent to the second direction.
- a numerical value based on the luminance difference from the target pixel and an indicator based on a numerical value indicating the low brightness of the neighboring pixels in the second direction are integrated in the first direction to derive the change in the integrated value in the second direction.
- the detection means is configured such that the integrated value derived by the first deriving means ranges in the first direction from the position where the integrated value is maximized to the position where the integrated value is minimized, and the second deriving means derives The integrated value is determined based on the range in the second direction from the maximum position to the minimum position. It is comprised so that it may detect.
- An image processing device is a specific test from a two-dimensional image including a plurality of pixels.
- a means for calculating an average value of pixel luminance, a means for calculating a variance value of pixel luminance, and the calculated average value and variance value are used.
- means for determining the priority of the detection method are used.
- An image processing system includes the image processing device according to any one of the fifth to sixteenth inventions, and an imaging device that generates an image processed by the image processing device.
- the detection target is a region including a nostril of a person in an image captured by the imaging device, the first direction is a horizontal direction, and the second direction is a vertical direction. To do.
- a computer program according to an eighteenth aspect of the invention is a computer program that causes a computer to detect a specific detection target from two-dimensional images in which pixels are arranged in different first and second directions, respectively.
- the procedure for deriving the change in the integrated value in the second direction by integrating the luminance of the pixels arranged in one direction and the computer as the position corresponding to the detection target candidate based on the derived change in the integrated value.
- a computer program according to a nineteenth invention is a computer program for causing a computer to detect a specific detection target from two-dimensional images in which pixels are arranged in different first and second directions, respectively.
- the luminance of the pixel is added based on the luminance of each other adjacent pixel, and the luminance of the pixel at a predetermined distance in the first direction from the one pixel and the luminance of the pixel at a predetermined distance in the second direction. It is characterized in that a procedure for performing conversion based on a result of subtraction based on pixel luminance and a procedure for causing a computer to detect a detection target based on the result of conversion are characterized.
- a computer program according to a twentieth invention is a computer program that causes a computer to detect a specific detection target from a two-dimensional image in which pixels are arranged in different first and second directions, respectively.
- a procedure for integrating the numerical values based on the change in luminance of pixels arranged in one direction in the second direction to derive the change in the integrated value in the first direction and A procedure for causing the computer to integrate the numerical value based on the change in luminance of the pixels arranged in the second direction in the first direction to derive the change in the integrated value in the second direction, and for the computer to calculate the derived integrated value in the first direction.
- a procedure for detecting a detection target based on a range in the first direction based on the change and a range in the second direction based on the derived change in the integrated value in the second direction is a computer program that causes a computer to detect a specific detection target from a two-dimensional image in which pixels are arranged in different first and second directions, respectively.
- a computer program according to a twenty-first invention is a computer program for causing a computer to detect a specific detection target from a two-dimensional image including a plurality of pixels by a plurality of detection methods.
- the region to the right and left noses including the nostrils in the face of a person is detected.
- the target is the horizontal direction and the vertical direction are the first direction and the second direction, respectively, paying attention to the luminance distribution in the vertical direction, a plurality of candidates including the eyebrows, eyes, and mouth are detected.
- the detection method position can be diversified to comprehensively determine the position of the detection target, thus further increasing the detection accuracy. It is possible.
- a region including a nostril and other regions are detected by detecting a range based on the luminance of pixels that are arranged in a horizontal direction in a plurality of candidates including regions such as eyebrows, eyes, and mouth. It is possible to detect so that the difference in the horizontal range from the difference is clear, and it is possible to improve the detection accuracy.
- the width of the face which is a detection region including the nostril to be detected, is detected and compared with the width of the face, for example, the horizontal range is 22% to 43% of the face width.
- the number of local minimum values in the horizontal direction that is, a portion that is highly likely to be a nostril If the number of positions is less than a predetermined number such as 2 set assuming the left and right nostrils, it is determined that the possibility of being a nostril is low and it is not a detection target. It is possible to reduce the possibility of erroneous detection of determining a site as a nostril.
- the eleventh invention if the continuity in the vertical direction exceeds a threshold value indicated by a predetermined number, it is determined that there is a high possibility that an ophthalmic frame is detected, and it is determined that it is not a detection target. Thus, the possibility of erroneous detection can be reduced.
- the twelfth invention, the thirteenth invention and the nineteenth invention for example, in a captured image, a nostril in a human face is a detection target, and the horizontal direction and the vertical direction. Is detected by performing a conversion process that emphasizes the area where the brightness of neighboring pixels is low and the surrounding brightness is high, that is, the small area where the brightness is low. Can be detected with high accuracy. In particular, when used in combination with other methods for detecting a detection target, the detection method can be diversified to comprehensively determine the position of the detection target, so that the detection accuracy can be further improved.
- the fourteenth invention, the fifteenth invention and the twentieth invention for example, in a captured image, a region around a nostril in a person's face is set as a detection target.
- the direction and the vertical direction are the first direction and the second direction, respectively, the portion where the luminance decrease is large and the portion where the luminance increase is large are small compared to the vertical direction and the horizontal direction, respectively. It is possible to detect a detection target with high accuracy by deriving a simple numerical value and detecting a detection target that is a rectangular region whose luminance is lower than that of the surroundings based on the derived numerical value.
- the nostril itself detects the downward area around the nostril, so it is possible to detect the detection target even when the face is tilted at an angle where the nostril is difficult to detect, especially when used in combination with other methods. Since the detection method can be diversified to comprehensively determine the position of the detection target, the detection accuracy can be further improved.
- the detection method can be diversified to comprehensively determine the position of the detection target, the detection accuracy can be further improved.
- the area is narrowed down based on the positional relationship with the detected part and then the detection according to the invention is performed. By performing this, it is possible to detect the detection target with higher accuracy.
- the sixteenth invention and the twenty-first invention for example, when detecting a detection target such as a nostril in a human face in a captured image, the average value and the variance value of the luminance are obtained. Base Then, it is determined whether there is a local change in illuminance on the face of the person, and the priority of the detection method is determined according to the determined status, so that the status can be selected from various detection methods. Therefore, it is possible to select a detection method and a detection order with high reliability according to the detection accuracy, so that the detection accuracy can be further improved.
- the region including the nostril of a person can be detected with high accuracy.
- the region obtained by imaging the driver's face using an imaging device such as an in-vehicle camera mounted on the vehicle can be applied to a system that detects a driver's face as a detection target from an image, and can be deployed to a system that detects a driver's situation and provides driving assistance such as a side-viewing warning.
- An image processing method, an image processing apparatus, an image processing system, and a computer program according to the present invention include, for example, an image obtained by imaging a driver's face with an imaging device such as an in-vehicle camera mounted on a vehicle. It is applied to the form where the area from the nostril to the left and right noses in the face of a person is detected.
- the image processing apparatus or the like according to the present invention derives a change in the integrated value in the vertical direction, which is the second direction, by integrating the luminance of the pixels arranged in the horizontal direction, which is the first direction.
- a plurality of positions that are minimal values are detected as candidates including the detection target, and further, the change in the intermediate force integrated value is second-order differentiated to narrow it down to a predetermined number, and in the first direction corresponding to each narrowed position.
- a range in the first direction that is a candidate for detection is detected based on the luminance of the pixel, and a detection target is selected from the detection target candidates based on the length of the detected range. Is identified.
- the present invention focuses on the vertical luminance distribution, detects a plurality of candidates including low-brightness eyebrows, eyes, and mouth, and selects from the candidates based on the horizontal width.
- the detection target By specifying the detection target, it is possible to detect the detection target with high accuracy, and so on.
- the detection method can be comprehensively determined by diversifying the detection method. It has excellent effects, such as being able to further improve accuracy.
- the present invention improves the accuracy of detection of a detection target, thereby When applied to systems that accurately detect and support driving such as side-viewing warnings, reliable driving assistance with few false detections even when driving in an environment where the ambient light conditions constantly change It is possible to construct a system, and it has excellent effects.
- the image processing apparatus or the like detects the width of the face, and compares the face width, for example, detects a candidate whose horizontal range is 22% to 43% with respect to the face width.
- a candidate whose horizontal range is 22% to 43% with respect to the face width.
- the vertical detection line for the detection target in the column of pixels arranged in the horizontal direction included in the specified detection target is less than the predetermined number indicating the number of right and left nostrils, that is, less than two.
- the predetermined number indicating the number of right and left nostrils that is, less than two.
- the image processing apparatus or the like if the number of pixels arranged in the vertical direction exceeds a predetermined number, it is determined that there is a high possibility that a spectacle frame is detected, It is possible to reduce the possibility of erroneous detection, and so on.
- An image processing method, an image processing device, an image processing system, and a computer program according to the present invention are obtained from an image obtained by imaging a driver's face by an imaging device such as an in-vehicle camera mounted on a vehicle, for example. Applies to forms that detect nostrils in human faces The Then, the image processing apparatus or the like according to the present invention adds the luminance of one pixel based on the luminance of each adjacent pixel, and a predetermined distance in the horizontal direction that is the first direction from the position of the one pixel.
- Conversion is performed based on the result of subtraction based on the luminance of the pixel at a distant position and the pixel at a predetermined distance in the vertical direction, which is the second direction, and the pixel with the smallest value after the conversion is detected as a detection target.
- the present invention performs a conversion process for emphasizing a portion where the luminance of neighboring pixels is low and the surrounding luminance is high, that is, a small region with low luminance. It has excellent effects such as being able to detect the nostril that is a region with high accuracy.
- the present invention can detect the nostril with high accuracy when the face is facing upward, so when used in combination with other methods for detecting the detection target, the detection method is diversified to integrate the position of the detection target. Therefore, the detection accuracy can be further improved, and an excellent effect can be obtained.
- the present invention when the present invention is applied to a system that accurately detects the situation of the driver by improving the accuracy of detection of the detection target and provides driving assistance such as a warning for looking aside, Even when driving in an environment where the conditions of the vehicle constantly change, it is possible to construct a reliable driving support system with few false detections.
- An image processing method, an image processing device, an image processing system, and a computer program according to the present invention are obtained from an image obtained by imaging a driver's face with an imaging device such as an in-vehicle camera mounted on a vehicle, for example. It is applied to a form in which a downward area around a nostril in a human face is detected.
- the image processing apparatus according to the present invention multiplies a numerical value based on a luminance difference between adjacent pixels in the horizontal direction that is the first direction and a numerical value indicating low brightness of the adjacent pixels in the first direction.
- the index is integrated in the vertical direction, which is the second direction, to derive the change in the integrated value in the first direction, and the numerical value based on the luminance difference between adjacent pixels in the second direction and adjacent in the second direction.
- the index multiplied by the numerical value indicating the low brightness of the pixel is integrated in the first direction to derive the change in the integrated value in the second direction, and each derived integrated value is minimized from the maximum position.
- the detection target is detected based on the range in the first direction and the range in the second direction up to the position.
- the present invention reduces the luminance with respect to each of the horizontal direction and the vertical direction.
- a numerical value is derived such that a large part becomes large and a part where the increase in luminance is large becomes small, and based on the derived numerical value, a detection target that is a rectangular area whose luminance is lower than the surrounding area is detected.
- the detection target is detected even when the face is tilted at an angle where the nostril is difficult to detect. It is possible to achieve an excellent effect.
- the present invention when a part such as the position of both eyes or the position of the nose is detected by another method, the area is narrowed down based on the positional relationship with the detected part, and then the invention is applied.
- the detection method can be comprehensively determined by diversifying the detection methods, and thus the detection accuracy can be further improved. Etc. have excellent effects.
- the present invention When the present invention is applied to a system that accurately detects the situation of the driver by improving the accuracy of detection of the detection target and provides driving assistance such as a warning for looking aside, Even when driving in an environment where the conditions of the vehicle constantly change, it is possible to construct a reliable driving support system with few false detections.
- An image processing method, an image processing device, an image processing system, and a computer program according to the present invention are obtained from an image obtained by imaging a driver's face with an imaging device such as an in-vehicle camera mounted on a vehicle, for example. It is applied to a form in which an area including a nostril of a person's face is a detection target.
- the image processing apparatus or the like according to the present invention calculates an average value of pixel luminances, calculates a variance value of pixel luminances, and based on the calculated average value and variance value, from among a plurality of detection methods. Determine the priority of detection methods.
- the present invention determines whether or not a local change in illuminance has occurred on the face of a person based on the average value and the variance value of the luminance, and detects it according to the determined situation.
- the priority of the method it is possible to select a highly reliable detection method and detection order according to the situation from various detection methods, so that the detection accuracy can be further improved, etc. Has an excellent effect.
- FIG. 1 is a block diagram showing a configuration example of an image processing system in Embodiment 1 of the present invention.
- FIG. 2 is a flowchart showing an example of processing of the image processing device used in the image processing system according to Embodiment 1 of the present invention.
- FIG. 3 is a flowchart showing an example of processing of the image processing apparatus used in the image processing system according to Embodiment 1 of the present invention.
- FIG. 4 is an explanatory diagram conceptually showing an example of processing up to detection of a candidate for a detection target of an image processing range of the image processing system in Embodiment 1 of the present invention.
- FIG. 5 is an explanatory diagram schematically showing an example of a range in which an edge detection process of the image processing system in Embodiment 1 of the present invention is performed.
- FIG. 6 is an explanatory diagram showing an example of coefficients used for edge detection processing of the image processing system according to Embodiment 1 of the present invention.
- FIG. 7 is an explanatory diagram schematically showing an example of a range in which an edge detection process of the image processing system in the first embodiment of the present invention is performed.
- FIG. 8 is an explanatory diagram schematically showing an example of a range in which an edge detection process of the image processing system in the first embodiment of the present invention is performed.
- FIG. 9 is an explanatory diagram showing detection target candidates of the image processing system in Embodiment 1 of the present invention.
- FIG. 10 is an explanatory diagram conceptually showing a nostril region score of the image processing system in Embodiment 1 of the present invention.
- FIG. 11 is a block diagram showing a configuration example of an image processing system in Embodiment 2 of the present invention.
- FIG. 12 is a flowchart showing an example of processing of the image processing device used in the image processing system in the second embodiment of the present invention.
- FIG. 13 is an explanatory diagram conceptually showing an example of setting a detection range of the image processing system in the second embodiment of the present invention.
- FIG. 14 is an explanatory diagram conceptually showing an example of setting a search range of the image processing system in Embodiment 2 of the present invention.
- FIG. 15 is an explanatory diagram showing an example of coefficients used for black region calculation filter processing of the image processing system in Embodiment 2 of the present invention.
- FIG. 16 is an explanatory diagram conceptually showing an example of detection using black region calculation filter processing of the image processing system in Embodiment 2 of the present invention.
- FIG. 17 is a block diagram showing a configuration example of an image processing system in Embodiment 3 of the present invention.
- FIG. 18 is a flowchart showing an example of processing of the image processing device used in the image processing system in the third embodiment of the present invention.
- FIG. 19 is an explanatory diagram conceptually showing an example of setting a search range of an image processing system in Embodiment 3 of the present invention.
- FIG. 20 is an explanatory diagram showing an example of coefficients used for horizontal edge filter processing in the image processing system according to Embodiment 3 of the present invention.
- FIG. 21 is an explanatory diagram showing an example of coefficients used for vertical edge filter processing in the image processing system according to Embodiment 3 of the present invention.
- FIG. 22 is an explanatory diagram showing a result of detection by the image processing system in the third embodiment of the present invention.
- FIG. 23 is a block diagram showing a configuration example of an image processing system in Embodiment 4 of the present invention.
- FIG. 24 is a flowchart showing an example of processing of the image processing device 2 used in the image processing system in the fourth embodiment of the present invention.
- FIG. 1 is a block diagram showing a configuration example of an image processing system according to Embodiment 1 of the present invention.
- 1 is an imaging device such as an in-vehicle camera mounted on a vehicle, and the imaging device 1 is configured by a communication line such as a dedicated cable or a wired or wireless line to the image processing device 2 that performs image processing. It is connected via a communication network such as in-vehicle LAN (Local Area Network).
- the imaging device 1 is disposed in front of a driver such as a steering wheel and a dashboard in a vehicle, and can capture images so that the horizontal and vertical sides of the driver's face are in the horizontal and vertical directions of the image. It has been adjusted to the state.
- the image pickup apparatus 1 includes an MPU (Micro Processor Unit) ll that controls the entire apparatus, various computer programs executed based on the control of the MPU11, and a ROM (Read Only Memory) 12 that records data, and a ROM 12 that records A random access memory (RAM) 13 that records various data temporarily generated when the computer program is executed, an imaging unit 14 configured using an imaging device such as a CCD (Charge Coupled Device), and an imaging unit 1 A / D converter 15 that converts analog image data obtained by imaging 4 into digital data, and a frame memory that temporarily stores the image data converted to digital by the A / D converter 15 16 And a communication interface 17 used for communication with the image processing apparatus 2.
- MPU Micro Processor Unit
- ROM Read Only Memory
- RAM random access memory
- the imaging unit 14 performs imaging processing continuously or intermittently, and generates, for example, 30 image data (image frames) per second based on the imaging processing to perform A / D
- the A / D converter 15 converts each pixel constituting the image into digital image data represented by a gradation such as 256 gradations (lBy te) and records it in the frame memory 16. Make it.
- the image data recorded in the frame memory 16 is output from the communication interface 17 to the image processing device 2 at a predetermined timing.
- Each pixel constituting the image is two-dimensionally arranged, and the image data includes the position of each pixel indicated by a plane rectangular coordinate system, a so-called xy coordinate system, and the luminance of each pixel indicated as a gradation value. It contains data indicating. Incidentally it may be as shown coordinates the order in which are arranged the Nag Data than indicating the coordinates in accordance with respective X y coordinates for each pixel.
- the horizontal direction of the image corresponds to the X-axis direction of the image data, and the vertical direction of the image corresponds to the y-axis direction of the image data.
- the image processing apparatus 2 includes a CPU (Central Processing Unit) 21 that controls the entire apparatus,
- the computer program 31 according to the first embodiment of the present invention and an auxiliary recording unit 22 such as a CD-ROM drive for reading information from a recording medium 41 such as a CD-ROM recording various information such as data and data, and an auxiliary recording unit 22
- a hard disk (hereinafter referred to as HD) 23 that records various information read by the computer
- a RAM 24 that records various data temporarily generated when the computer program 31 recorded on the HD 23 is executed
- a volatile memory A frame memory 25 and a communication interface 26 used for communication with the imaging device 1 are provided.
- the computer program 31 of the present invention and various types of information such as data are read from the HD 23, recorded in the RAM 24, and various procedures included in the computer program 31 are executed by the CPU 21. It operates as the image processing device 2 of Data recorded on the HD 23 includes data related to execution of the computer program 31, for example, various data such as mathematical expressions, filters, and various constants described later, and data that indicates detected detection targets or detection target candidates. There is.
- the image data output from the imaging apparatus 1 is transferred to the communication interface.
- the received image data is recorded in the frame memory 25, and the image data recorded in the frame memory 25 is read out to perform various image processing.
- the various image processes to be performed on the received image data are various processes necessary for detecting areas such as the contour of the driver's face, eyes, and nose from the image data.
- the brightness of the image arranged in the vertical direction of the image is integrated, and the integrated value is compared with a predetermined threshold value.
- a contour width detection process for detecting a horizontal range can be mentioned.
- there is an outline width detection process that identifies the position where the change is large by differentiating the horizontal change of the integrated value and detects the boundary between the background and the face outline where the brightness changes greatly. it can.
- FIGS. 1 and 3 are flowcharts showing an example of processing of the image processing apparatus 2 used in the image processing system according to Embodiment 1 of the present invention.
- the image processing device 2 extracts from the frame memory 25 the image data obtained through the imaging of the imaging device 1 and received via the communication interface 26 under the control of the CPU 21 that executes the computer program 31 recorded in the RAM 24 ( S101), from the extracted image data, the width of the driver's face, ie, the horizontal range (first direction) that is the boundary of the region showing the face is detected by, for example, the above-described contour width detection process (S102). ), A range of image processing to be performed thereafter is set based on the detected result (S103). The area range (contour width) and image processing range detected in step S102 are recorded in the HD 23 or RAM 24.
- the image processing apparatus 2 determines the brightness of the pixels arranged in the horizontal direction (first direction) with respect to the image in the range extracted in step S101 and set in step S103. Accumulated (S104), and the resulting resultant force also derives a change in the integrated value in the vertical direction (second direction) (S105), and the vertical direction corresponding to the detection target candidate is derived from the change in the integrated value in the derived vertical direction. A plurality of positions indicating minimum values are detected as the positions at (S106). In step S106, a plurality of candidates including low-brightness eyebrows, eyes, and mouth are detected only in the region up to the left and right noses including the nostril that is the original detection target.
- the image processing device 2 secondarily differentiates the change in the integrated value derived in step S105 (S107), and at a plurality of positions indicating the minimum values detected in step S106. Among them, a predetermined number of positions, such as ten, are detected as candidate positions including detection targets in the vertical direction in order from the lowest secondary differential value (S108).
- the process of step S107 for secondarily differentiating the change in the integrated value is performed using, for example, the following Equation 1.
- Step S107 is processing for narrowing down the candidates detected in step S106, and a maximum of 10 candidates are detected by the processing of steps S104 and S108.
- the number of candidates detected in step S106 is less than 10
- the number of candidates detected in step S108 must be less than 10. Needless to say.
- the predetermined number is a numerical value that can be changed according to the necessity recorded in advance in the HD 23 or the RAM 24. Data indicating the candidate detected in step S108 is recorded in the HD 23 or RAM 24.
- Second derivative ? ⁇ ) '2_? ⁇ _8) _? + 8)... Equation 1
- the image processing apparatus 2 reads each position in the vertical direction, which is the detection target candidate detected in step S108, from the HD 23 or the RAM 24, and corresponds to each position read.
- the horizontal end of the detection target candidate that is, the left and right ends
- S109 the change in the luminance of the pixel
- S110 a horizontal range that is a candidate for detection
- step S109 is performed by detecting, as an end, a point where a state where pixels having lower luminance than the surroundings are interrupted is interrupted. Further, the processing in steps S109 to S110 is performed for all detection target candidates, and further narrowing down of the detection target candidates and range detection are performed.
- the image processing apparatus 2 detects the length of the horizontal range of the detection target candidates detected and recorded in step S110, and the driver's detected and recorded in step S102. Compare the length of the horizontal range (contour width) of the face area. Among the candidates for detection, the length of the horizontal range is the length of the horizontal range of the driver's face area. On the other hand, the detection target candidates that fall within the range of 22% to 43% are identified as detection targets (SI 11), and the contents recorded as detection target candidates in HD23 or RAM24 are updated.
- SI 11 detection targets
- step S111 for all detection target candidates, the detection area including the detection target, that is, the face area, is compared to the length of the horizontal range of the detection target, that is, the region to the left and right nostrils including the nostril. It is determined whether or not the horizontal length falls within a predetermined range, here 22% to 43%, and detection candidates that fall within the predetermined range are set as detection targets. To detect.
- step SI 11 when there are a plurality of detection target candidates that fall within the predetermined range, based on the index (horizontal edge score described later) used during the range detection process shown in steps S 109 to S 110, The candidate for detection is specified.
- the numerical value of 22 43% shown as the predetermined range is a numerical value that is appropriately set according to factors such as the race of the driver that are not fixed.
- the image processing device 2 includes the detection target specified in step S111 under the control of the CPU 21, the horizontal width matches the detection target, and the vertical width is a predetermined pixel such as 3, 5 or the like.
- Set the number of test areas (S112) integrate the luminance of the pixels aligned in the vertical direction within the set test area, derive the change in the horizontal integrated value (S113), and derive the horizontal integration
- the minimum value is detected from the change in value (S114), the number of detected minimum values is counted (S115), and the number of the counted minimum values is determined to be a predetermined number, here, whether or not it is less than two. (S116).
- step S115 when the number of counted minimum values is less than the predetermined number (S116: YES), the image processing apparatus 2 determines that the detection target specified in step S111 is false under the control of the CPU 21, and the detection target It is determined that the detection is not possible (S117), and based on the determination result, the recorded contents of the HD 23 or RAM 24 are updated, and the process is terminated.
- Detected force to be detected If the region includes the nostril and the left and right nostrils, the position corresponding to the nostril has a minimum value in the change in the integrated value derived in step S113. The value can be counted. Therefore, when the minimum value is 0 or 1, it is determined that the specified detection target is false. Since there is a possibility that the position near the outline of the nose will become a shadow and indicate a minimum value, even if there are three or more minimum values, it will not be judged as false.
- step S116 when the number of counted minimum values is equal to or more than the predetermined number 2 (S116: NO), the image processing apparatus 2 includes a pixel corresponding to the minimum value under the control of the CPU 21.
- the number of pixels that have the same brightness as that of the pixel and are continuous in the vertical direction is counted (S 118), and whether or not the number of consecutive pixels counted is equal to or greater than a predetermined number is preset. Judgment is made (S119).
- step S118 the continuity in the vertical direction of pixels with low luminance is determined.
- counting the number of pixels is intended to determine the continuity of pixels with low luminance, so it is not always necessary to count only the pixels with the same luminance as the pixels corresponding to the minimum value.
- the luminance is expressed as a gradation classified into 256 levels
- the pixel level indicating the minimum value is displayed.
- the key is 20, it is desirable to count the continuity at a gradation with a width of 20 ⁇ 5.
- step S119 when the number of consecutive pixels counted exceeds a predetermined number set in advance (S119: YES), the image processing apparatus 2 performs the detection specified in step S111 under the control of the CPU 21.
- the target is false, and it is determined that the detection target cannot be detected (S117).
- the recorded content of the HD 23 or RAM 24 is updated, and the process is terminated. If the continuity of the low-luminance pixels is equal to or greater than the predetermined number, it is determined that the frame of the glasses is erroneously detected.
- step S119 when the number of consecutive pixels counted is equal to or less than a predetermined number set in advance (S119: NO), the image processing apparatus 2 is identified in step S111 under the control of the CPU 21. It is determined that the detected object is true and the detected object has been detected (S120), and the recorded content of the HD 23 or RAM 24 is updated based on the determined result, and the process is terminated.
- FIG. 4 is an explanatory diagram conceptually showing an example of processing from determination of the image processing range of the image processing system to detection of a detection target candidate in the first embodiment of the present invention.
- Fig. 4 (a) shows the situation where the processing range of the image is determined.
- the outer frame force indicated by the solid line S of the image indicated by the image data extracted in step S101 is shown.
- the whole image of the image of the driver's face and the area to the left and right noses including the nostrils in the driver's face to be detected are shown.
- the line in the vertical direction (y-axis direction) of the image indicated by the alternate long and short dash line is the range of the area detected in step S102, that is, the width of the contour of the driver's face.
- the range of the image processing set in step S103 is a region force surrounded by the width of the face outline indicated by the one-dot chain line and the upper and lower frames in the entire image indicated by the solid line.
- FIG. 4 (b) is a graph showing the distribution of the integrated value of the luminance in the vertical direction derived in step S105, by integrating the luminance of the pixels arranged in the horizontal direction in step S104.
- Figure 4 (b) Fig. 4 (a) shows the distribution of the integrated luminance value in the vertical direction of the image. The vertical axis shows the vertical coordinate corresponding to Fig. 4 (a), and the horizontal axis shows the integrated luminance value. Show.
- the integrated value of the luminance in the vertical direction changes so as to take the minimum value indicated by the arrow at the parts such as the eyebrows, eyes, nostril, mouth, etc. It can be seen that the candidate for detection can be detected based on the local minimum.
- FIG. 5 is an explanatory diagram schematically showing an example of a range in which the edge detection process of the image processing system according to the first embodiment of the present invention is performed
- FIG. 6 is an image process according to the first embodiment of the present invention.
- It is explanatory drawing which shows the example of the coefficient used for the edge part detection process of a system.
- Fig. 5 shows the pixels around the detection target candidate.
- the numbers shown in the upper part of Fig. 5 indicate the horizontal position of the pixel, that is, the X coordinate, and the symbol on the left indicates the vertical direction of the pixel. Indicates the position of y, ie the y coordinate.
- the pixel ⁇ IJy which is the value of y-coordinate indicated by the diagonal lines from the upper right to the lower left in the horizontal direction in Fig. 5, indicates the candidate for detection, and the pixel column y and the pixel column y ⁇ ijy + 1 and column y ⁇ 1 of pixels whose y coordinate values are y + 1 and y ⁇ 1, respectively, shown by diagonal lines from the upper left to the lower right adjacent to are used for edge detection. Then, by multiplying each pixel included in the pixel column y, y + 1, y-1 by the coefficient shown in Fig. 6, the edge of the region where the pixels are arranged in the horizontal direction has low luminance and the pixels are continuous. Functions as a clarified horizontal edge filter.
- the horizontal edge coefficient of the pixel at the center position is calculated as the horizontal edge coefficient of the pixel at the center position.
- the horizontal edge coefficient is obtained by multiplying the brightness of the pixel adjacent to the left by "1" and the brightness of the pixel adjacent to the right by "1". It is obtained by adding numerical values.
- the 3 X 3 area indicated by the bold line in FIG. 5 shows a state in which the horizontal edge filter is associated with the pixel indicated by the coordinates (2, y + 1), and the coordinates (2, y + 1)
- the horizontal edge coefficient of the pixel indicated by () is calculated using the following equation (2).
- horizontal edge coefficients are calculated for the pixels included in the horizontal rows ⁇ ijy of the pixels arranged in the horizontal direction that are candidates for detection and the pixels y-1 and y + 1 adjacent to the upper and lower sides thereof. .
- FIG. 7 is an explanatory diagram schematically showing an example of a range in which the edge detection process of the image processing system in the first embodiment of the present invention is performed.
- the numbers shown in the upper part of FIG. 7 indicate the horizontal position of the pixel, that is, the X coordinate, and the symbols shown on the left indicate the vertical position of the pixel, that is, the y coordinate.
- the pixel ⁇ ijy with the y-coordinate value power Sy in Fig. 7 indicates the candidate for detection, and the y-coordinate values adjacent to the pixel row y and the pixel ⁇ 'Jy above and below are respectively
- the lateral edge coefficients are calculated for ⁇ ljy + l and column y ⁇ 1 of the pixels that are y + 1 and y ⁇ 1.
- the horizontal edge coefficient is calculated for one pixel and nine pixels totaling eight pixels adjacent to the pixel.
- a predetermined threshold value set in advance is compared.
- an index indicating the number of pixels whose horizontal edge coefficient exceeds a predetermined threshold is calculated as the horizontal edge score of one pixel.
- a 3 ⁇ 3 region surrounded by a thick line in FIG. 7 indicates a pixel required for calculating the horizontal edge score of the pixel indicated by the coordinates (3, y).
- the number of pixels having a horizontal edge coefficient exceeding the threshold is the horizontal edge score of the pixel indicated by coordinates (3, y).
- a pixel having a horizontal edge score indicating a value of 0-9 is 5 or more is determined to be within the horizontal range of the detection target candidate. That is, the image processing apparatus 2 detects the leftmost pixel having a horizontal edge score of 5 or more as the left end in the horizontal direction of the detection target candidate in step S 109, and the horizontal edge score is 5 One or more rightmost pixels are detected as being the right end of the detection target candidate in the horizontal direction. In step S 110, a horizontal range of detection target candidates is detected based on the detected left and right ends.
- FIG. 8 shows the edge detection process of the image processing system according to Embodiment 1 of the present invention. It is explanatory drawing which shows the example of the range to be shown typically.
- FIG. 8 shows the candidate pixels to be detected and the horizontal edge score, and the numbers shown in the upper part of FIG. 8 indicate the horizontal positions of the pixels, that is, the X coordinates.
- the range from the pixel with the X coordinate of 5 to the pixel with the X coordinate of 636 is detected as the horizontal range to be detected.
- FIG. 9 is an explanatory diagram showing detection target candidates of the image processing system according to Embodiment 1 of the present invention.
- FIG. 9 shows an image of the driver's face and detection target candidates in which the horizontal range is detected by the processing in steps S109 to S110.
- the X mark indicates the left and right edges detected at step 109, and the line segment connecting the left and right edges indicated by the X marks indicates detection target candidates. Candidates whose left and right edges are not detected at this stage are excluded from the detection targets.
- the eyebrows, eyes, nostrils, mouth and chin positions are candidates for detection.
- the horizontal edge score is a predetermined value among the horizontal pixels forming each identified detection target. The number of pixels as described above is counted for each specified detection target, and an index indicated by the obtained numerical value is used as a nostril region score. Then, the detection target having the maximum nostril area score is set as a true detection target, and the other detection targets are excluded as false.
- FIG. 10 is an explanatory diagram schematically showing a nostril region score of the image processing system in the first embodiment of the present invention.
- FIG. 10 shows the pixels included in the horizontal range of the detection target detected in FIG. 8 and the numerical values indicating the horizontal edge score of the pixels.
- the pixels surrounded by a numerical value 0 are horizontal pixels. It shows that the edge score is a predetermined value, here 5 or more. Then, the number of pixels whose horizontal edge score is equal to or greater than a predetermined value is counted as a nostril area score.
- the authenticity of the specified detection target is further determined by the processing from step S112 onward.
- FIG. 11 is a block diagram illustrating a configuration example of the image processing system according to the second embodiment of the present invention.
- 1 is an imaging device, and the imaging device 1 is connected to the image processing device 2 by, for example, a dedicated cable.
- the imaging device 1 includes an MPU 11, a ROM 12, a RAMI 3, an imaging unit 14, an A / D converter 15, a frame memory 16, and a communication interface 17.
- the image processing apparatus 2 includes a CPU 21, an auxiliary recording unit 22, an HD 23, a RAM 24, a frame memory 25, and a communication interface 26, and includes a computer program 32 and data according to the second embodiment of the present invention.
- the image processing apparatus 2 implements the present invention by reading the various information from the recording medium 42 on which various information is recorded by the auxiliary recording unit 22, recording the information on the HD 23, recording it on the RAM 24, and executing it on the CPU 21.
- Various procedures related to Form 2 are executed.
- Embodiment 1 Note that the detailed description of each device is the same as that in Embodiment 1, and therefore, Embodiment 1 will be referred to and description thereof will be omitted.
- FIG. 12 is a flowchart showing an example of processing of the image processing apparatus 2 used in the image processing system according to Embodiment 2 of the present invention.
- the CPU 21 that executes the computer program 32 recorded in the RAM 24 extracts, from the frame memory 25, the image data obtained by the imaging of the imaging device 1 and received via the communication interface 26 (S201).
- the width of the driver's face that is, the horizontal range of the contour that is the boundary of the region showing the face is detected by, for example, the contour width detection process shown in the first embodiment (S202), Further, the area from the left and right nostrils including the nostril is detected as a nostril peripheral area (S203).
- the horizontal range of the contour detected in step S202 and the nostril peripheral area detected in step S203 are recorded in HD 23 or RAM 24.
- the nostril area is detected. For example, the method described in Embodiment 1 is used.
- the image processing device 2 derives, as the minimum point, two points that take the minimum value from the change in luminance of the pixel in the horizontal direction in the peripheral region of the nostril detected in step S203 under the control of the CPU 21 (S204).
- a search range for detecting a detection target is set based on the two derived minimum points (S205).
- the search range set in step S205 is recorded in HD23 or RAM24. If there are 3 or more local minimum points in step S205, 2 points with low brightness are derived as local minimum points.
- the range set in step S205 is 15 pixels for the horizontal direction and 5 pixels for the vertical direction for each of the two points of the derived minimum value. Range. Note that the two minimum points derived in step S204 are considered to be points related to the left and right nostrils, so in step S205, it can be considered that the search range has been set for each of the left and right nostrils. .
- the image processing apparatus 2 adds to all the pixels within the search range set in step S205 based on the luminance of other adjacent pixels and performs predetermined in the horizontal direction.
- the luminance is converted by the black area calculation filter processing that performs subtraction based on the luminance of the pixel at a distance and the luminance of the pixel at a predetermined distance in the vertical direction (S206), and the converted value is minimized.
- a pixel is detected as a detection target (S207), and the detection result is recorded in the HD 23 or RAM 24.
- the processing in steps S206 to S207 is performed on the pixels included in the search ranges of the left and right nostrils.
- FIG. 13 is an explanatory diagram conceptually showing an example of setting a detection range of the image processing system in the second embodiment of the present invention.
- the outer frame force indicated by the solid line S is the entire image indicated by the image data extracted at step S201, and the vertical line (y-axis direction) of the image indicated by the alternate long and short dash line is indicated by step S202. This is the range of the area detected at, that is, the width of the driver's face outline.
- the horizontal line force indicated by the thick solid line is the region up to the left and right nostrils including the nostril detected as the nostril peripheral region in step S203.
- FIG. 14 is an explanatory diagram conceptually showing an example of setting a search range of the image processing system in the second embodiment of the present invention.
- FIG. 14 shows a surrounding image including the nostril.
- the rectangular range indicated by the solid line in FIG. 14 represents the left and right nostrils in step S205. Is a search range set for.
- FIG. 15 is an explanatory diagram showing an example of coefficients used for the black region calculation filter processing of the image processing system according to Embodiment 2 of the present invention.
- “1” is set to a coefficient that is multiplied by the luminance of one pixel to be converted and the luminance of eight adjacent pixels.
- a coefficient that is multiplied by the luminance of each of the two pixels that are arranged at a predetermined distance from each other on the left and right is set to “1”, and the coefficient of each of the two pixels that are arranged at a predetermined distance up and down from one pixel is set.
- the coefficient to be multiplied by the luminance is set to “1”.
- the predetermined distance is set to 1Z18 of the area detected in step S202.
- FIG. 16 is an explanatory diagram conceptually showing an example of detection using the black region calculation filter process of the image processing system in the second embodiment of the present invention.
- FIG. 16 shows a black area calculation filter when black area calculation filter processing is performed on a detection target detected as a nostril on the left side in FIG. 16 in the search range set as shown in FIG. The position of is shown.
- the black area calculation filter set under the conditions shown in FIG. 15 is such that, when black area calculation filter processing is performed on the pixels near the center of the nostril as shown in FIG. However, since the coefficient for subtraction is located outside the nostril, it is possible to clarify the center of the nostril.
- FIG. 17 is a block diagram illustrating a configuration example of the image processing system according to the third embodiment of the present invention.
- reference numeral 1 denotes an image pickup apparatus, and the image pickup apparatus 1 is connected to the image processing apparatus 2 by, for example, a dedicated cable.
- the imaging device 1 includes an MPU 11, a ROM 12, a RAMI 3, an imaging unit 14, an A / D converter 15, a frame memory 16, and a communication interface 17.
- the image processing apparatus 2 includes a CPU 21, an auxiliary recording unit 22, an HD 23, a RAM 24, a frame memory 25, and a communication interface 26, and the computer according to the third embodiment of the present invention.
- the image processing device is configured to read various information from the recording medium 43 on which the computer program 33 and various information such as data are recorded by the auxiliary recording unit 22, record in the HD 23, record in the RAM 24, and execute on the CPU 21. 2 executes various procedures according to Embodiment 3 of the present invention.
- FIG. 18 is a flowchart showing an example of processing of the image processing apparatus 2 used in the image processing system according to Embodiment 3 of the present invention.
- the image processing device 2 the image data obtained by the imaging of the imaging device 1 and received via the communication interface 26 is extracted from the frame memory 25 under the control of the CPU 21 that executes the computer program 33 recorded in the RAM 24 (S301).
- the width of the driver's face that is, the horizontal range of the contour that is the boundary of the region showing the face is detected by, for example, the contour width detection process described in the first embodiment (S302), and
- the positions of both eyes and nose tip are detected by detecting both eyes and nose tip using processing such as pattern matching (S303), and the search range is set based on the horizontal range of the detected contour and the positions of both eyes and nose tip (S304). ).
- the horizontal range of the contour detected in step S302, the positions of both eyes and nose tip detected in step S303, and the search range set in step S304 are recorded in HD23 or RAM24.
- the search range set in step S304 is, for example, a position where the upper end in the vertical direction is lower than the average y coordinate indicating the vertical position of both eyes by a distance of 1Z16 of the horizontal width of the contour,
- the lower end is a position that is 3/8 of the contour width below the average y coordinate of both eyes, and the horizontal left end is the horizontal position of the nose tip from the X coordinate.
- the left position by the distance of, and the right end is the width of the outline from the X coordinate of the nose tip It is set to an area based on the right position by 1 / 8th of the distance.
- the image processing apparatus 2 uses a numerical value based on a luminance difference from a pixel adjacent in the horizontal direction and a horizontal value for all pixels within the search range set in step S304.
- a horizontal score is derived (S30 6).
- the image processing apparatus 2 indicates the numerical value based on the luminance difference from the pixels adjacent in the vertical direction and the low luminance of the pixels adjacent in the vertical direction for all the pixels in the search range under the control of the CPU 21.
- a vertical pixel score that is an index multiplied by a numerical value is derived (S307), and the derived vertical pixel score is accumulated in the horizontal direction to derive a vertical direction score that is an index indicating a change in the accumulated value in the vertical direction. (S308). Then, under the control of the CPU 21, the image processing apparatus 2 uses the horizontal range from the maximum value to the minimum value in the horizontal direction derived in step S306 and the maximum value in the vertical direction score derived in step S307. An area based on the vertical range up to the minimum value is detected as a detection target (S309), and the detection result is recorded in the HD 23 or RAM 24.
- FIG. 19 is an explanatory diagram conceptually showing an example of setting a search range of the image processing system in the third embodiment of the present invention.
- the outer frame force indicated by the solid line S is the entire image indicated by the image data extracted at step S301, and the vertical line (y-axis direction) of the image indicated by the alternate long and short dash line is indicated by step S302.
- This is the range of the area detected at, that is, the width of the contour of the driver's face, and the position of both eyes and nose apex detected at the point force S shown at X in step S303.
- the area indicated by the dotted line is the search range set in step S304.
- Equation 3 The horizontal pixel score derived in step S305 is expressed by the following Equation 3.
- FIG. 20 is an explanatory diagram showing an example of coefficients used for the horizontal edge filter processing of the image processing system according to the third embodiment of the present invention.
- ⁇ (X, y) represents the result of horizontal edge filter processing performed using the coefficients shown in FIG.
- FIG. 20 shows a coefficient multiplied by the luminance of nine pixels as a 3 ⁇ 3 matrix. Coefficients corresponding to the luminance of one central pixel and the luminance of eight adjacent pixels are shown. The result of multiplication and the sum of the results is calculated as the horizontal edge filter processing result of the pixel at the center position.
- the luminance of the pixel adjacent to the left is multiplied by "1”
- the luminance of the pixel adjacent to the right is The resulting index is obtained by adding the numbers multiplied by “1”.
- a numerical value based on a luminance difference from a pixel adjacent in the horizontal direction is obtained by the horizontal edge filter processing.
- the formula that multiplies the result of the horizontal edge filter processing is obtained by subtracting the brightness of pixels adjacent in the horizontal direction from "255", which indicates the highest gradation value obtained by classifying the brightness of the pixels into 256 levels. It is a numerical value indicating the low brightness.
- the horizontal pixel score is an index based on a numerical value based on a luminance difference from a pixel adjacent in the horizontal direction and a numerical value indicating low brightness of the pixel adjacent in the horizontal direction. Note that the pixels whose luminance is used in the formula to multiply the result of the horizontal edge filter processing differ depending on whether the result of the horizontal edge filter processing is positive or negative.
- the horizontal score derived in step S306 is obtained by subtracting the horizontal pixel score.
- This is an index indicating the relationship between the accumulated value and the X coordinate, which is the horizontal position derived by integrating in the straight direction. That is, the horizontal score indicates a change in the horizontal direction of the horizontal pixel score.
- the horizontal direction score can be shown so that the value of the portion where the luminance of the pixel is greatly decreased in the horizontal direction is increased and the value of the portion where the luminance of the pixel is greatly increased is decreased.
- Equation 4 The vertical pixel score derived in step S307 is expressed by Equation 4 below.
- V (x, y) Vertical edge filter processing result of pixel at coordinate 0,
- FIG. 21 is an explanatory diagram showing an example of coefficients used for the vertical direction edge filter processing of the image processing system according to the third embodiment of the present invention.
- V (x, y) represents the result of the vertical edge filtering performed using the coefficients shown in FIG.
- FIG. 21 shows a coefficient by which the luminance of nine pixels is multiplied as a 3 ⁇ 3 matrix. The coefficients corresponding to the luminance of the central pixel and the luminance of eight adjacent pixels are shown. The result of multiplication and the total of the results is calculated as the vertical edge filter processing result of the pixel at the center position.
- the brightness of the upper adjacent pixel is multiplied by "1", and the brightness of the lower adjacent pixel is set to "1 1".
- the resulting index is obtained by adding the numbers multiplied by ".” In other words, a numerical value based on the luminance difference between adjacent pixels in the vertical direction is obtained by the vertical edge filter processing.
- the formula to multiply the result of the vertical edge filter processing subtracts the brightness of the adjacent pixels in the vertical direction from "255", which indicates the highest gradation value obtained by classifying the brightness of the pixels into 256 levels. It is a numerical value indicating the low brightness.
- the vertical pixel score is This is an index based on a numerical value based on a luminance difference from a pixel adjacent in the vertical direction and a numerical value indicating a low luminance of a pixel adjacent in the vertical direction. Note that the pixel whose luminance is used in the formula to multiply the result of the vertical edge filter processing differs depending on whether the result of the vertical edge filter processing is positive or negative.
- the vertical score derived in step S308 is an index indicating the relationship between the integrated value and the y coordinate that is the vertical position derived by integrating the vertical pixel score in the horizontal direction. It is. That is, the vertical score indicates a change in the vertical direction of the vertical pixel score. Specifically, the vertical score can indicate that the value of the portion where the luminance of the pixel is greatly reduced in the vertical direction is large and the value of the portion where the luminance of the pixel is greatly increased is small.
- step S309 the maximum horizontal score is the left end in the horizontal direction, the minimum value is the right end in the horizontal direction, the maximum value in the vertical score is the top end in the vertical direction, and the minimum value is the bottom end in the vertical direction. Is detected.
- FIG. 22 is an explanatory diagram showing the result of detection by the image processing system in the third embodiment of the present invention.
- FIG. 22 shows the search range set in step S304.
- a rectangular area indicated by a diagonal line surrounded by the left end L, the right end upper end U, and the lower end D is a downward area around the detected nostril. It is.
- the various conditions including the numerical values shown in the third embodiment are merely examples, and the system configuration, purpose, etc. can be appropriately set according to the situation.
- the vertical score is derived after the horizontal score is derived.
- the horizontal score may be derived after the vertical score is derived.
- the left and right edges in the horizontal direction may be the width of the face outline, and only the vertical score may be derived.
- FIG. 23 is a block diagram illustrating a configuration example of the image processing system according to the fourth embodiment of the present invention.
- reference numeral 1 denotes an imaging device, and the imaging device 1 is connected to the image processing device 2 by, for example, a dedicated cable.
- the imaging device 1 includes an MPU 11, a ROM 12, a RAMI 3, an imaging unit 14, an A / D converter 15, a frame memory 16, and a communication interface 17.
- the imaging device 1 includes an MPU 11, a ROM 12, a RAMI 3, an imaging unit 14, an A / D converter 15, a frame memory 16, and a communication interface 17.
- the image processing apparatus 2 includes a CPU 21, an auxiliary recording unit 22, an HD 23, a RAM 24, a frame memory 25, and a communication interface 26, and the computer program 34, data, and the like according to the fourth embodiment of the present invention are provided.
- the image processing apparatus 2 implements the present invention by reading various information from the recording medium 44 on which various information has been recorded by the auxiliary recording unit 22, recording it on the HD 23, recording it on the RAM 24, and executing it on the CPU 21.
- Various procedures related to Form 4 are executed.
- a plurality of detection methods including the detection method described in the first to third embodiments of the present invention are recorded.
- Embodiment 1 The detailed description of each device is the same as in Embodiment 1, and therefore, Embodiment 1 is referred to and the description thereof is omitted.
- Embodiment 4 of the present invention detects a region including a nostril of a driver's face from an image obtained by imaging the driver's face with an imaging device 1 such as an in-vehicle power camera mounted on a vehicle, for example. set to target.
- FIG. 24 is a flowchart showing an example of processing of the image processing apparatus 2 used in the image processing system according to Embodiment 4 of the present invention.
- the image processing device 2 extracts from the frame memory 25 the image data obtained through the imaging of the imaging device 1 and received via the communication interface 26 under the control of the CPU 21 that executes the computer program 34 recorded in the RAM 24.
- the average value of the luminance of the pixels included in the extracted image data is calculated (S402), and the calculated average value is compared with a threshold value set in advance for the average value of the luminance. (S403). Further, in the image processing apparatus 2, the variance value of the pixel luminance is calculated under the control of the CPU 21 (S404), and the calculated variance value is compared with a threshold value set in advance for the luminance variance value. (S 405). In the image processing apparatus 2, the priority of the plurality of detection methods recorded on the HD 23 is controlled based on the comparison result between the average value of the luminance and the threshold value and the comparison result of the luminance dispersion value and the threshold value under the control of the CPU 21. The ranking is determined (S406).
- processing for determining the priority of detection methods is performed based on the average value and variance value of luminance.
- the priority order determined is the necessity and the order of execution of the plurality of detection methods. According to the determined contents, the above-described embodiment of the present invention is performed.
- the detection method described in the states 1 to 3 and other detection methods are executed.
- the irradiation status is judged from the average value and the variance value, and the optimum detection method is Make a choice. Specifically, if there is a partial change in which only the left half of the face is exposed to sunlight, the average value of the luminance is below the threshold value, and the variance value of the luminance is above the threshold value. It gives priority to detection methods that can be determined to have occurred and are less susceptible to bias change.
- the nostril area is detected by the detection method shown in the first embodiment, and the nostril is detected by the detection method shown in the second embodiment using the detection result.
- the range to be subjected to image processing is limited, so that the processing speed is improved and the detection accuracy is improved.
- the luminance saturation is likely to occur, so the reliability of the process using the lateral edge filter of the first embodiment is lowered. Therefore, priority is given to detection of the nostril by the detection method shown in the second embodiment.
- one threshold value of the average value of luminance and the threshold value of the variance value is set.
- the present invention is not limited to this, and a plurality of threshold values are set, and detection methods according to various situations are set. It is also possible to determine the priority order of the image, and it is also possible to determine the processing conditions such as various setting values required for image processing for detection based on the average value and the variance value. ,.
- the processing for the image data represented by the planar rectangular coordinate system has been described.
- the present invention is not limited to this.
- image data of various coordinate systems such as applying to image data shown in a coordinate system where the first direction and the second direction intersect at an angle of 60 degrees.
- the vehicle driver is detected.
- the present invention is not limited to this, and various people, and other organisms or inanimate objects can be detected.
- the target form may be sufficient.
- Embodiments 1 to 4 the form in which the detection target is detected from the image generated by the imaging of the imaging device using the in-vehicle camera is shown, but the present invention is not limited to this. Images generated in various ways by various devices can be recorded in HD, and can be applied to various image processing to detect a specific detection target from the recorded images.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Closed-Circuit Television Systems (AREA)
- Studio Devices (AREA)
Abstract
Description
Claims
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2005800478099A CN101116106B (zh) | 2005-02-23 | 2005-02-23 | 图像处理方法、图像处理装置以及图像处理系统 |
| PCT/JP2005/002928 WO2006090449A1 (ja) | 2005-02-23 | 2005-02-23 | 画像処理方法、画像処理装置、画像処理システム及びコンピュータプログラム |
| JP2007504583A JP4364275B2 (ja) | 2005-02-23 | 2005-02-23 | 画像処理方法、画像処理装置及びコンピュータプログラム |
| US11/844,097 US8457351B2 (en) | 2005-02-23 | 2007-08-23 | Image object detection using separate ranges from both image detections |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/JP2005/002928 WO2006090449A1 (ja) | 2005-02-23 | 2005-02-23 | 画像処理方法、画像処理装置、画像処理システム及びコンピュータプログラム |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| US11/844,097 Continuation US8457351B2 (en) | 2005-02-23 | 2007-08-23 | Image object detection using separate ranges from both image detections |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2006090449A1 true WO2006090449A1 (ja) | 2006-08-31 |
Family
ID=36927100
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/JP2005/002928 Ceased WO2006090449A1 (ja) | 2005-02-23 | 2005-02-23 | 画像処理方法、画像処理装置、画像処理システム及びコンピュータプログラム |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US8457351B2 (ja) |
| JP (1) | JP4364275B2 (ja) |
| CN (1) | CN101116106B (ja) |
| WO (1) | WO2006090449A1 (ja) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008197762A (ja) * | 2007-02-09 | 2008-08-28 | Fujifilm Corp | 撮影装置および方法並びにプログラム |
| JP2009199417A (ja) * | 2008-02-22 | 2009-09-03 | Denso Corp | 顔追跡装置及び顔追跡方法 |
| EP2015568A3 (en) * | 2007-06-14 | 2012-02-22 | FUJIFILM Corporation | Digital image pickup apparatus |
| KR101788070B1 (ko) * | 2012-12-14 | 2017-10-20 | 동국대학교 산학협력단 | 코 영역 검출 방법 |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4518157B2 (ja) * | 2008-01-31 | 2010-08-04 | カシオ計算機株式会社 | 撮像装置及びそのプログラム |
| CN101978392B (zh) * | 2008-03-26 | 2013-01-16 | 本田技研工业株式会社 | 车辆用图像处理装置 |
| CN102339466B (zh) * | 2010-07-15 | 2016-04-13 | 韩华泰科株式会社 | 用于检测具有特定形状的区域的方法和相机 |
| JP5367037B2 (ja) * | 2011-09-26 | 2013-12-11 | 本田技研工業株式会社 | 顔向き検出装置 |
| JP6222900B2 (ja) * | 2012-07-09 | 2017-11-01 | キヤノン株式会社 | 画像処理装置、画像処理方法およびプログラム |
| JP6169366B2 (ja) * | 2013-02-08 | 2017-07-26 | 株式会社メガチップス | 物体検出装置、プログラムおよび集積回路 |
| JP6337949B1 (ja) * | 2016-12-15 | 2018-06-06 | オムロン株式会社 | スジ状領域検出装置およびスジ状領域検出方法 |
| CN108304764B (zh) * | 2017-04-24 | 2021-12-24 | 中国民用航空局民用航空医学中心 | 模拟飞行驾驶过程中疲劳状态检测装置及检测方法 |
| CN110580676A (zh) * | 2018-06-07 | 2019-12-17 | 富泰华工业(深圳)有限公司 | 人脸漫画形象制作方法、电子装置和存储介质 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH04174309A (ja) * | 1990-11-07 | 1992-06-22 | Nissan Motor Co Ltd | 運転車の眼位置検出装置及び状態検出装置 |
| JPH08300978A (ja) * | 1995-05-10 | 1996-11-19 | Mitsubishi Electric Corp | 顔画像の処理装置 |
| JP2001067459A (ja) * | 1999-08-26 | 2001-03-16 | Toshiba Corp | 顔画像処理方法および顔画像処理装置 |
| JP2001331795A (ja) * | 2000-05-18 | 2001-11-30 | Victor Co Of Japan Ltd | 画像正規化装置及び方法 |
| JP2004234367A (ja) * | 2003-01-30 | 2004-08-19 | Fujitsu Ltd | 顔の向き検出装置、顔の向き検出方法及びコンピュータプログラム |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2674354A1 (fr) * | 1991-03-22 | 1992-09-25 | Thomson Csf | Procede d'analyse de sequences d'images routieres, dispositif pour sa mise en óoeuvre et son application a la detection d'obstacles. |
| US6061400A (en) * | 1997-11-20 | 2000-05-09 | Hitachi America Ltd. | Methods and apparatus for detecting scene conditions likely to cause prediction errors in reduced resolution video decoders and for using the detected information |
| CN1131699C (zh) * | 1999-06-09 | 2003-12-24 | 现代自动车株式会社 | 瞌睡驾驶警告系统的驾驶员眼睛检测方法 |
| JP3895238B2 (ja) * | 2002-08-28 | 2007-03-22 | 株式会社東芝 | 障害物検出装置及びその方法 |
| JP4162503B2 (ja) | 2003-01-31 | 2008-10-08 | 富士通株式会社 | 眼の状態判定装置、眼の状態判定方法及びコンピュータプログラム |
-
2005
- 2005-02-23 CN CN2005800478099A patent/CN101116106B/zh not_active Expired - Lifetime
- 2005-02-23 JP JP2007504583A patent/JP4364275B2/ja not_active Expired - Lifetime
- 2005-02-23 WO PCT/JP2005/002928 patent/WO2006090449A1/ja not_active Ceased
-
2007
- 2007-08-23 US US11/844,097 patent/US8457351B2/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH04174309A (ja) * | 1990-11-07 | 1992-06-22 | Nissan Motor Co Ltd | 運転車の眼位置検出装置及び状態検出装置 |
| JPH08300978A (ja) * | 1995-05-10 | 1996-11-19 | Mitsubishi Electric Corp | 顔画像の処理装置 |
| JP2001067459A (ja) * | 1999-08-26 | 2001-03-16 | Toshiba Corp | 顔画像処理方法および顔画像処理装置 |
| JP2001331795A (ja) * | 2000-05-18 | 2001-11-30 | Victor Co Of Japan Ltd | 画像正規化装置及び方法 |
| JP2004234367A (ja) * | 2003-01-30 | 2004-08-19 | Fujitsu Ltd | 顔の向き検出装置、顔の向き検出方法及びコンピュータプログラム |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008197762A (ja) * | 2007-02-09 | 2008-08-28 | Fujifilm Corp | 撮影装置および方法並びにプログラム |
| EP2015568A3 (en) * | 2007-06-14 | 2012-02-22 | FUJIFILM Corporation | Digital image pickup apparatus |
| CN102438105A (zh) * | 2007-06-14 | 2012-05-02 | 富士胶片株式会社 | 摄影设备 |
| CN102438099A (zh) * | 2007-06-14 | 2012-05-02 | 富士胶片株式会社 | 摄影设备 |
| US9065998B2 (en) | 2007-06-14 | 2015-06-23 | Fujifilm Corporation | Photographing apparatus provided with an object detection function |
| US9131138B2 (en) | 2007-06-14 | 2015-09-08 | Fujifilm Corporation | Photographing apparatus |
| JP2009199417A (ja) * | 2008-02-22 | 2009-09-03 | Denso Corp | 顔追跡装置及び顔追跡方法 |
| KR101788070B1 (ko) * | 2012-12-14 | 2017-10-20 | 동국대학교 산학협력단 | 코 영역 검출 방법 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101116106A (zh) | 2008-01-30 |
| CN101116106B (zh) | 2011-11-09 |
| US8457351B2 (en) | 2013-06-04 |
| JPWO2006090449A1 (ja) | 2008-07-17 |
| JP4364275B2 (ja) | 2009-11-11 |
| US20070291999A1 (en) | 2007-12-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8457351B2 (en) | Image object detection using separate ranges from both image detections | |
| EP2031571B1 (en) | Vehicle type determination device, program, and method | |
| JP4781467B2 (ja) | 画像処理方法、画像処理装置、画像処理システム及びコンピュータプログラム | |
| EP2698762B1 (en) | Eyelid-detection device, eyelid-detection method, and program | |
| JP2008286725A (ja) | 人物検出装置および方法 | |
| JP4592744B2 (ja) | 画像処理方法、画像処理システム、画像処理装置及びコンピュータプログラム | |
| CN102368946B (zh) | 瞳孔遮蔽状态检测装置和具有它的车载照相机 | |
| JP5034623B2 (ja) | 画像処理方法、画像処理装置、画像処理システム及びコンピュータプログラム | |
| CN101107625B (zh) | 图像处理方法、图像处理系统、摄像装置及图像处理装置 | |
| CN101320477B (zh) | 一种人体跟踪方法及其设备 | |
| JP2011165170A (ja) | 対象物検出装置及びプログラム | |
| JP5370375B2 (ja) | 画像認識装置 | |
| US20070292000A1 (en) | Image processing method, image processing system, image processing device, and computer program | |
| JP6002564B2 (ja) | 鋼材における炭化物の球状化率測定装置および球状化率測定方法、並びにプログラム | |
| JP2009193130A (ja) | 車両周辺監視装置、車両、車両周辺監視用プログラム、車両周辺監視方法 | |
| JP7472772B2 (ja) | 画像処理装置、過積載検知システム、画像処理方法および画像処理プログラム | |
| CN101714210B (zh) | 图像处理方法、图像处理装置以及图像处理系统 | |
| JPH09128548A (ja) | 車両検出方法 | |
| WO2011155152A1 (ja) | 車両の周辺監視装置 | |
| KR20220145080A (ko) | 차량용 듀얼 카메라 장착 위치 결정 장치 및 방법 | |
| JP2021077249A (ja) | 判定装置 | |
| JP2021152826A (ja) | 情報処理装置、被写体分類方法、及び被写体分類プログラム | |
| CN115311639B (zh) | 面向自动驾驶的车道线多聚类虚拟生成方法 | |
| JPH1091798A (ja) | 駐車車両検出方法およびその装置 | |
| JPH05183784A (ja) | 車載用道路画像解析装置およびその装置が用いられた車載用撮像制御装置 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
| WWE | Wipo information: entry into national phase |
Ref document number: 2007504583 Country of ref document: JP |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 200580047809.9 Country of ref document: CN |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 11844097 Country of ref document: US |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| WWP | Wipo information: published in national office |
Ref document number: 11844097 Country of ref document: US |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 05719436 Country of ref document: EP Kind code of ref document: A1 |
|
| WWW | Wipo information: withdrawn in national office |
Ref document number: 5719436 Country of ref document: EP |
