WO2019061293A1 - 对象检测方法、对象检测终端及计算机可读介质 - Google Patents
对象检测方法、对象检测终端及计算机可读介质 Download PDFInfo
- Publication number
- WO2019061293A1 WO2019061293A1 PCT/CN2017/104383 CN2017104383W WO2019061293A1 WO 2019061293 A1 WO2019061293 A1 WO 2019061293A1 CN 2017104383 W CN2017104383 W CN 2017104383W WO 2019061293 A1 WO2019061293 A1 WO 2019061293A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- target image
- type
- image
- area
- target
- Prior art date
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 156
- 230000002159 abnormal effect Effects 0.000 claims abstract description 65
- 238000000034 method Methods 0.000 claims abstract description 54
- 238000012545 processing Methods 0.000 claims description 35
- 230000001815 facial effect Effects 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 12
- 238000000605 extraction Methods 0.000 claims description 10
- 238000005192 partition Methods 0.000 claims description 5
- 230000036760 body temperature Effects 0.000 description 35
- 238000001228 spectrum Methods 0.000 description 19
- 238000010586 diagram Methods 0.000 description 10
- 208000035473 Communicable disease Diseases 0.000 description 4
- 241001465754 Metazoa Species 0.000 description 4
- 238000007689 inspection Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 206010037660 Pyrexia Diseases 0.000 description 3
- 230000009977 dual effect Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 208000015181 infectious disease Diseases 0.000 description 3
- 238000009434 installation Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 208000002979 Influenza in Birds Diseases 0.000 description 1
- 241000700605 Viruses Species 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000002547 anomalous effect Effects 0.000 description 1
- 206010064097 avian influenza Diseases 0.000 description 1
- 210000000746 body region Anatomy 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 206010022000 influenza Diseases 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J5/00—Radiation pyrometry, e.g. infrared or optical thermometry
- G01J5/0022—Radiation pyrometry, e.g. infrared or optical thermometry for sensing the radiation of moving bodies
- G01J5/0025—Living bodies
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
Definitions
- the present invention relates to the field of image processing technologies, and in particular, to an object detecting method, an object detecting terminal, and a computer readable medium.
- body temperature can reflect the physical health of the human body to a certain extent, especially in the period of serious spread of infectious diseases such as influenza, the body temperature abnormality is detected by body temperature detection in various occasions and fields (such as airport, customs security, abnormal epidemic situation). Monitoring, etc.) are very important.
- Embodiments of the present invention provide an object detecting method, which can quickly detect an object with a temperature anomaly.
- an embodiment of the present invention provides an object detection method, including:
- the first type of target image is an image for representing a temperature distribution
- the second type of target image is an image for performing image object recognition
- the target image object is an object located in the second type of target image and determined to be in a temperature abnormal state according to a temperature distribution represented by the first type of target image.
- an embodiment of the present invention provides an object detection terminal, including a processor and a memory:
- the memory is configured to store a computer program, the computer program comprising program instructions
- the processor is configured to invoke the program instruction to execute:
- the first type of target image is an image for representing a temperature distribution
- the second type of target image is an image for performing image object recognition
- the target image object is an object in a temperature abnormal state determined in the second type of target image and determined according to a temperature distribution represented by the first type of target image.
- an embodiment of the present invention provides a computer readable storage medium, where the computer storage medium stores a computer program, where the computer program includes program instructions, and the program instructions, when executed by a processor, cause the processing The method of the first aspect described above is performed.
- a second type of target image for performing image object recognition such as a first type of target image for representing a temperature distribution and an RGB image, such as a thermal infrared image, and the like, are obtained by photographing, and according to the first type of target image pair.
- the second type of target image is subjected to analysis processing, and the target image object in the temperature abnormal state is determined from the second type of target image, and the object with abnormal temperature can be quickly detected, and the intelligence of the object detection can be improved.
- the efficiency of object detection can be greatly improved.
- FIG. 1 is a schematic flow chart of an object detecting method according to an embodiment of the present invention.
- FIG. 2 is a schematic flow chart of another object detecting method according to an embodiment of the present invention.
- FIG. 3 is a schematic diagram of dividing an object area from a second type of target image according to an embodiment of the present invention.
- FIG. 4 is a schematic diagram of dividing a temperature detection area corresponding to an object area shown in FIG. 3 from a first type of target image according to an embodiment of the present invention
- FIG. 5 is a schematic diagram of dividing a temperature anomaly region from a first type of target image according to an embodiment of the present invention
- FIG. 6 is a schematic diagram of dividing an object detection area corresponding to a temperature abnormal area shown in FIG. 5 from a second type of target image according to an embodiment of the present invention
- FIG. 7 is a view showing a positional relationship between an object detecting terminal and a photographic subject according to an embodiment of the present invention.
- FIG. 8 is a diagram showing a positional relationship between another object detecting terminal and a photographic subject according to an embodiment of the present invention.
- FIG. 9 is a diagram showing a position and a proportional relationship between a face image area and a target object area according to an embodiment of the present invention.
- FIG. 10 is a schematic structural diagram of an object detecting terminal according to an embodiment of the present invention.
- body temperature can reflect the health of the human body to a certain extent, body temperature is one of the most important indicators in security screening.
- body temperature is one of the most important indicators in security screening.
- only some individuals can be routinely examined for body temperature. This often fails to detect individuals who may carry abnormal viruses (ie, individuals with abnormal body temperature), and there are still missed tests. risk. If a large number of individuals are concentrated in a certain location, the temperature detection by the staff one by one will cause the workload of body temperature detection to be huge, time-consuming and labor-intensive, and the detection efficiency is low.
- an embodiment of the present invention provides an object detection method, which may be specifically applied to an object detection terminal.
- the object detecting terminal first acquires a thermal infrared image and an RGB (red red, green green, blue blue) image.
- the object detection terminal may include a thermal infrared camera (also referred to as a thermal infrared camera) and a visible light camera (also referred to as a visible light camera).
- the thermal infrared camera and the visible light camera are respectively used for capturing a thermal infrared image and an RGB image.
- the object detecting terminal can acquire the thermal infrared image and the RGB image by controlling the thermal infrared camera and the visible light camera to respectively perform a photographing operation.
- the object detecting terminal may control the thermal infrared camera and the visible light camera to perform a shooting operation simultaneously in the same posture, or the object detecting terminal may control the thermal infrared camera.
- the visible light camera simultaneously performs a shooting operation on the same scene.
- the object detecting terminal can be installed in a place where an individual who needs to perform body temperature detection such as a ticket gate or a boarding gate must pass.
- the object detection terminal can include a dual spectrum camera (also referred to as a dual spectrum camera).
- the dual-spectrum camera can be used to capture thermal infrared images and RGB images.
- the dual-spectrum camera may have a first shooting mode and a second shooting mode. In the first shooting mode, the dual-spectrum camera can be used to capture a thermal infrared image. In the second shooting mode, the dual-spectrum camera can be used to capture RGB images.
- the object detecting terminal can acquire the thermal infrared image and the RGB image by controlling the dual-spectrum camera to perform a photographing operation in the first photographing mode and the second photographing mode, respectively.
- the object detecting terminal may control the dual-spectrum camera to perform a photographing operation in the first photographing mode at a first moment to acquire the thermal infrared image, and control the dual-spectrum camera.
- the RGB image is acquired by performing a shooting operation in the second shooting mode at a second time, wherein a time interval between the first time and the second time is a preset time interval. It can be understood that in order for the thermal infrared image and the RGB image to be images of two different properties taken on the same scene, the preset time interval must be sufficiently small. And, at the first time and the second time, the shooting attitude of the dual-spectrum camera is the same.
- the object detecting terminal may be disposed at a ticket gate, a boarding gate, and the like where an individual who needs to perform body temperature detection must pass.
- the object detecting terminal may acquire the thermal infrared image and the RGB image from a photographing terminal. That is, the thermal infrared image and the RGB image are captured by the photographing terminal and transmitted to the object detecting terminal by wire or wirelessly.
- the photographing terminal may respectively obtain the thermal infrared image and the RGB image by a thermal infrared camera and a visible light camera disposed therein.
- the photographing terminal may further capture the thermal infrared camera and the RGB image by using a dual-spectrum camera disposed therein.
- the photographing terminal may be disposed at a ticket gate, a boarding gate, and the like where an individual who needs to perform body temperature detection must pass, and the object detecting terminal may be disposed in the monitoring room.
- the object detecting terminal may acquire the hot infrared image and the RGB image sent by the plurality of shooting terminals.
- the thermal infrared image and the RGB image are images of two different attributes.
- the thermal infrared image is an image for indicating a temperature distribution
- a color image captured by a terminal having a photographing function such as a mobile phone, a tablet computer, or a digital camera is an RGB image.
- the object detecting terminal may align and fuse the captured thermal infrared image and the RGB image, so that the temperature value of each position (such as each pixel point) in the RGB image may be obtained. .
- the object detecting terminal may determine an image region in the RGB image whose temperature is within an abnormal body temperature range as an image region of interest according to a temperature distribution represented by the thermal infrared image. Then, the object detecting terminal may perform human body detecting processing on the image region of interest using a human body detecting algorithm to determine a human body image region in the image region of interest.
- the normal body temperature range of the human body is generally about 36 to 37.5 °C.
- the body temperature of the human body exceeds 37.5 ° C, it is called fever (or fever).
- the body temperature does not exceed 42 ° C when the body heats up.
- 30 to 36 ° C and 37.5 to 42 ° C can be preset as the abnormal body temperature range in the object detecting terminal.
- °C is an abnormal body temperature range, that is, a human body whose body temperature is 37.5 to 42 ° C is determined to be a human body that may have this infectious disease.
- the object detecting terminal may be from the human body according to the determined ratio of the human body image area, the face image area in the human body image area, and the face image area occupying the human body image area.
- a face image area is determined in the image area.
- the human body is generally in a standing posture, in which case the face image area is at the top of the body image area.
- the ratio of the face image area to the body image area may be a constant (eg, 20%) preset in the object detecting terminal.
- the object detecting terminal may perform face recognition processing on the face image region, and extract face feature data in the face image region. Then, the object detecting terminal may further match the extracted facial feature data with facial feature data in a database (such as a public security system) to determine identity information of the human body whose body temperature is abnormal.
- a database such as a public security system
- the object detecting terminal can determine position information of the human body with abnormal body temperature relative to the camera. Positioning information (such as coordinate values) of an object detecting terminal or a photographing terminal that photographs the RGB image and photographing of the camera that photographs the RGB image The posture is known.
- the object detecting terminal may further determine the body temperature according to the position of the human body image area in the RGB image, the mounting position information, and the shooting posture and position information thereof. Absolute position information of an abnormal human body.
- the object detecting method of the embodiment of the present invention can conveniently determine an individual whose body temperature is abnormal from a crowd, and can also determine according to a shooting posture (or installation position information and a shooting attitude) of a photographing device (such as the object detecting terminal or the photographing terminal).
- the position information of the abnormal body temperature is obtained, thereby finding the individual whose body temperature is abnormal, and using the face recognition technology, the identity information of the individual with abnormal body temperature can also be obtained from the public security system.
- the object detection method of the embodiment of the present invention can be applied to security inspection work.
- a specific occasion such as an airport, a gate, a train station, etc.
- the object detection terminal or the photographing terminal is installed in a security check-in intersection or a ticket gate, a boarding gate, and the like where an individual who needs to perform body temperature detection must pass
- the object detection The terminal can automatically perform the detecting work, determine the individual whose body temperature is abnormal, and provide the position information and the identity information of the individual whose body temperature is abnormal, and the security inspection staff only needs to find the abnormal body temperature according to the position information, and The identity information is confirmed, the basic security inspection work is completed efficiently, and the detection efficiency is greatly improved.
- the object detection method of the embodiment of the invention can also be applied to abnormal epidemic monitoring.
- body temperature detection is particularly important, especially in densely populated public places such as squares and commercial streets. Individuals with abnormal body temperature need to be detected in time and isolated in time. To prevent further spread of the epidemic.
- the object detecting terminal or the photographing terminal can be installed at the exit, entrance, and the like of these public places.
- the object detecting method of the embodiment of the invention can also be applied to the abnormal temperature area monitoring of the factory building.
- the object detecting terminal or the photographing terminal can be installed at a position such as four corners in the factory building.
- the preset normal temperature range or abnormal temperature range
- an image of the area where the temperature is abnormal may be Reported to the system to provide a reference for security checks, with a focus on the area.
- the method of the embodiment of the present invention may be applied to other occasions or fields, which is not limited by the embodiment of the present invention.
- the object detecting terminal or the photographing terminal can be installed on a drone or a patrol vehicle, thereby improving the mobility and flexibility of the detecting work.
- FIG. 1 is a schematic flowchart diagram of an object detecting method according to an embodiment of the present invention. As shown in FIG. 1, the object detecting method may include:
- S101 The first type target image and the second type target image are captured.
- the first type of target image is an image for representing a temperature distribution.
- the first type of target image is a thermal infrared image captured by a thermal infrared camera.
- the second type of target image is an image for performing image object recognition.
- the second type of target image is an RGB image captured by a visible light camera, or any image including RGB information.
- the rear camera and the front camera among the terminals having a photographing function such as a mobile phone, a tablet computer, a digital camera, and a SLR camera are visible light cameras, and the color images captured are RGB images.
- the second type of target image may also be an HSV image, a YUV image, or the like, which is captured by a visible light camera, which is not limited in the embodiment of the present invention.
- the object detection method in the embodiment of the present invention may be specifically applied to the object detection terminal.
- the object detecting terminal may include a hot red hot camera and a visible light camera. Therefore, the object detecting terminal performing the capturing to obtain the first type of target image and the second type of target image may specifically include: controlling the hot infrared camera to perform a shooting operation to obtain a first type of target image; and controlling the visible light camera to perform shooting. The operation obtains the second type of target image.
- the object detecting terminal may control the thermal infrared camera and the visible light camera to be the same The gesture is simultaneously performed.
- S102 Perform analysis processing on the second type of target image according to the first type of target image, and determine a target image object from the second type of target image.
- the target image object is an object in a temperature abnormal state determined in the second type of target image and determined according to a temperature distribution represented by the first type of target image.
- the object detecting terminal performs analysis processing on the second type of target image according to the first type of target image, and determines from the second type of target image.
- the target image object may specifically include: performing object detection processing on the second type of target image to obtain at least one object region; determining, according to the temperature distribution represented by the first type of target image, from the at least one object region a target object area; a target image object is identified from the target object area.
- the object detection process may be, for example, a human body detection process.
- the object area may specifically be a human body image area.
- the target image object may be specifically a human body.
- the object detection process may also be an animal detection process, or a human body and an animal detection process, etc., which are not limited in the embodiment of the present invention.
- the object area may be a rectangular area.
- the object detecting terminal detects that the second type target image includes three human body image regions after performing human body detecting processing on the second type target image, the object detecting terminal may Three rectangular object areas are determined in the second type of target image.
- the object detecting terminal may be configured according to a region correspondence relationship between the first type of target image and the second type of target image. A temperature detection area corresponding to the target area in the first type of target image is determined. Further, the object detecting terminal may, according to the temperature distribution represented by the first type of target image, an object corresponding to a temperature detecting area of a pixel point in which the temperature is within an abnormal temperature range in the second type of target image. The area is the target object area.
- the abnormal temperature range may be preset in the object detecting terminal.
- the object detecting terminal performs analysis processing on the second type of target image according to the first type of target image, and determines a target from the second type of target image.
- the image object may include: determining an object detection area from the second type of target image according to a temperature distribution represented by the first type of target image; performing object detection processing on the object detection area to obtain a target object area ; identifying a target image object from the target object area.
- the object detection area may be a rectangular area.
- the object detecting terminal may determine that there is a temperature of a pixel in the first type of target image whose temperature is within an abnormal temperature range. Anomalous area. Therefore, the object detecting terminal may determine an object corresponding to the temperature abnormal region in the second type target image according to a region correspondence relationship between the first type target image and the second type target image. Detection area.
- the pixel whose temperature is within an abnormal temperature range The ratio of the points to the total pixel points in the abnormal temperature region is at least a preset ratio.
- the object detection process may be, for example, a human body detection process.
- the target object area may specifically be a human body image area.
- the target object area may be a rectangular area.
- the object detecting terminal when the object detecting terminal detects that the object detecting area includes one body image area after performing the human body detecting process on the object detecting area, the object detecting terminal may be from the object detecting area. Determine a target object area.
- a second type of target image for performing image object recognition such as a first type of target image for representing a temperature distribution and an RGB image, such as a thermal infrared image
- the image is analyzed and processed by the second type of target image, and the target image object in the temperature abnormal state is determined from the second type of target image, and the object with abnormal temperature can be detected quickly, and the intelligence of the object detection can be improved. Sex. When the number of objects requiring temperature detection is large, the efficiency of object detection can be greatly improved.
- FIG. 2 is a schematic flowchart diagram of another object detecting method according to an embodiment of the present invention.
- the object detecting method may include:
- S201 The first type of target image and the second type of target image are captured.
- the first type of target image is an image for representing a temperature distribution
- the second type of target image is an image for performing image object recognition.
- the first type of target image is a thermal infrared image captured by a thermal infrared camera
- the second type of target image is an RGB image captured by a visible light camera
- the rear camera and the front camera among the terminals having a photographing function such as a mobile phone, a tablet computer, a digital camera, and a SLR camera are visible light cameras, and the color images captured are RGB images.
- the object detection method in the embodiment of the present invention may be specifically applied to the object detection terminal.
- the object detecting terminal may include a hot red hot camera and a visible light camera. Therefore, the object detecting terminal performing the capturing to obtain the first type of target image and the second type of target image may specifically include: controlling the hot infrared camera to perform a shooting operation to obtain a first type of target image; and controlling the visible light camera to perform shooting. The operation obtains the second type of target image.
- the object detecting terminal may control the thermal infrared camera and the visible light camera to perform a photographing operation simultaneously in the same posture (ie, the thermal infrared camera and the visible light camera have the same shooting posture).
- the first type of target image and the second type of target image are respectively a thermal infrared image and an RGB image captured by a dual spectrum camera.
- the dual-spectrum camera may include a first shooting mode and a second shooting mode. In the first shooting mode, the dual-spectrum camera can be used to capture a thermal infrared image. In the second shooting mode, the dual-spectrum camera can be used to capture RGB images. Thereby, the object detecting terminal can obtain the thermal infrared image and the RGB image by controlling the dual-spectrum camera to perform a photographing operation in the first photographing mode and the second photographing mode, respectively.
- the object detecting terminal may control the dual-spectrum camera to perform a photographing operation in the first photographing mode at a first moment to acquire the thermal infrared image, and control the dual-spectrum camera.
- the RGB image is acquired by performing a shooting operation in the second shooting mode at a second time, wherein a time interval between the first time and the second time is a preset time interval. It can be understood that in order for the thermal infrared image and the RGB image to be images of two different attributes captured on the same scene, the preset time interval must be sufficiently small. And, at the first time and the second time, the shooting attitude of the dual-spectrum camera is the same.
- the second type of target image may also be an HSV image, a YUV image, or the like, which is captured by a visible light camera, which is not limited in the embodiment of the present invention.
- the object detecting terminal may perform a photographing operation on the environment to be detected to obtain a first type of image for representing a temperature distribution and a second type of image for performing image object recognition. And determining the first type image and the second type image as the first type target image and the second type target image, respectively.
- the environment to be detected is an environment in which the object detection terminal is responsible for performing object detection.
- the lobby is the environment to be detected.
- the object detecting terminal performing the capturing to obtain the first type of target image and the second type of target image may specifically include: capturing a first type of image of the environment to be detected, and according to the The first type of image to be detected determines the target detection area in the environment to be detected; and the target detection area is photographed to obtain a first type of target image and a second type of target image.
- the determining, by the object detecting terminal, the target detection area in the to-be-detected environment according to the first type of image of the to-be-detected environment may specifically include: determining, according to a preset temperature range, the to-be-detected environment A type of image is subjected to an analysis process to determine a target image region, and a temperature distribution corresponding to the target image region satisfies a preset partition condition; and a target detection region in the to-be-detected environment is determined according to the target image region.
- the target detection area is a partial area in an environment in which the object detection terminal is responsible for performing object detection.
- the preset temperature range may be specifically an abnormal temperature range such as a preset abnormal body temperature range, and the preset partition condition may be specifically a pixel point in the target image area where the temperature is within the preset temperature range.
- the object detecting terminal may determine that a pixel located in a left side of the first type of image has a pixel whose temperature is within the preset temperature range, and is located at a pixel
- the area on the right side of the first type of image does not have a pixel at a temperature within the preset temperature range (ie, the temperature in each position in the area of the hall that is located on the right side of the object detection terminal is at a normal temperature
- the object detecting terminal may correspond to the area on the left side of the first type of image in the second type of image according to the area correspondence between the first type of image and the second type of image.
- the area is determined as the target image area, and the object detecting terminal can also determine the area located on the left side
- S202 Perform analysis processing on the second type of target image according to the first type of target image, and determine a target image object from the second type of target image.
- the target image object is an object in a temperature abnormal state determined in the second type of target image and determined according to a temperature distribution represented by the first type of target image.
- the object detecting terminal performs analysis processing on the second type of target image according to the first type of target image, and determines a target image from the second type of target image.
- the object may specifically include: performing object detection processing on the second type of target image to obtain at least one object region; determining a target object from the at least one object region according to a temperature distribution represented by the first type of target image An area; a target image object is identified from the target object area.
- the object detection process may be, for example, a human body detection process.
- the object area may specifically be a human body image area.
- the target image object may be specifically a human body.
- the object detection process may also be an animal detection process, or a human body and an animal detection process, etc., which are not limited in the embodiment of the present invention.
- the object area may be a rectangular area.
- the first image object 111 and the second image object 112 are included in the area 10 where the second type of target image is located.
- the first object area 121 and the second object area 122 may be obtained.
- the first object area 121 and the second object area 122 may be specifically rectangular areas.
- the region 20 in which the first type of target image is located is composed of a first region 211 whose temperature is within an abnormal temperature range, a second region 212 whose temperature is within an abnormal temperature range, and The third region 213 whose temperature is not within the abnormal temperature range is composed.
- the object detecting terminal may determine, respectively, the first object area 121 in the area 20 where the first type of target image is located, according to the area correspondence relationship between the first type of target image and the second type of target image. a first temperature detecting area 221 and a second temperature detecting area 222 corresponding to the second object area 122.
- the area 10 where the second type of target image is located has a mutual correspondence with the area 20 where the first type of target image is located.
- the area 10 where the second type of target image is located is the same shape and area as the area 20 where the first type of target image is located, and, in the same coordinate system, where the second type of target image is located.
- a pixel at a certain coordinate position in the region 10 corresponds to a pixel located at the coordinate position in the region 20 in which the first type of target image is located.
- the first temperature detecting region 221 includes the first region 211, that is, the first temperature detecting region 221 includes a pixel point whose temperature is within an abnormal temperature range, and the second temperature detecting region 222 The pixel points in the image are not in the abnormal temperature range, so the object detecting terminal can determine the first object region 121 corresponding to the first temperature detecting region 221 in the region 10 where the second type target image is located as the target.
- the object area since the first temperature detecting region 221 includes the first region 211, that is, the first temperature detecting region 221 includes a pixel point whose temperature is within an abnormal temperature range, and the second temperature detecting region 222 The pixel points in the image are not in the abnormal temperature range, so the object detecting terminal can determine the first object region 121 corresponding to the first temperature detecting region 221 in the region 10 where the second type target image is located as the target.
- the object area is not in the abnormal temperature range.
- the object detecting terminal may perform an image processing operation such as contour extraction on the first object region 121, identify the first image object 111, and determine the first image object 111 as a target image object.
- the object detecting terminal performs the analyzing process on the second type of target image according to the first type of target image, and determines from the second type of target image.
- the target image object may include: determining, according to a temperature distribution represented by the first type of target image, an object detection area from the second type of target image; performing object detection processing on the object detection area to obtain a target An object area; a target image object is identified from the target object area.
- the region 20 in which the target image of the first type is located is composed of a first region 211 whose temperature is within an abnormal temperature range, a second region 212 whose temperature is within an abnormal temperature range, and a third region 213 whose temperature is not within an abnormal temperature range.
- the object detecting terminal may divide the first area 211 and the second area 212 into the first temperature abnormal area 231 and the second temperature abnormal area 232, respectively.
- the first temperature abnormal region 231 and the second temperature abnormal region 232 may be specifically rectangular regions.
- the area of the first area 211 occupies the area of the first temperature abnormal area 231
- the area of the second area 212 occupies the area of the second temperature abnormal area 232. The proportions exceed the preset ratio.
- the first image object 111 and the second image object 112 are included in the area 10 where the second type of target image is located.
- the object detecting terminal may determine, respectively, the first temperature abnormal region 231 and the region in the region 10 where the second type target image is located, according to the region correspondence relationship between the first type of target image and the second type of target image.
- the first object detection area 131 and the second object detection area 132 corresponding to the second temperature abnormal area 232 are described.
- the object detecting terminal may perform object detecting processing on the first object detecting area 131 and the second object detecting area 132, respectively.
- the object detecting terminal can determine that there is an image object in the first object detecting area 131, and there is no image object in the second object detecting area 132, so that the first object detecting area 131 can be used as a target object. region.
- the object detecting terminal may perform an image processing operation such as contour extraction on the first object detecting area 131, identify the first image object 111, and determine the first image object 111 as a target image object. .
- S203 Acquire location description information of the shooting location, and determine a location of the detection object indicated by the target image object according to the location description information.
- the photographing position refers to a position where the camera of the first type target image or the second type target image is captured.
- the shooting location may be specifically a location where the object detecting terminal is located, that is, an installation location of the object detecting terminal.
- the location description information may be relative location description information or absolute location description information.
- the relative position description information refers to a position of the photographing position relative to a certain reference object. Set the information.
- the relative position description information may include a mounting height (ie, a shooting height) of the object detecting terminal.
- the relative position description information may be, for example, that the object detecting terminal is located in Hall 1, and its installation height is 3 m.
- the absolute location description information may be, for example, a coordinate value of the object detection terminal in a three-dimensional coordinate system.
- the location description information may further include shooting attitude information of the object detecting terminal, such as a shooting angle of the object detecting terminal in a vertical direction (first shooting angle) and a shooting angle of the object detecting terminal in a horizontal direction ( Second shooting angle).
- shooting attitude information of the object detecting terminal such as a shooting angle of the object detecting terminal in a vertical direction (first shooting angle) and a shooting angle of the object detecting terminal in a horizontal direction ( Second shooting angle).
- the determining, by the object detecting terminal, the determining, by the location description information, the location of the detection object indicated by the target image object may include: obtaining the shooting object and shooting according to the shooting height and the first shooting angle. a distance between the positions; obtaining a position coordinate of the photographic subject according to a second photographic angle, a distance between the photographic subject and the photographing position, and a position coordinate described by the position description information; The position coordinates determine the position of the detected object.
- the object detecting terminal 31 can determine the position of the photographic subject 32 with respect to the object detecting terminal 31.
- the object detecting terminal 31 can determine the coordinate value of the photographic subject 32 in the three-dimensional Cartesian coordinate system.
- the abscissa value of the object detecting terminal 31 on the x-axis is a
- the object detecting terminal 31 is at z
- the vertical coordinate value of the axis is c
- the photographic subject is the detection object.
- the photographic subject is one of the detection objects.
- the object detecting terminal performing the determining the position of the detection object according to the position coordinates of the photographic object may specifically include: determining the photographic object according to the second type of target image a relative positional relationship between the other detected objects; determining the position coordinates of the other detected objects based on the position coordinates of the photographic subject and the relative positional relationship.
- the object detecting terminal can determine the position coordinates of the other detected objects according to the actual positional relationship and the position coordinates of the photographic subject.
- S204 Extract face feature data of the target image object from the second type of target image, and determine identity information of the detected object according to face feature data matching.
- the object detecting terminal performs the extraction of the facial feature data of the target image object from the second type of target image, and determines the identity information of the detected object according to the matching of the facial feature data.
- the method may further include: determining whether the detection object indicated by the target image object is a human body; if yes, performing the facial feature data of extracting the target image object from the second type of target image. That is to say, the object detecting terminal needs to perform face recognition processing on the second type of target image only when the detecting object is a human body.
- the performing, by the object detecting terminal, the method for extracting the facial feature data of the target image object from the second type of image may include: positioning the human face according to the location of the human body and occupying the human body region according to the human face region a ratio, a face image region of the target image object is determined in the second type of target image; and face feature data is extracted from the face image region.
- the object detecting terminal may have an area 20 at the top of the target object area 41 and an area of the target object area 41.
- the area of % is determined as the face image area 42.
- the length of the target object region 41 is h'
- the length of the face image region 42 is 0.2 h'.
- the object detecting terminal matches the extracted facial feature data with facial feature data in a database (such as a public security system), if the extracted facial feature data and the preset number are The object detection terminal may use the identity information corresponding to the face feature data as the identity information of the detection target indicated by the target image object.
- a database such as a public security system
- the identity information may include, but is not limited to, name, gender, age, ID number, home address, marital status, education, graduate school, occupation, work place, criminal record, and the like.
- the object detection terminal may adjust the shooting parameters and perform a shooting operation under the adjusted shooting parameters to obtain a new second class. Target image. Then, the object detecting terminal may perform face recognition processing on the new second type target image.
- the object detecting terminal obtains a second type of target image for performing image object recognition, such as a first type of target image for representing a temperature distribution, such as a thermal infrared image, and the like, and
- the first type of target image is analyzed and processed on the second type of target image, and the target image object in the temperature abnormal state is determined from the second type of target image, and the object with abnormal temperature can be detected quickly, and Improve the intelligence of object detection.
- the efficiency of object detection can be greatly improved.
- the object detection terminal may further determine location information and identity information of the detection object; according to the location information, a specific occasion such as an airport, a gateway, a railway station, and the like
- the worker can find the detection object and further observe or isolate it; according to the identity information, the worker can determine the identity of the detection object.
- the embodiments of the present invention can facilitate object detection work for specific occasions.
- FIG. 10 is a schematic block diagram of an object detecting terminal according to an embodiment of the present invention.
- the terminal in this embodiment as shown in FIG. 10 may include one or more processors 51 and a memory 52.
- the processor 51 and the memory 52 are connected by a bus 53.
- the memory 53 is for storing a computer program, the computer program comprising program instructions.
- the processor 51 is configured to invoke the program instruction execution:
- the first type of target image is an image for representing a temperature distribution
- the second type of target image is An image for image object recognition
- the target image object is an object in a temperature abnormal state determined in the second type of target image and determined according to a temperature distribution represented by the first type of target image.
- the processor 51 is configured to invoke the program instruction to perform the analyzing process on the second type of target image according to the first type of target image, from the second type of target image. Specific execution when determining the target image object:
- a target image object is identified from the target object area.
- the processor 51 is configured to invoke the program instruction to perform the analyzing process on the second type of target image according to the first type of target image, from the second type of target image. Specific execution when determining the target image object:
- a target image object is identified from the target object area.
- the processor 51 is configured to execute the program instruction to perform the capturing to obtain the first type of target image and the second type of target image:
- the target detection area is photographed to obtain a first type of target image and a second type of target image.
- the processor 51 is configured to: when the program instruction is executed to perform the determining, according to the first type of image of the environment to be detected, the target detection area in the environment to be detected:
- Determining a target detection area in the environment to be detected according to the target image area Determining a target detection area in the environment to be detected according to the target image area.
- processor 51 is configured to invoke the program instructions to further perform:
- processor 51 is configured to invoke the program instructions to further perform:
- the current photographing direction is adjusted to face the subject, and the subject is photographed.
- the processor 51 is configured to: when the program instruction is executed to perform, determining, according to the location description information, a location of the detection object indicated by the target image object:
- the first photographing angle is a photographing angle in a vertical direction
- the second photographing angle is a photographing angle in a horizontal direction.
- the detecting object includes a plurality of; the processor 51 is configured to invoke the program instruction to perform the determining, according to the position coordinates of the photographic object, the location of the detecting object:
- the position coordinates of the other detection objects are determined according to the position coordinates of the photographic subject and the relative positional relationship.
- processor 51 is configured to invoke the program instructions to further perform:
- processor 51 is configured to invoke the program instructions to further perform:
- the processor 51 is configured to: when the program instruction is executed to perform the extracting the facial feature data of the target image object from the second type of image:
- the face feature data is extracted from the face image area.
- processor 51 is configured to invoke the program instructions to further perform:
- the processor 51 may be a central processing unit (CPU), and the processor 51 may also be another general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, and the like.
- the general purpose processor may be a microprocessor or the processor 51 or any conventional processor or the like.
- the memory 52 may include a Read-Only Memory (ROM) and a Random Access Memory (RAM), and provides the processor 51 with computer programs and data. A portion of the memory 52 may also include a non-volatile random access memory. For example, the memory 52 can also store information of the device type.
- ROM Read-Only Memory
- RAM Random Access Memory
- the processor 52 described in the embodiment of the present invention may implement the implementation of the object detection method described in FIG. 1 or FIG. 2 of the present application.
- the description of related parts of the method in the embodiment of the present invention I will not repeat them here.
- the processor 51 calls a program instruction stored in the memory 52 to obtain a thermal infrared image or the like, a first type of target image for indicating a temperature distribution, an RGB image, or the like for performing an image. a second type of target image recognized by the object, and analyzing the second type of target image according to the first type of target image, and determining a target image object in a temperature abnormal state from the second type of target image, It can quickly detect objects with abnormal temperature and improve the intelligence of object detection. When the number of objects requiring temperature detection is large, the efficiency of object detection can be greatly improved. Further, when the detection object indicated by the target image is a human body, the processor 51 calls a program instruction stored in the memory 52, and may also determine location information and identity information of the detection object, thereby being specific Facilitating object inspection work.
- a computer readable storage medium is also provided in an embodiment of the present invention, the computer readable storage medium storing a computer program including program instructions, the processor being configured to invoke the program The object performs the object detection method shown in FIG. 1 or FIG. 2 of the present application.
- the computer readable storage medium may be an internal storage unit of the object detection terminal described in any of the foregoing embodiments, such as a hard disk or a memory of the object detection terminal.
- the computer readable storage medium may also be an external storage device of the object detecting terminal, such as a plug-in hard disk equipped on the object detecting terminal, a smart memory card (SMC), and a secure digital (Secure Digital) , SD) card, flash card (Flash Card), etc.
- the computer readable storage medium may also include both an internal storage unit of the object detecting terminal and an external storage device.
- the computer readable storage medium is for storing the computer program and other programs and data required by the object detection terminal.
- the computer readable storage medium can also be used to temporarily store data that has been output or is about to be output.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
一种对象检测方法、对象检测终端及计算机可读介质,其中,所述方法包括:拍摄得到第一类目标图像和第二类目标图像(S101);根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象(S102);所述第一类目标图像是用于表示温度分布的图像,所述第二类目标图像是用于进行图像对象识别的图像;所述目标图像对象是位于所述第二类目标图像中且根据所述第一类目标图像所表示的温度分布确定的处于温度异常状态的对象。可以快速地检测出温度异常的对象。
Description
本发明涉及图像处理技术领域,尤其涉及一种对象检测方法、对象检测终端及计算机可读介质。
由于体温可以在一定程度上反映人体的身体健康状况,特别是在流感等传染病扩散严重的时期,通过体温检测排查出体温异常的人体在多种场合和领域(如机场、关口安检,异常疫情监控等等)都显得非常重要。
如果想要对出入某些特定场合(如机场、关口、火车站等)的个体进行体温检测,需要安排专门的工作人员对这些个体逐个进行体温检测或对部分个体进行体温抽查。可以看出,传统的检测体温的方法非常低效。
发明内容
本发明实施例提供了一种对象检测方法,可以快速地检测出温度异常的对象。
第一方面,本发明实施例提供了一种对象检测方法,包括:
拍摄得到第一类目标图像和第二类目标图像;
根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象;
所述第一类目标图像是用于表示温度分布的图像,所述第二类目标图像是用于进行图像对象识别的图像;
所述目标图像对象是位于所述第二类目标图像中且根据所述第一类目标图像所表示的温度分布确定处于温度异常状态的对象。
第二方面,本发明实施例提供了一种对象检测终端,包括处理器和存储器:
所述存储器,用于存储计算机程序,所述计算机程序包括程序指令;
所述处理器,用于调用所述程序指令执行:
拍摄得到第一类目标图像和第二类目标图像;
根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第
二类目标图像中确定出目标图像对象;
所述第一类目标图像是用于表示温度分布的图像,所述第二类目标图像是用于进行图像对象识别的图像;
所述目标图像对象是位于所述第二类目标图像中且根据所述第一类目标图像所表示的温度分布确定的处于温度异常状态的对象。
第三方面,本发明实施例提供了一种计算机可读存储介质,所述计算机存储介质存储有计算机程序,所述计算机程序包括程序指令,所述程序指令当被处理器执行时使所述处理器执行上述第一方面的方法。
本发明实施例通过拍摄得到热红外图像等用于表示温度分布的第一类目标图像以及RGB图像等用于进行图像对象识别的第二类目标图像,并根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出处于温度异常状态的目标图像对象,可以快速地检测出温度异常的对象,还可以提高对象检测的智能性。当需要进行温度检测的对象的数目较大时,可以大大提高对象检测的效率。
为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1是本发明实施例提供的一种对象检测方法的流程示意图;
图2是本发明实施例提供的另一种对象检测方法的流程示意图;
图3是本发明实施例提供的一种从第二类目标图像中划分出对象区域的示意图;
图4是本发明实施例提供的一种从第一类目标图像中划分出与如图3所示的对象区域对应的温度检测区域的示意图;
图5是本发明实施例提供的一种从第一类目标图像中划分出温度异常区域的示意图;
图6是本发明实施例提供的一种从第二类目标图像中划分出如图5所示的温度异常区域对应的对象检测区域的示意图;
图7是本发明实施例提供的一种对象检测终端与拍摄对象之间的位置关系图;
图8是本发明实施例提供的另一种对象检测终端与拍摄对象之间的位置关系图;
图9是本发明实施例提供的一种人脸图像区域与目标对象区域之间的位置和比例关系图;
图10是本发明实施例提供的一种对象检测终端的结构示意图。
下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例,都属于本发明保护的范围。
由于体温可以在一定程度上反映人体的身体健康状况,因此在安检工作中,体温是非常重要的检测指标之一。但是,在机场、关口、火车站等人流非常密集的场合,只能偶尔对部分个体进行体温抽查,这样往往不能及时发现可能携带异常病毒的个体(即体温异常的个体),还存在漏检的风险。如果将大量个体集中在某个地点,由工作人员逐个地进行体温检测,又会导致体温检测的工作量巨大,耗时耗力,且检测效率低下。
为了解决上述问题,本发明实施例提供一种对象检测方法,所述对象检测方法可具体应用于对象检测终端。在本发明实施例的对象检测方法中,所述对象检测终端首先获取热红外图像和RGB(red红、green绿、blue蓝)图像。
在一个具体的实施例中,所述对象检测终端可以包括热红外摄像头(也可称为热红外相机)和可见光摄像头(也可称为可见光相机)。其中,所述热红外摄像头和所述可见光摄像头分别用于拍摄得到热红外图像和RGB图像。从而,所述对象检测终端可以通过控制所述热红外摄像头和所述可见光摄像头分别进行拍摄操作来获取所述热红外图像和所述RGB图像。作为一种可选的实施方式,所述对象检测终端可以控制所述热红外摄像头和所述可见光摄像头以同样的姿态同时进行拍摄操作,或者说,所述对象检测终端可以控制所述热红外摄像头和所述可见光摄像头对同一场景同时进行拍摄操作。具体实现中,所
述对象检测终端可以设置在检票口、登机口等需要进行体温检测的个体必经的地方。
在另一个具体的实施例中,所述对象检测终端可以包括双光谱摄像头(也可称为双光谱相机)。其中,所述双光谱摄像头可用于拍摄得到热红外图像和RGB图像。具体地,所述双光谱摄像头可以具有第一拍摄模式和第二拍摄模式。在所述第一拍摄模式下,所述双光谱摄像头可用于拍摄得到热红外图像。在所述第二拍摄模式下,所述双光谱摄像头可用于拍摄得到RGB图像。从而,所述对象检测终端可以通过控制所述双光谱摄像头以所述第一拍摄模式和所述第二拍摄模式分别进行拍摄操作来获取所述热红外图像和所述RGB图像。作为一种可选的实施方式,所述对象检测终端可以控制所述双光谱摄像头在第一时刻以所述第一拍摄模式进行拍摄操作来获取所述热红外图像,以及控制所述双光谱摄像头在第二时刻以所述第二拍摄模式进行拍摄操作来获取所述RGB图像,其中,所述第一时刻与所述第二时刻之间的时间间隔为预设时间间隔。可以理解的是,为了使得所述热红外图像和所述RGB图像是对同一场景拍摄的两种不同属性的图像,所述预设时间间隔必须足够小。并且,在所述第一时刻和所述第二时刻,所述双光谱摄像头的拍摄姿态相同。具体实现中,所述对象检测终端可以设置在检票口、登机口等需要进行体温检测的个体必经的地方。
在又一个具体的实施例中,所述对象检测终端可以从拍摄终端处获取所述热红外图像和所述RGB图像。也就是说,所述热红外图像和所述RGB图像是由所述拍摄终端拍摄得到并通过有线或无线的方式发送给所述对象检测终端的。可选地,所述拍摄终端可以通过设置在其中的热红外摄像头和可见光摄像头分别拍摄得到所述热红外图像和所述RGB图像。可选地,所述拍摄终端还可以通过设置在其中的双光谱摄像头拍摄得到所述热红外摄像头和所述RGB图像。具体实现中,所述拍摄终端可以设置在检票口、登机口等需要进行体温检测的个体必经的地方,所述对象检测终端可以设置在监控室。可选地,所述对象检测终端可以获取多个拍摄终端发送的热红外图像和RGB图像。
需要说明的是,所述热红外图像和所述RGB图像是两种不同属性的图像。具体地,热红外图像是用于表示温度分布的图像,手机、平板电脑、数码相机等具有拍摄功能的终端拍摄得到的彩色图像就是RGB图像。
进一步地,所述对象检测终端可以将拍摄得到的所述热红外图像和所述RGB图像进行对齐并融合,从而可以得到所述RGB图像中的每个位置(如每个像素点)的温度值。
进一步地,所述对象检测终端可以根据所述热红外图像所表示的温度分布,将所述RGB图像中温度在异常体温范围内的图像区域确定为感兴趣图像区域。然后,所述对象检测终端可以使用人体检测算法对所述感兴趣图像区域进行人体检测处理,确定出所述感兴趣图像区域中的人体图像区域。
需要说明的是,人体的正常体温范围一般约为36~37.5℃。当人体的体温超过37.5℃时,称为发热(或发烧)。一般来说,人体发热时体温不会超过42℃。并且,当人体的体温低于30℃时,人就会失去知觉。从而,30~36℃以及37.5~42℃可以作为异常体温范围预置在所述对象检测终端中。举例来说,在某种会引起患者发热的传染性疾病扩散严重的时期,如果需要在机场、关口、火车站等特定场合检测出可能患有此传染性疾病的人体时,可以将37.5~42℃作为异常体温范围,也就是将体温在37.5~42℃的人体确定为可能患有此传染性疾病的人体。
进一步地,根据确定出的所述人体图像区域、人脸图像区域在所述人体图像区域中的位置以及人脸图像区域占所述人体图像区域的比例,所述对象检测终端可以从所述人体图像区域中确定出人脸图像区域。举例来说,在检票口、登机口等地方,人体一般处于站立姿态,在这种情形下,人脸图像区域在所述人体图像区域的顶部。可选地,人脸图像区域占所述人体图像区域的比例可以是一个预置在所述对象检测终端中的常数(如20%)。
进一步地,所述对象检测终端可以对所述人脸图像区域进行人脸识别处理,提取出所述人脸图像区域中的人脸特征数据。然后,所述对象检测终端还可以将提取出的所述人脸特征数据与数据库(如公安系统)中的人脸特征数据进行匹配,确定出体温异常的人体的身份信息。
进一步地,由于拍摄所述RGB图像的对象检测终端或拍摄终端在拍摄所述RGB图像时的拍摄姿态是已知的,根据所述人体图像区域在所述RGB图像中的位置以及所述拍摄姿态,所述对象检测终端可以确定出体温异常的人体相对于所述摄像头的位置信息。由于拍摄所述RGB图像的对象检测终端或拍摄终端的安装位置信息(如坐标值)及其在拍摄所述RGB图像的摄像头的拍摄
姿态都是已知的,根据所述人体图像区域在所述RGB图像中的位置,所述安装位置信息以及所述拍摄姿态及其位置信息,所述对象检测终端还可以进一步确定出所述体温异常的人体的绝对位置信息。
本发明实施例的对象检测方法可以方便地从人群中确定出体温异常的个体,还可以根据拍摄设备(如上述对象检测终端或拍摄终端)的拍摄姿态(或者安装位置信息以及拍摄姿态),确定出所述体温异常的个体的位置信息,从而找到所述体温异常的个体,利用人脸识别技术,还可以从公安系统中获取所述体温异常的个体的身份信息。
本发明实施例的对象检测方法可以应用到安检工作中。在机场、关口、火车站等特定场合,如果将所述对象检测终端或拍摄终端安装在安检的必经路口或者检票口、登机口等需要进行体温检测的个体必经的地方,那么对象检测终端可以自动进行检测工作,确定出体温异常的个体,并提供所述体温异常的个体的位置信息和身份信息,安检工作人员只需要根据所述位置信息找到所述体温异常的个体,并对所述身份信息进行确认,高效地完成了基础安检工作,大大提高了检测效率。
本发明实施例的对象检测方法还可以应用于异常疫情监控。举例来说,在SARS、禽流感等特殊疫情扩散严重的时期,体温检测尤为重要,特别是在广场、商业街等人口密集的公共场合,需要及时检测出体温异常的个体并及时对其进行隔离,防止疫情的进一步扩散。在这种情形下,可以将对象检测终端或拍摄终端安装在这些公共场合的出口处、入口处等位置。
本发明实施例的对象检测方法还可以应用于厂房异常温度区域监控。在高危工业厂房中,如果存在泄漏点,就会引发安全隐患,而且通常会伴随着周围的温度异常。在这种情形下,可以将对象检测终端或拍摄终端安装在厂房内的四个角落等位置。根据预设的正常温度范围(或异常温度范围),如果对象检测终端检测到某个区域的温度与正常温度相差太大(或处于异常温度范围内),则可以将该温度异常的区域的图像上报到系统中,为安全检查提供参考,重点关注该区域。
当然,本发明实施例的方法还可以应用到其他场合或领域中,本发明实施例不做限定。例如,可以将对象检测终端或拍摄终端安装在无人机或巡检车辆上,提高检测工作的机动性和灵活性。
下面结合附图1-10,对本发明实施例的对象检测方法及对象检测终端进行详细的描述。
请参见图1,是本发明实施例提供的一种对象检测方法的流程示意图。如图1所示,所述对象检测方法可以包括:
S101:拍摄得到第一类目标图像和第二类目标图像。
其中,所述第一类目标图像是用于表示温度分布的图像。在一个具体的实施例中,所述第一类目标图像为热红外摄像头拍摄得到的热红外图像。
其中,所述第二类目标图像是用于进行图像对象识别的图像。在一个具体的实施例中,所述第二类目标图像为可见光摄像头拍摄得到的RGB图像,或者包括RGB信息的任何图像。
需要说明的是,手机、平板电脑、数码相机、单反相机等具有拍摄功能的终端中的后置摄像头和前置摄像头即为可见光摄像头,拍摄得到的彩色图像即为RGB图像。
可选地,所述第二类目标图像也可以是可见光摄像头拍摄得到的HSV图像、YUV图像等等,本发明实施例不做限定。
还需要说明的是,本发明实施例的对象检测方法可具体应用于对象检测终端中。其中,所述对象检测终端可以包括热红热摄像头和可见光摄像头。从而,所述对象检测终端执行所述拍摄得到第一类目标图像和第二类目标图像可以具体包括:控制所述热红外摄像头进行拍摄操作得到第一类目标图像;控制所述可见光摄像头进行拍摄操作得到第二类目标图像。为了保证所述第一类目标图像和所述第二类目标图像是对同一场景拍摄得到的两种不同属性的图像,所述对象检测终端可以控制所述热红外摄像头和所述可见光摄像头以同样的姿态同时进行拍摄操作。
S102:根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象。
其中,所述目标图像对象是位于所述第二类目标图像中且根据所述第一类目标图像所表示的温度分布确定的处于温度异常状态的对象。
作为一种可选的实施方式,所述对象检测终端执行所述根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出
目标图像对象可以具体包括:对所述第二类目标图像进行对象检测处理,得到至少一个对象区域;根据所述第一类目标图像所表示的温度分布,从所述至少一个对象区域中确定出目标对象区域;从所述目标对象区域中识别出目标图像对象。
其中,所述对象检测处理例如可以是人体检测处理。此时,所述对象区域可以具体是人体图像区域。在这种情形下,所述目标图像对象可以具体是人体。当然,所述对象检测处理也可以是动物检测处理,或者是人体和动物检测处理等等,本发明实施例不做限定。
在一个具体的实施例中,所述对象区域可以是矩形区域。举例来说,当所述对象检测终端通过对所述第二类目标图像进行人体检测处理之后,检测出所述第二类目标图像中包括三个人体图像区域,那么所述对象检测终端可以从所述第二类目标图像中确定出三个矩形的对象区域。
在对所述第二类目标图像进行对象检测处理,得到至少一个对象区域之后,根据所述第一类目标图像与所述第二类目标图像之间的区域对应关系,所述对象检测终端可以确定出所述第一类目标图像中与所述对象区域对应的温度检测区域。进一步地,根据所述第一类目标图像所表示的温度分布,所述对象检测终端可以将所述第二类目标图像中与存在温度在异常温度范围内的像素点的温度检测区域对应的对象区域作为目标对象区域。其中,所述异常温度范围可以预置在所述对象检测终端中。
作为另一种可选的实施方式,所述对象检测终端执行所述根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象可以具体包括:根据所述第一类目标图像所表示的温度分布,从所述第二类目标图像中确定出对象检测区域;对所述对象检测区域进行对象检测处理,得到目标对象区域;从所述目标对象区域中识别出目标图像对象。
可选地,所述对象检测区域可以是矩形区域。在一个具体的实施例中,根据所述第一类目标图像所表示的温度分布,所述对象检测终端可以确定出所述第一类目标图像中存在温度在异常温度范围内的像素点的温度异常区域。从而,根据所述第一类目标图像与所述第二类目标图像之间的区域对应关系,所述对象检测终端可以确定出所述第二类目标图像中与所述温度异常区域对应的对象检测区域。可选地,在所述温度异常区域中,温度在异常温度范围内的像素
点占所述温度异常区域中总的像素点的比例至少为预设比例。
其中,所述对象检测处理例如可以是人体检测处理。此时,所述目标对象区域可以具体是人体图像区域。可选地,所述目标对象区域可以是矩形区域。举例来说,当所述对象检测终端通过对所述对象检测区域进行人体检测处理之后,检测出所述对象检测区域包括一个人体图像区域,那么所述对象检测终端可以从所述对象检测区域中确定出一个目标对象区域。
在本发明实施例中,通过拍摄得到热红外图像等用于表示温度分布的第一类目标图像以及RGB图像等用于进行图像对象识别的第二类目标图像,并根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出处于温度异常状态的目标图像对象,可以快速地检测出温度异常的对象,还可以提高对象检测的智能性。当需要进行温度检测的对象的数目较大时,可以大大提高对象检测的效率。
进一步地,请参见图2,是本发明实施例提供的另一种对象检测方法的流程示意图。在图1所示的实施例的基础上,如图2所示,所述对象检测方法可以包括:
S201:拍摄得到第一类目标图像和第二类目标图像。
其中,所述第一类目标图像是用于表示温度分布的图像,所述第二类目标图像是用于进行图像对象识别的图像。
在一个具体的实施例中,所述第一类目标图像为热红外摄像头拍摄得到的热红外图像,所述第二类目标图像为可见光摄像头拍摄得到的RGB图像。
需要说明的是,手机、平板电脑、数码相机、单反相机等具有拍摄功能的终端中的后置摄像头和前置摄像头即为可见光摄像头,拍摄得到的彩色图像即为RGB图像。
还需要说明的是,本发明实施例的对象检测方法可具体应用于对象检测终端中。其中,所述对象检测终端可以包括热红热摄像头和可见光摄像头。从而,所述对象检测终端执行所述拍摄得到第一类目标图像和第二类目标图像可以具体包括:控制所述热红外摄像头进行拍摄操作得到第一类目标图像;控制所述可见光摄像头进行拍摄操作得到第二类目标图像。为了使得所述第一类目标图像和所述第二类目标图像是对同一场景拍摄得到的两种不同属性的图像,所
述对象检测终端可以控制所述热红外摄像头和所述可见光摄像头以同样的姿态(即所述热红外摄像头和所述可见光摄像头的拍摄姿态相同)同时进行拍摄操作。
在另一个具体的实施例中,所述第一类目标图像和所述第二类目标图像分别为双光谱摄像头拍摄得到的热红外图像和RGB图像。具体地,所述双光谱摄像头可以包括第一拍摄模式和第二拍摄模式。在所述第一拍摄模式下,所述双光谱摄像头可用于拍摄得到热红外图像。在所述第二拍摄模式下,所述双光谱摄像头可用于拍摄得到RGB图像。从而,所述对象检测终端可以通过控制所述双光谱摄像头以所述第一拍摄模式和所述第二拍摄模式分别进行拍摄操作,得到所述热红外图像和所述RGB图像。作为一种可选的实施方式,所述对象检测终端可以控制所述双光谱摄像头在第一时刻以所述第一拍摄模式进行拍摄操作来获取所述热红外图像,以及控制所述双光谱摄像头在第二时刻以所述第二拍摄模式进行拍摄操作来获取所述RGB图像,其中,所述第一时刻与所述第二时刻之间的时间间隔为预设时间间隔。可以理解的是,为了使得所述热红外图像和所述RGB图像是对同一场景拍摄得到的两种不同属性的图像,所述预设时间间隔必须足够小。并且,在所述第一时刻和所述第二时刻,所述双光谱摄像头的拍摄姿态相同。
可选地,所述第二类目标图像也可以是可见光摄像头拍摄得到的HSV图像、YUV图像等等,本发明实施例不做限定。
需要说明的是,作为一种可选的实施方式,所述对象检测终端可以对待检测环境进行拍摄操作得到用于表示温度分布的第一类图像和用于进行图像对象识别的第二类图像,并将所述第一类图像和所述第二类图像分别确定为所述第一类目标图像和所述第二类目标图像。
其中,所述待检测环境为所述对象检测终端负责进行对象检测的环境。举例来说,当所述对象检测终端负责对某个大厅进行对象检测时,所述大厅即为所述待检测环境。
作为另一种可选的实施方式,所述对象检测终端执行所述拍摄得到第一类目标图像和第二类目标图像可以具体包括:拍摄得到待检测环境的第一类图像,并根据所述待检测环境的第一类图像确定所述待检测环境中的目标检测区域;对所述目标检测区域进行拍摄,得到第一类目标图像和第二类目标图像。
进一步地,所述对象检测终端执行所述根据所述待检测环境的第一类图像确定所述待检测环境中的目标检测区域可以具体包括:根据预设温度范围对所述待检测环境的第一类图像进行分析处理,确定出目标图像区域,所述目标图像区域所对应的温度分布满足预设分区条件;根据所述目标图像区域确定所述待检测环境中的目标检测区域。
其中,所述目标检测区域为所述对象检测终端负责进行对象检测的环境中的部分区域。所述预设温度范围可以具体为预设的异常体温范围等异常温度范围,所述预设分区条件可以具体为所述目标图像区域中存在温度在所述预设温度范围内的像素点。
举例来说,当所述对象检测终端负责对某个大厅进行对象检测,且所述对象检测终端在水平方向的拍摄角度为0°(即所述对象检测终端没有向左或向右转动)时,根据所述第一类图像所表示的温度分布,所述对象检测终端可以判断出位于所述第一类图像左侧的区域存在温度在所述预设温度范围内的像素点,而位于所述第一类图像右侧的区域不存在温度在所述预设温度范围内的像素点(即所述大厅中位于所述对象检测终端右侧的区域中的各个位置的温度均在正常温度范围内),那么根据所述第一类图像与所述第二类图像的区域对应关系,所述对象检测终端可以将所述第二类图像中与所述第一类图像左侧的区域对应的区域确定为目标图像区域,并且,所述对象检测终端还可以将所述大厅中位于所述对象检测终端左侧的区域确定为目标检测区域。此时,所述目标检测区域成为了新的待检测环境。
S202:根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象。
其中,所述目标图像对象是位于所述第二类目标图像中且根据所述第一类目标图像所表示的温度分布确定的处于温度异常状态的对象。
作为一种可选的实施方式,所述对象检测终端执行所述根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象可以具体包括:对所述第二类目标图像进行对象检测处理,得到至少一个对象区域;根据所述第一类目标图像所表示的温度分布,从所述至少一个对象区域中确定出目标对象区域;从所述目标对象区域中识别出目标图像对象。
其中,所述对象检测处理例如可以是人体检测处理。此时,所述对象区域可以具体是人体图像区域。在这种情形下,所述目标图像对象可以具体是人体。当然,所述对象检测处理也可以是动物检测处理,或者是人体和动物检测处理等等,本发明实施例不做限定。在一个具体的实施例中,所述对象区域可以是矩形区域。
具体地,请参见图3,第二类目标图像所在的区域10中包括第一图像对象111和第二图像对象112。所述对象检测终端对所述第二类目标图像所在的区域10进行对象检测之后,可以得到第一对象区域121和第二对象区域122。其中,所述第一对象区域121和第二对象区域122可以具体为矩形区域。
请一并参见图3和图4,如图4所示,第一类目标图像所在的区域20由温度在异常温度范围内的第一区域211、温度在异常温度范围内的第二区域212以及温度不在异常温度范围内的第三区域213组成。进一步地,根据第一类目标图像和第二类目标图像的区域对应关系,所述对象检测终端可以在所述第一类目标图像所在的区域20中确定出分别与所述第一对象区域121和所述第二对象区域122对应的第一温度检测区域221和第二温度检测区域222。
需要说明的是,所述第二类目标图像所在的区域10与所述第一类目标图像所在的区域20具有相互对应关系。具体地,所述第二类目标图像所在的区域10与所述第一类目标图像所在的区域20形状和面积相同,并且,在同样的坐标系下,在所述第二类目标图像所在的区域10中位于某一坐标位置的像素点对应于在所述第一类目标图像所在的区域20中位于所述坐标位置的像素点。
进一步地,由于所述第一温度检测区域221包括所述第一区域211,即所述第一温度检测区域221中包括温度在异常温度范围内的像素点,而所述第二温度检测区域222中的像素点均不在异常温度范围内,因此所述对象检测终端可以将所述第二类目标图像所在的区域10中与所述第一温度检测区域221对应的第一对象区域121确定为目标对象区域。
进一步地,所述对象检测终端可以对所述第一对象区域121进行轮廓提取等图像处理操作,识别出所述第一图像对象111,并将所述第一图像对象111确定为目标图像对象。
作为另一种可选的实施方式,所述对象检测终端执行所述根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定
出目标图像对象可以具体包括:根据所述第一类目标图像所表示的温度分布,从所述第二类目标图像中确定出对象检测区域;对所述对象检测区域进行对象检测处理,得到目标对象区域;从所述目标对象区域中识别出目标图像对象。
请参见图5,第一类目标图像所在的区域20由温度在异常温度范围内的第一区域211、温度在异常温度范围内的第二区域212以及温度不在异常温度范围内的第三区域213组成。所述对象检测终端可以将所述第一区域211和所述第二区域212分别划分到第一温度异常区域231和第二温度异常区域232中。其中,所述第一温度异常区域231和所述第二温度异常区域232可以具体为矩形区域。在一个具体的实施例中,所述第一区域211的面积占所述第一温度异常区域231的面积的比例,以及所述第二区域212的面积占所述第二温度异常区域232的面积的比例均超过预设比例。
请一并参见图5和图6,如图6所示,第二类目标图像所在的区域10中包括第一图像对象111和第二图像对象112。根据第一类目标图像和第二类目标图像的区域对应关系,所述对象检测终端可以在所述第二类目标图像所在的区域10中确定出分别与所述第一温度异常区域231和所述第二温度异常区域232对应的第一对象检测区域131和第二对象检测区域132。
进一步地,所述对象检测终端可以分别对所述第一对象检测区域131和所述第二对象检测区域132进行对象检测处理。所述对象检测终端可以判断出所述第一对象检测区域131中存在图像对象,而所述第二对象检测区域132中不存在图像对象,从而可以将所述第一对象检测区域131作为目标对象区域。进一步地,所述对象检测终端可以对所述第一对象检测区域131进行轮廓提取等图像处理操作,识别出所述第一图像对象111,并将所述第一图像对象111确定为目标图像对象。
S203:获取拍摄位置的位置描述信息,并根据所述位置描述信息确定所述目标图像对象所指示的检测对象的位置。
其中,所述拍摄位置指的是拍摄得到的所述第一类目标图像或所述第二类目标图像的摄像头所在的位置。在本发明实施例中,所述拍摄位置可以具体是所述对象检测终端所在的位置,即所述对象检测终端的安装位置。
其中,所述位置描述信息可以为相对位置描述信息或绝对位置描述信息。其中,所述相对位置描述信息指的是所述拍摄位置相对于某个参照物而言的位
置信息。具体地,所述相对位置描述信息可以包括所述对象检测终端的安装高度(即拍摄高度)。举例来说,所述相对位置描述信息例如可以是:所述对象检测终端位于1号大厅,其安装高度为3m。其中,所述绝对位置描述信息例如可以是所述对象检测终端在三维坐标系中的坐标值。
所述位置描述信息还可以包括所述对象检测终端的拍摄姿态信息,如所述对象检测终端在竖直方向的拍摄角度(第一拍摄角度)和所述对象检测终端在水平方向的拍摄角度(第二拍摄角度)。
具体地,所述对象检测终端执行所述根据所述位置描述信息确定所述目标图像对象所指示的检测对象的位置可具体包括:根据拍摄高度和第一拍摄角度,得到所述拍摄对象和拍摄位置之间的距离;根据第二拍摄角度、所述拍摄对象和拍摄位置之间的距离以及所述位置描述信息所描述的位置坐标,得到所述拍摄对象的位置坐标;根据所述拍摄对象的位置坐标确定所述检测对象的位置。
进一步具体地,以三维直角坐标系为例,如图7所示,当所述对象检测终端31相对于所述拍摄对象32的高度(即拍摄高度)为h,第一拍摄角度为θ时,所述对象检测终端31和所述拍摄对象32的安装位置(即拍摄位置)对应的地面位置O之间的直线距离d可以表示为:d=h·tanθ。如图8所示,当第二拍摄角度为α时,所述直线距离d在x轴的分量d1可以表示为:d1=d·cosα,所述直线距离d在y轴的分量d2可以表示为:d2=d·sinα。根据d1和d2,所述对象检测终端31可以确定出所述拍摄对象32相对于所述对象检测终端31的位置。
进一步地,当所述对象检测终端31在三维直角坐标系中的坐标值已知时,所述对象检测终端31可以确定出所述拍摄对象32在三维直角坐标系中的坐标值。具体地,当所述对象检测终端31在x轴的横坐标值为a时,所述拍摄对象32在x轴的横坐标值a1可以表示为:a1=a+d1;当所述对象检测终端31在y轴的纵坐标值为b时,所述拍摄对象32在y轴的横坐标值b1可以表示为:b1=b+d2;当所述对象检测终端31在z轴的竖坐标值为c时,所述拍摄对象32在y轴的竖坐标值c1可以表示为:c1=c-h。根据a1、b1和c1,所述对象检测终端31可以确定出所述拍摄对象32在三维直角坐标系中的绝对位置。
需要说明的是,当所述检测对象为单个对象时,所述拍摄对象即为所述检测对象。当所述检测对象包括多个对象时,所述拍摄对象为其中一个检测对象。
当所述检测对象为多个对象时,所述对象检测终端执行所述根据拍摄对象的位置坐标确定所述检测对象的位置可以具体包括:根据所述第二类目标图像确定所述拍摄对象与其他检测对象之间的相对位置关系;根据所述拍摄对象的位置坐标以及所述相对位置关系,确定其他检测对象的位置坐标。
可以理解的是,根据所述拍摄对像与其他检测对象在所述第二类目标图像中的相对位置关系,以及图像中的位置关系和实际位置关系之间的转化规则,所述对象检测终端可以确定出所述拍摄对象与其他检测对象之间的实际位置关系。其中,所述实际位置关系例如可以是:所述其他检测对象中的某一检测对象位于所述拍摄对象的正后方1m处。从而,根据所述实际位置关系和所述拍摄对象的位置坐标,所述对象检测终端可以确定出所述其他检测对象的位置坐标。
S204:从所述第二类目标图像中提取所述目标图像对象的人脸特征数据,并根据人脸特征数据匹配确定所述检测对象的身份信息。
可以理解的是,所述对象检测终端在执行所述从所述第二类目标图像中提取所述目标图像对象的人脸特征数据,并根据人脸特征数据匹配确定所述检测对象的身份信息之前还可以包括:判断所述目标图像对象所指示的检测对象是否为人体;如果是,则执行所述从所述第二类目标图像中提取所述目标图像对象的人脸特征数据。也就是说,只有当检测对象为人体时,所述对象检测终端才需要对所述第二类目标图像进行人脸识别处理。
具体地,所述对象检测终端在执行所述从所述第二类图像中提取所述目标图像对象的人脸特征数据可以具体包括:根据人脸位于人体的位置以及人脸区域占人体区域的比例,在所述第二类目标图像中确定所述目标图像对象的人脸图像区域;从所述人脸图像区域中提取人脸特征数据。
举例来说,当人体处于站立姿态时,人脸位于人体的顶部。当人脸区域占人体区域的比例设置为预设数值20%时,如图9所示,所述对象检测终端可以将位于目标对象区域41顶部且面积为所述目标对象区域41的面积的20%的区域确定为人脸图像区域42。在图9中,所述目标对象区域41的长度为h′,所述人脸图像区域42的长度为0.2h′。
进一步地,所述对象检测终端将提取出的人脸特征数据与数据库(如公安系统)中的人脸特征数据进行匹配,如果提取出的人脸特征数据与所述预置数
据库中的某一人脸特征数据相同,那么所述对象检测终端可以将所述人脸特征数据对应的身份信息作为所述目标图像对象所指示的检测对象的身份信息。
可以理解的是,所述数据库中关联存储了人脸特征数据和身份信息。其中,所述身份信息可以包括但不限于姓名、性别、年龄、身份证号、家庭住址、婚姻状况、学历、毕业院校、职业、工作地点、犯罪记录等等。
需要说明的是,如果所述第二类目标图像不满足人脸特征提取条件,那么所述对象检测终端可以调整拍摄参数,并在调整后的拍摄参数下进行拍摄操作,得到新的第二类目标图像。然后,所述对象检测终端可以对所述新的第二类目标图像进行人脸识别处理。
在本发明实施例中,所述对象检测终端通过拍摄得到热红外图像等用于表示温度分布的第一类目标图像以及RGB图像等用于进行图像对象识别的第二类目标图像,并根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出处于温度异常状态的目标图像对象,可以快速地检测出温度异常的对象,还可以提高对象检测的智能性。当需要进行温度检测的对象的数目较大时,可以大大提高对象检测的效率。此外,当所述目标图像所指示的检测对象为人体时,所述对象检测终端还可以确定所述检测对象的位置信息和身份信息;根据所述位置信息,机场、关口、火车站等特定场合的工作人员可以找到所述检测对象并对其进行进一步观察或隔离;根据所述身份信息,工作人员可以确定所述检测对象的身份。从而,本发明实施例可以为特定场合的对象检测工作提供便利。
请参见图10,是本发明实施例提供的一种对象检测终端的示意性框图。如图10所示的本实施例中的终端可以包括:一个或多个处理器51和存储器52。所述处理器51和所述存储器52通过总线53连接。所述存储器53用于存储计算机程序,所述计算机程序包括程序指令。
具体地,所述处理器51被配置用于调用所述程序指令执行:
拍摄得到第一类目标图像和第二类目标图像;
根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象;
所述第一类目标图像是用于表示温度分布的图像,所述第二类目标图像是
用于进行图像对象识别的图像;
所述目标图像对象是位于所述第二类目标图像中且根据所述第一类目标图像所表示的温度分布确定的处于温度异常状态的对象。
可选地,所述处理器51被配置用于调用所述程序指令执行所述根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象时具体执行:
对所述第二类目标图像进行对象检测处理,得到至少一个对象区域;
根据所述第一类目标图像所表示的温度分布,从所述至少一个对象区域中确定出目标对象区域;
从所述目标对象区域中识别出目标图像对象。
可选地,所述处理器51被配置用于调用所述程序指令执行所述根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象时具体执行:
根据所述第一类目标图像所表示的温度分布,从所述第二类目标图像中确定出对象检测区域;
对所述对象检测区域进行对象检测处理,得到目标对象区域;
从所述目标对象区域中识别出目标图像对象。
可选地,所述处理器51被配置用于调用所述程序指令执行所述拍摄得到第一类目标图像和第二类目标图像时具体执行:
拍摄得到待检测环境的第一类图像,并根据所述待检测环境的第一类图像确定所述待检测环境中的目标检测区域;
对所述目标检测区域进行拍摄,得到第一类目标图像和第二类目标图像。
可选地,所述处理器51被配置用于调用所述程序指令执行所述根据所述待检测环境的第一类图像确定所述待检测环境中的目标检测区域时具体执行:
根据预设温度范围对所述待检测环境的第一类图像进行分析处理,确定出目标图像区域,所述目标图像区域所对应的温度分布满足预设分区条件;
根据所述目标图像区域确定所述待检测环境中的目标检测区域。
可选地,所述处理器51被配置用于调用所述程序指令还执行:
获取拍摄位置的位置描述信息;
根据所述位置描述信息确定所述目标图像对象所指示的检测对象的位置。
可选地,所述处理器51被配置用于调用所述程序指令还执行:
从所述检测对象中确定出拍摄对象;
将当前拍摄方向调整至正对所述拍摄对象,并对所述拍摄对象进行拍摄。
可选地,所述处理器51被配置用于调用所述程序指令执行所述根据所述位置描述信息确定所述目标图像对象所指示的检测对象的位置时具体执行:
根据拍摄高度和第一拍摄角度,得到所述拍摄对象和拍摄位置之间的距离;
根据第二拍摄角度、所述拍摄对象和拍摄位置之间的距离以及所述位置描述信息所描述的位置坐标,得到所述拍摄对象的位置坐标;
根据所述拍摄对象的位置坐标确定所述检测对象的位置;
所述第一拍摄角度是在竖直方向的拍摄角度,所述第二拍摄角度是在水平方向的拍摄角度。
可选地,所述检测对象包括多个;所述处理器51被配置用于调用所述程序指令执行所述根据拍摄对象的位置坐标确定所述检测对象的位置时具体执行:
根据所述第二类目标图像确定所述拍摄对象与其他检测对象之间的相对位置关系;
根据所述拍摄对象的位置坐标以及所述相对位置关系,确定其他检测对象的位置坐标。
可选地,所述处理器51被配置用于调用所述程序指令还执行:
从所述第二类目标图像中提取所述目标图像对象的人脸特征数据;
根据人脸特征数据匹配确定所述目标图像对象所指示的检测对象的身份信息。
可选地,所述处理器51被配置用于调用所述程序指令还执行:
判断所述目标图像对象所指示的检测对象是否为人体;
如果是,则执行所述从所述第二类目标图像中提取所述目标图像对象的人脸特征数据。
可选地,所述处理器51被配置用于调用所述程序指令执行所述从所述第二类图像中提取所述目标图像对象的人脸特征数据时具体执行:
根据人脸位于人体的位置以及人脸区域占人体区域的比例,在所述第二类目标图像中确定所述目标图像对象的人脸图像区域;
从所述人脸图像区域中提取人脸特征数据。
可选地,所述处理器51被配置用于调用所述程序指令还执行:
如果所述第二类目标图像不满足人脸特征提取条件,调整拍摄参数;
在调整后的拍摄参数下进行拍摄,得到新的第二类目标图像。
应当理解,在本发明实施例中,所述处理器51可以是中央处理单元(Central Processing Unit,CPU),所述处理器51还可以是其他通用处理器、数字信号处理器(Digital Signal Processor,DSP)、专用集成电路(Application Specific Integrated Circuit,ASIC)、现场可编程门阵列(Field-Programmable Gate Array,FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。所述通用处理器可以是微处理器或者所述处理器51也可以是任何常规的处理器等。
所述存储器52可以包括只读存储器(Read-Only Memory,ROM)和随机存取存储器(Random Access Memory,RAM),并向所述处理器51提供计算机程序和数据。所述存储器52的一部分还可以包括非易失性随机存取存储器。例如,所述存储器52还可以存储设备类型的信息。
具体实现中,本发明实施例中所描述的处理器52可执行本申请图1或图2所描述的对象检测方法的实现方式,具体技术细节可以参考本发明实施例方法的相关部分的描述,在此不再赘述。
在本发明实施例中,所述处理器51调用存储在所述存储器52中的程序指令,通过拍摄得到热红外图像等用于表示温度分布的第一类目标图像以及RGB图像等用于进行图像对象识别的第二类目标图像,并根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出处于温度异常状态的目标图像对象,可以快速地检测出温度异常的对象,还可以提高对象检测的智能性。当需要进行温度检测的对象的数目较大时,可以大大提高对象检测的效率。此外,当所述目标图像所指示的检测对象为人体时,所述处理器51调用存储在所述存储器52中的程序指令,还可以确定所述检测对象的位置信息和身份信息,从而为特定场合的对象检测工作提供便利。
在本发明的实施例中还提供一种计算机可读存储介质,所述计算机可读存储介质存储有计算机程序,所述计算机程序中包括程序指令,所述处理器被配置用于调用所述程序指令,执行本申请图1或图2所示的对象检测方法。
所述计算机可读存储介质可以是前述任一实施例所述的对象检测终端的内部存储单元,例如所述对象检测终端的硬盘或内存。所述计算机可读存储介质也可以是所述对象检测终端的外部存储设备,例如所述对象检测终端上配备的插接式硬盘,智能存储卡(Smart Media Card,SMC),安全数字(Secure Digital,SD)卡,闪存卡(Flash Card)等。进一步地,所述计算机可读存储介质还可以既包括所述对象检测终端的内部存储单元也包括外部存储设备。所述计算机可读存储介质用于存储所述计算机程序以及所述对象检测终端所需的其他程序和数据。所述计算机可读存储介质还可以用于暂时地存储已经输出或者将要输出的数据。
以上所述,仅为本发明的具体实施方式,但本发明的保护范围并不局限于此,任何熟悉本技术领域的技术人员在本发明揭露的技术范围内,可轻易想到各种等效的修改或替换,这些修改或替换都应涵盖在本发明的保护范围之内。因此,本发明的保护范围应以权利要求的保护范围为准。
Claims (27)
- 一种对象检测方法,其特征在于,包括:拍摄得到第一类目标图像和第二类目标图像;根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象;所述第一类目标图像是用于表示温度分布的图像,所述第二类目标图像是用于进行图像对象识别的图像;所述目标图像对象是位于所述第二类目标图像中且根据所述第一类目标图像所表示的温度分布确定的处于温度异常状态的对象。
- 根据权利要求1所述的方法,其特征在于,所述根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象,包括:对所述第二类目标图像进行对象检测处理,得到至少一个对象区域;根据所述第一类目标图像所表示的温度分布,从所述至少一个对象区域中确定出目标对象区域;从所述目标对象区域中识别出目标图像对象。
- 根据权利要求1所述的方法,其特征在于,所述根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象,包括:根据所述第一类目标图像所表示的温度分布,从所述第二类目标图像中确定出对象检测区域;对所述对象检测区域进行对象检测处理,得到目标对象区域;从所述目标对象区域中识别出目标图像对象。
- 根据权利要求1-3任一项所述的方法,其特征在于,所述拍摄得到第一类目标图像和第二类目标图像,包括:拍摄得到待检测环境的第一类图像,并根据所述待检测环境的第一类图像 确定所述待检测环境中的目标检测区域;对所述目标检测区域进行拍摄,得到第一类目标图像和第二类目标图像。
- 根据权利要求4所述的方法,其特征在于,所述根据所述待检测环境的第一类图像确定所述待检测环境中的目标检测区域,包括:根据预设温度范围对所述待检测环境的第一类图像进行分析处理,确定出目标图像区域,所述目标图像区域所对应的温度分布满足预设分区条件;根据所述目标图像区域确定所述待检测环境中的目标检测区域。
- 根据权利要求1-5任一项所述的方法,其特征在于,还包括:获取拍摄位置的位置描述信息;根据所述位置描述信息确定所述目标图像对象所指示的检测对象的位置。
- 根据权利要求6所述的方法,其特征在于,还包括:从所述检测对象中确定出拍摄对象;将当前拍摄方向调整至正对所述拍摄对象,并对所述拍摄对象进行拍摄。
- 根据权利要求7所述的方法,其特征在于,所述根据所述位置描述信息确定所述目标图像对象所指示的检测对象的位置,包括:根据拍摄高度和第一拍摄角度,得到所述拍摄对象和拍摄位置之间的距离;根据第二拍摄角度、所述拍摄对象和拍摄位置之间的距离以及所述位置描述信息所描述的位置坐标,得到所述拍摄对象的位置坐标;根据所述拍摄对象的位置坐标确定所述检测对象的位置;所述第一拍摄角度是在竖直方向的拍摄角度,所述第二拍摄角度是在水平方向的拍摄角度。
- 根据权利要求8所述的方法,其特征在于,所述检测对象包括多个,所述根据拍摄对象的位置坐标确定所述检测对象的位置,包括:根据所述第二类目标图像确定所述拍摄对象与其他检测对象之间的相对位置关系;根据所述拍摄对象的位置坐标以及所述相对位置关系,确定其他检测对象的位置坐标。
- 根据权利要求1-9任一项所述的方法,其特征在于,还包括:从所述第二类目标图像中提取所述目标图像对象的人脸特征数据;根据人脸特征数据匹配确定所述目标图像对象所指示的检测对象的身份信息。
- 根据权利要求10所述的方法,其特征在于,在所述从所述第二类目标图像中提取所述目标图像对象的人脸特征数据之前,还包括:判断所述目标图像对象所指示的检测对象是否为人体;如果是,则执行所述从所述第二类目标图像中提取所述目标图像对象的人脸特征数据。
- 根据权利要求10所述的方法,其特征在于,所述从所述第二类图像中提取所述目标图像对象的人脸特征数据,包括:根据人脸位于人体的位置以及人脸区域占人体区域的比例,在所述第二类目标图像中确定所述目标图像对象的人脸图像区域;从所述人脸图像区域中提取人脸特征数据。
- 根据权利要求10-12任一项所述的方法,其特征在于,还包括:如果所述第二类目标图像不满足人脸特征提取条件,调整拍摄参数;在调整后的拍摄参数下进行拍摄,得到新的第二类目标图像。
- 一种对象检测终端,其特征在于,包括:处理器和存储器;所述存储器,用于存储计算机程序,所述计算机程序包括程序指令;所述处理器,用于调用所述程序指令执行:拍摄得到第一类目标图像和第二类目标图像;根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象;所述第一类目标图像是用于表示温度分布的图像,所述第二类目标图像是用于进行图像对象识别的图像;所述目标图像对象是位于所述第二类目标图像中且根据所述第一类目标图像所表示的温度分布确定的处于温度异常状态的对象。
- 根据权利要求14所述的终端,其特征在于,所述处理器,用于调用所述程序指令执行所述根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象时具体执行:对所述第二类目标图像进行对象检测处理,得到至少一个对象区域;根据所述第一类目标图像所表示的温度分布,从所述至少一个对象区域中确定出目标对象区域;从所述目标对象区域中识别出目标图像对象。
- 根据权利要求14所述的终端,其特征在于,所述处理器,用于调用所述程序指令执行所述根据所述第一类目标图像对所述第二类目标图像进行分析处理,从所述第二类目标图像中确定出目标图像对象时具体执行:根据所述第一类目标图像所表示的温度分布,从所述第二类目标图像中确定出对象检测区域;对所述对象检测区域进行对象检测处理,得到目标对象区域;从所述目标对象区域中识别出目标图像对象。
- 根据权利要求14-15任一项所述的终端,其特征在于,所述处理器,用于调用所述程序指令执行所述拍摄得到第一类目标图像和第二类目标图像时具体执行:拍摄得到待检测环境的第一类图像,并根据所述待检测环境的第一类图像确定所述待检测环境中的目标检测区域;对所述目标检测区域进行拍摄,得到第一类目标图像和第二类目标图像。
- 根据权利要求17所述的终端,其特征在于,所述处理器,用于调用所述程序指令执行所述根据所述待检测环境的第一类图像确定所述待检测环境中的目标检测区域时具体执行:根据预设温度范围对所述待检测环境的第一类图像进行分析处理,确定出目标图像区域,所述目标图像区域所对应的温度分布满足预设分区条件;根据所述目标图像区域确定所述待检测环境中的目标检测区域。
- 根据权利要求14-18任一项所述的终端,其特征在于,所述处理器,用于调用所述程序指令还执行:获取拍摄位置的位置描述信息;根据所述位置描述信息确定所述目标图像对象所指示的检测对象的位置。
- 根据权利要求19所述的终端,其特征在于,所述处理器,用于调用所述程序指令还执行:从所述检测对象中确定出拍摄对象;将当前拍摄方向调整至正对所述拍摄对象,并对所述拍摄对象进行拍摄。
- 根据权利要求20所述的终端,其特征在于,所述处理器,用于调用所述程序指令执行所述根据所述位置描述信息确定所述目标图像对象所指示的检测对象的位置时具体执行:根据拍摄高度和第一拍摄角度,得到所述拍摄对象和拍摄位置之间的距离;根据第二拍摄角度、所述拍摄对象和拍摄位置之间的距离以及所述位置描述信息所描述的位置坐标,得到所述拍摄对象的位置坐标;根据所述拍摄对象的位置坐标确定所述检测对象的位置;所述第一拍摄角度是在竖直方向的拍摄角度,所述第二拍摄角度是在水平方向的拍摄角度。
- 根据权利要求21所述的终端,其特征在于,所述检测对象包括多个;所述处理器,用于调用所述程序指令执行所述根据拍摄对象的位置坐标确定所述检测对象的位置时具体执行:根据所述第二类目标图像确定所述拍摄对象与其他检测对象之间的相对位置关系;根据所述拍摄对象的位置坐标以及所述相对位置关系,确定其他检测对象的位置坐标。
- 根据权利要求14-22任一项所述的终端,其特征在于,所述处理器,用于调用所述程序指令还执行:从所述第二类目标图像中提取所述目标图像对象的人脸特征数据;根据人脸特征数据匹配确定所述目标图像对象所指示的检测对象的身份信息。
- 根据权利要求23所述的终端,其特征在于,所述处理器,用于调用所述程序指令还执行:判断所述目标图像对象所指示的检测对象是否为人体;如果是,则执行所述从所述第二类目标图像中提取所述目标图像对象的人脸特征数据。
- 根据权利要求23所述的终端,其特征在于,所述处理器,用于调用所述程序指令执行所述从所述第二类图像中提取所述目标图像对象的人脸特征数据时具体执行:根据人脸位于人体的位置以及人脸区域占人体区域的比例,在所述第二类目标图像中确定所述目标图像对象的人脸图像区域;从所述人脸图像区域中提取人脸特征数据。
- 根据权利要求23-25任一项所述的终端,其特征在于,所述处理器,用于调用所述程序指令还执行:如果所述第二类目标图像不满足人脸特征提取条件,调整拍摄参数;在调整后的拍摄参数下进行拍摄,得到新的第二类目标图像。
- 一种计算机可读存储介质,其特征在于,所述计算机存储介质存储有 计算机程序,所述计算机程序包括程序指令,所述程序指令当被处理器调用时使所述处理器执行如权利要求1-13任一项所述的方法。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2017/104383 WO2019061293A1 (zh) | 2017-09-29 | 2017-09-29 | 对象检测方法、对象检测终端及计算机可读介质 |
CN201780003638.2A CN108700468A (zh) | 2017-09-29 | 2017-09-29 | 对象检测方法、对象检测终端及计算机可读介质 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2017/104383 WO2019061293A1 (zh) | 2017-09-29 | 2017-09-29 | 对象检测方法、对象检测终端及计算机可读介质 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2019061293A1 true WO2019061293A1 (zh) | 2019-04-04 |
Family
ID=63843741
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2017/104383 WO2019061293A1 (zh) | 2017-09-29 | 2017-09-29 | 对象检测方法、对象检测终端及计算机可读介质 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN108700468A (zh) |
WO (1) | WO2019061293A1 (zh) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110175114A (zh) * | 2019-04-19 | 2019-08-27 | 深圳壹账通智能科技有限公司 | 测试终端、app与摄像头兼容性测试方法及存储介质 |
CN111126267A (zh) * | 2019-12-24 | 2020-05-08 | 厦门瑞为信息技术有限公司 | 基于视觉感知的人脸离港信息设备 |
CN111339854A (zh) * | 2020-02-14 | 2020-06-26 | 北京猎户星空科技有限公司 | 一种测温方法、测温装置、机器人以及计算机存储介质 |
CN111598865A (zh) * | 2020-05-14 | 2020-08-28 | 上海锘科智能科技有限公司 | 基于热红外和rgb双摄的手足口病检测方法、装置及系统 |
CN111626125A (zh) * | 2020-04-26 | 2020-09-04 | 浙江大华技术股份有限公司 | 人脸温度检测的方法、系统、装置和计算机设备 |
CN111626985A (zh) * | 2020-04-20 | 2020-09-04 | 北京农业信息技术研究中心 | 基于图像融合的家禽体温检测方法及禽舍巡检系统 |
CN111738132A (zh) * | 2020-06-17 | 2020-10-02 | 银河水滴科技(北京)有限公司 | 人体温度的测量方法、装置、电子设备及可读存储介质 |
CN111968713A (zh) * | 2020-08-05 | 2020-11-20 | 七海行(深圳)科技有限公司 | 一种数据采集方法及巡检装置 |
CN112597957A (zh) * | 2020-12-30 | 2021-04-02 | 核工业北京地质研究院 | 一种用于检测公共场所人群体温异常的图像处理方法 |
CN113096077A (zh) * | 2021-03-25 | 2021-07-09 | 深圳力维智联技术有限公司 | 异常比例检测方法、装置、设备及计算机可读存储介质 |
WO2021204947A1 (en) | 2020-04-08 | 2021-10-14 | Heba Bevan | Infection and disease sensing systems |
Families Citing this family (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109657576B (zh) * | 2018-12-06 | 2023-10-31 | 联想(北京)有限公司 | 图像采集控制方法、装置、存储介质及系统 |
CN109870250A (zh) * | 2019-01-27 | 2019-06-11 | 武汉星巡智能科技有限公司 | 区域异常体温监测方法、装置及计算机可读存储介质 |
CN110222725B (zh) * | 2019-05-15 | 2022-01-04 | 广州影子科技有限公司 | 猪只核验方法、猪只核验装置与猪只核验系统 |
CN110956118B (zh) * | 2019-11-26 | 2023-04-18 | 浙江大华技术股份有限公司 | 目标对象检测方法和装置、存储介质及电子装置 |
CN111289110A (zh) * | 2020-02-14 | 2020-06-16 | 北京都是科技有限公司 | 人体温度检测方法、系统、装置及热红外图像处理器 |
CN111353438A (zh) * | 2020-03-02 | 2020-06-30 | 成都三零凯天通信实业有限公司 | 一种基于ai的公共场所应急医疗寻人方法 |
CN111027525B (zh) * | 2020-03-09 | 2020-06-30 | 中国民用航空总局第二研究所 | 疫情期间公共场所潜在被感染者追踪方法、装置及系统 |
CN113447128B (zh) | 2020-03-10 | 2023-06-23 | 百度在线网络技术(北京)有限公司 | 一种多人体温检测方法、装置、电子设备及存储介质 |
CN111432052A (zh) * | 2020-04-13 | 2020-07-17 | 北京年管家信息科技有限公司 | 一种带有温度测量和夜视仪功能的智能手机、测量方法及计算机可读存储介质 |
CN111540092A (zh) * | 2020-04-30 | 2020-08-14 | 济南浪潮高新科技投资发展有限公司 | 一种门禁闸机启停的控制方法、装置、设备及存储介质 |
CN113168530A (zh) * | 2020-06-23 | 2021-07-23 | 深圳市大疆创新科技有限公司 | 目标检测设备及方法、成像装置和可移动平台 |
CN111912545A (zh) * | 2020-08-21 | 2020-11-10 | 深圳市中易通安全芯科技有限公司 | 温感式人证核验终端使用方法 |
CN112336316B (zh) * | 2020-11-04 | 2023-03-24 | 安捷光通科技成都有限公司 | 一种多态无感测温系统及其工作方法 |
CN113516146A (zh) * | 2020-12-21 | 2021-10-19 | 腾讯科技(深圳)有限公司 | 一种数据分类方法、计算机及可读存储介质 |
CN114493291B (zh) * | 2022-01-28 | 2022-11-01 | 中铁北京工程局集团有限公司 | 一种高填方质量智能检测方法及系统 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104820967A (zh) * | 2015-05-26 | 2015-08-05 | 中国科学院自动化研究所 | 在轨计算成像方法 |
CN105005234A (zh) * | 2015-06-02 | 2015-10-28 | 哈尔滨工业大学(威海) | 基于zigbee的智能温室环境遥测与控制系统 |
CN105931254A (zh) * | 2016-05-17 | 2016-09-07 | 北京市检验检疫科学技术研究院 | 一种国境口岸出入境检验检疫低温查验方法及系统 |
CN106407900A (zh) * | 2016-08-31 | 2017-02-15 | 上海交通大学 | 基于多源航片的异常场景识别方法 |
Family Cites Families (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7457441B2 (en) * | 2005-02-25 | 2008-11-25 | Aptina Imaging Corporation | System and method for detecting thermal anomalies |
CN101771539B (zh) * | 2008-12-30 | 2012-07-04 | 北京大学 | 一种基于人脸识别的身份认证方法 |
EP2230496A1 (en) * | 2009-03-16 | 2010-09-22 | Autoliv Development AB | Method and system for automatically detecting objects in front of a motor vehicle |
CN102045423A (zh) * | 2009-10-16 | 2011-05-04 | 英华达(上海)电子有限公司 | 移动终端、使用移动终端进行温度检测的方法 |
CN103914679B (zh) * | 2013-01-05 | 2017-07-21 | 联想(北京)有限公司 | 一种图像识别方法、装置和电子设备 |
CN104980710B (zh) * | 2015-06-30 | 2018-05-04 | 上海青橙实业有限公司 | 温度检测方法及移动终端 |
CN105844240A (zh) * | 2016-03-23 | 2016-08-10 | 深圳云天励飞技术有限公司 | 一种红外测温系统中的人脸检测方法及装置 |
US9693695B1 (en) * | 2016-09-23 | 2017-07-04 | International Business Machines Corporation | Detecting oral temperature using thermal camera |
CN107172349B (zh) * | 2017-05-19 | 2020-12-04 | 崔祺 | 移动终端拍摄方法、移动终端及计算机可读存储介质 |
CN107197170A (zh) * | 2017-07-14 | 2017-09-22 | 维沃移动通信有限公司 | 一种曝光控制方法及移动终端 |
-
2017
- 2017-09-29 CN CN201780003638.2A patent/CN108700468A/zh active Pending
- 2017-09-29 WO PCT/CN2017/104383 patent/WO2019061293A1/zh active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104820967A (zh) * | 2015-05-26 | 2015-08-05 | 中国科学院自动化研究所 | 在轨计算成像方法 |
CN105005234A (zh) * | 2015-06-02 | 2015-10-28 | 哈尔滨工业大学(威海) | 基于zigbee的智能温室环境遥测与控制系统 |
CN105931254A (zh) * | 2016-05-17 | 2016-09-07 | 北京市检验检疫科学技术研究院 | 一种国境口岸出入境检验检疫低温查验方法及系统 |
CN106407900A (zh) * | 2016-08-31 | 2017-02-15 | 上海交通大学 | 基于多源航片的异常场景识别方法 |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110175114A (zh) * | 2019-04-19 | 2019-08-27 | 深圳壹账通智能科技有限公司 | 测试终端、app与摄像头兼容性测试方法及存储介质 |
CN111126267A (zh) * | 2019-12-24 | 2020-05-08 | 厦门瑞为信息技术有限公司 | 基于视觉感知的人脸离港信息设备 |
CN111339854A (zh) * | 2020-02-14 | 2020-06-26 | 北京猎户星空科技有限公司 | 一种测温方法、测温装置、机器人以及计算机存储介质 |
CN111339854B (zh) * | 2020-02-14 | 2023-06-30 | 北京猎户星空科技有限公司 | 一种测温方法、测温装置、机器人以及计算机存储介质 |
WO2021204947A1 (en) | 2020-04-08 | 2021-10-14 | Heba Bevan | Infection and disease sensing systems |
CN111626985A (zh) * | 2020-04-20 | 2020-09-04 | 北京农业信息技术研究中心 | 基于图像融合的家禽体温检测方法及禽舍巡检系统 |
CN111626125A (zh) * | 2020-04-26 | 2020-09-04 | 浙江大华技术股份有限公司 | 人脸温度检测的方法、系统、装置和计算机设备 |
CN111626125B (zh) * | 2020-04-26 | 2023-04-28 | 浙江大华技术股份有限公司 | 人脸温度检测的方法、系统、装置和计算机设备 |
CN111598865B (zh) * | 2020-05-14 | 2023-05-16 | 上海锘科智能科技有限公司 | 基于热红外和rgb双摄的手足口病检测方法、装置及系统 |
CN111598865A (zh) * | 2020-05-14 | 2020-08-28 | 上海锘科智能科技有限公司 | 基于热红外和rgb双摄的手足口病检测方法、装置及系统 |
CN111738132A (zh) * | 2020-06-17 | 2020-10-02 | 银河水滴科技(北京)有限公司 | 人体温度的测量方法、装置、电子设备及可读存储介质 |
CN111738132B (zh) * | 2020-06-17 | 2024-03-05 | 银河水滴科技(北京)有限公司 | 人体温度的测量方法、装置、电子设备及可读存储介质 |
CN111968713A (zh) * | 2020-08-05 | 2020-11-20 | 七海行(深圳)科技有限公司 | 一种数据采集方法及巡检装置 |
CN111968713B (zh) * | 2020-08-05 | 2023-10-27 | 七海行(深圳)科技有限公司 | 一种数据采集方法及巡检装置 |
CN112597957A (zh) * | 2020-12-30 | 2021-04-02 | 核工业北京地质研究院 | 一种用于检测公共场所人群体温异常的图像处理方法 |
CN113096077A (zh) * | 2021-03-25 | 2021-07-09 | 深圳力维智联技术有限公司 | 异常比例检测方法、装置、设备及计算机可读存储介质 |
CN113096077B (zh) * | 2021-03-25 | 2024-05-03 | 深圳力维智联技术有限公司 | 异常比例检测方法、装置、设备及计算机可读存储介质 |
Also Published As
Publication number | Publication date |
---|---|
CN108700468A (zh) | 2018-10-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2019061293A1 (zh) | 对象检测方法、对象检测终端及计算机可读介质 | |
KR102021999B1 (ko) | 인체 감시 발열 경보 장치 | |
US11030464B2 (en) | Privacy processing based on person region depth | |
CN109002786B (zh) | 人脸检测方法、设备及计算机可读存储介质 | |
WO2015165365A1 (zh) | 一种人脸识别方法及系统 | |
CN111916219B (zh) | 检验检疫智能安全预警方法、装置及电子系统 | |
US20150269437A1 (en) | Image processing apparatus, method thereof and program therefor | |
US10433119B2 (en) | Position determination device, position determining method, and storage medium | |
WO2020173258A1 (zh) | 图像识别系统及其方法 | |
US11983951B2 (en) | Human detection device and human detection method | |
US10496874B2 (en) | Facial detection device, facial detection system provided with same, and facial detection method | |
US20220198803A1 (en) | Person detection device and person detection method | |
WO2018149322A1 (zh) | 图像识别方法、装置、设备及存储介质 | |
CN113837030A (zh) | 用于疫情防控的人员智能管控方法、系统及计算机设备 | |
KR20220056279A (ko) | 인공 지능 기반 비전 감시 장치 | |
Ma et al. | A cloud-edge-terminal collaborative system for temperature measurement in COVID-19 prevention | |
US11256910B2 (en) | Method and system for locating an occupant | |
CN108234932B (zh) | 视频监控图像中的人员形态提取方法及装置 | |
JP2017054229A (ja) | 人物判別システム | |
US20190122228A1 (en) | Examination device | |
Al Maashri et al. | A novel drone-based system for accurate human temperature measurement and disease symptoms detection using thermography and AI | |
JPWO2021131050A5 (zh) | ||
WO2021140844A1 (ja) | 人体検出装置および人体検出方法 | |
US11481996B2 (en) | Calculation device, information processing method, and storage medium | |
TWM606799U (zh) | 體溫監控裝置 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 17927606 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 17927606 Country of ref document: EP Kind code of ref document: A1 |