CN110244314B - Low-slow small target detection and identification system and method - Google Patents
Low-slow small target detection and identification system and method Download PDFInfo
- Publication number
- CN110244314B CN110244314B CN201910548691.1A CN201910548691A CN110244314B CN 110244314 B CN110244314 B CN 110244314B CN 201910548691 A CN201910548691 A CN 201910548691A CN 110244314 B CN110244314 B CN 110244314B
- Authority
- CN
- China
- Prior art keywords
- target
- information
- image
- visible light
- suspicious
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/02—Systems using the reflection of electromagnetic waves other than radio waves
- G01S17/06—Systems determining position data of a target
- G01S17/42—Simultaneous measurement of distance and other co-ordinates
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/02—Systems using the reflection of electromagnetic waves other than radio waves
- G01S17/50—Systems of measurement based on relative movement of target
- G01S17/58—Velocity or trajectory determination systems; Sense-of-movement determination systems
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/66—Tracking systems using electromagnetic waves other than radio waves
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/88—Lidar systems specially adapted for specific applications
- G01S17/89—Lidar systems specially adapted for specific applications for mapping or imaging
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Closed-Circuit Television Systems (AREA)
- Studio Devices (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a low-slow small target detection and identification system and a low-slow small target detection and identification method, belongs to the technical field of target detection, and solves the problem of detection and identification of a low-slow small unmanned aerial vehicle target. The system comprises a panoramic staring unit, a panoramic staring unit and a panoramic vision unit, wherein the panoramic staring unit is used for early warning and detecting a full-view target; the target information sensing unit is used for sensing distance information, visible light image information and infrared image information of a suspicious target; the image comprehensive processing box is used for searching a target area according to the early warning detection information and calculating azimuth angle and pitch angle information of a suspicious target; controlling the target information sensing unit to aim at a suspicious target in an early warning area to sense target information; and according to the received perception result information, extracting and comprehensively identifying the characteristic information of the suspicious target, and determining the type of the suspicious target. The invention can alarm the low-slow small unmanned aerial vehicle target with the distance not more than 2km and the wingspan not less than 0.6m multiplied by 0.6m in the air space covered by 360 degrees multiplied by 30 degrees in the urban complex environment.
Description
Technical Field
The invention relates to the technical field of target detection, in particular to a low-speed and small-target detection and identification system and a method.
Background
The radar scattering area of the low-slow small unmanned aerial vehicle target is small, the flying height is low, the speed is low, the Doppler frequency shift is not obvious, and in addition, the use scene is complex, the environmental interference is more, so that the detection and the identification of the low-slow small unmanned aerial vehicle target become a worldwide difficult problem. At present, the photoelectric means for detecting and identifying the target of the low-slow small unmanned aerial vehicle mainly comprises detection technologies such as laser ranging, visible light imaging or array type camera imaging, infrared imaging and the like. The characteristics of the detection means are different, the detection means have long characteristics in the aspects of distance finding, sensitivity, energy efficiency range, severe weather resistance, simultaneous multi-target tracking capability and the like, but the problem of low-speed small unmanned aerial vehicle target detection and identification in urban complex environment is still difficult to effectively solve.
Disclosure of Invention
In view of the above analysis, the present invention aims to provide a system and a method for detecting and identifying a low-slow small target, so as to solve the problem of detecting and identifying a low-slow small unmanned aerial vehicle target in a complex urban environment.
The purpose of the invention is mainly realized by the following technical scheme:
a low-slow small target detection and recognition system comprises a panoramic staring unit (1), a target information sensing unit (2) and an image comprehensive processing box (3);
the panoramic staring unit (1) is used for carrying out full-view target early warning detection on a detection task area; sending the early warning detection information to an image comprehensive processing box (3);
the target information sensing unit (2) is used for sensing distance information, visible light image information and infrared image information of a suspicious target; and sending the perception result information to an image comprehensive processing box (3);
the image comprehensive processing box (3) is used for searching a suspicious target according to the early warning detection information and calculating azimuth angle and pitch angle information of the suspicious target; controlling the target information sensing unit (2) to align to a suspicious target in an early warning area according to the azimuth angle and the pitch angle information, and sensing target information; and according to the received perception result information, extracting and comprehensively identifying the characteristic information of the suspicious target, and determining the type of the suspicious target.
Further, the panoramic staring unit (1) comprises a plurality of early warning subunits, each early warning subunit corresponds to one early warning sub-area, and the early warning sub-areas are spliced to form a 360-degree full-view-angle early warning area;
the early warning subunit adopts a high-resolution high-frame frequency visible light detection component to detect early warning image information.
Further, the target information sensing unit (2) comprises a laser ranging sensor (201), a visible light image sensor (202), an infrared image sensor (203) and an optical-mechanical platform (204);
the laser ranging sensor (201) is used for performing laser ranging on the suspicious target through laser and sending measured target distance information to the image comprehensive processing box (3);
the visible light image sensor (202) is used for carrying out visible light imaging on the suspicious target and sending visible light image information to the image comprehensive processing box (3);
the infrared image sensor (203) is used for carrying out infrared imaging on a target in the early warning space region and sending infrared image information to the image comprehensive processing box (3);
the optical mechanical platform (204) is used for bearing the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203); and driving the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203) to align to a target area according to the received photoelectric detection instruction sent by the image comprehensive processing box (3).
Further, the image comprehensive processing box (3) comprises an early warning image processing unit (301), a holder control unit (302), a comprehensive identification unit (304) and a target information output unit (305);
the early warning image processing unit (301) is used for processing the early warning detection information uploaded by the panoramic gaze fixation unit (1) to obtain images, searching suspicious targets and determining azimuth angle and pitch angle information of the suspicious targets;
the holder control unit (302) is used for generating a photoelectric detection instruction according to the azimuth angle and pitch angle information of the suspicious target, and sending the photoelectric detection instruction to the optical mechanical platform (204), so that the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203) borne by the optical mechanical platform (204) point to the direction of the suspicious target;
a comprehensive identification unit (303) for sensing the suspicious target distance information and the visible light image information sensed by the target information sensing unit (2)
Or extracting suspicious target characteristic information from the infrared image information, and performing comprehensive identification according to the suspicious target characteristic information to determine the target type.
Further, the image comprehensive processing box (3) further comprises a perception control unit (304) for controlling the visible light image sensor (202) and the infrared image sensor (203) to acquire images according to the distance of the suspicious target;
when the distance of the suspicious target is within a first preset distance range, starting a visible light image sensor (202) to acquire a visible light image; and when the target distance information is within a second preset distance range, simultaneously starting the visible light sensor (202) and the infrared image sensor (203), and simultaneously acquiring a visible light image and an infrared image.
An object detection and identification method adopting the object detection and identification system is characterized by comprising the following steps:
step S1, the system is powered on, and the panoramic staring unit (1) collects the panoramic image of the detection task area in real time; the image comprehensive processing box (3) processes the acquired image and searches whether a suspicious target exists;
step S2, after searching the suspicious target, the image comprehensive processing box (3) calculates the azimuth angle and the pitch angle information of the suspicious target, converts the azimuth angle and the pitch angle information into a photoelectric detection instruction signal and outputs the photoelectric detection instruction signal to the optical-mechanical platform (204), and moves the optical-mechanical platform (204) to drive the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203) to point to the direction of the suspicious target;
step S3, the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (202) detect the possible target and sense the distance information, the visible light image information and the infrared image information of the suspicious target;
and step S4, extracting suspicious target characteristic information according to the sensed suspicious target distance information, visible light image information or infrared image information, and performing comprehensive identification according to the suspicious target characteristic information to determine the target type.
Further, in step S1, the image integration processing box (3) processes the acquired image by using a frame difference method to determine whether there is a suspicious object.
Further, in step S3, including,
the laser ranging sensor (201) carries out laser ranging on the target to obtain distance information of the suspicious target, and the distance information is sent to the image comprehensive processing box (3);
the image comprehensive processing box (3) judges the distance information, and when the target distance information is within a first preset distance range, a visible light image sensor (202) is started to collect a visible light image; and when the target distance information is within a second preset distance range, simultaneously starting the visible light sensor (202) and the infrared image sensor (203) to acquire a visible light image and an infrared image.
Further, in step S4, the suspicious target feature information includes target contour information, and the target contour information is obtained by performing median filtering, binarization, edge detection, and image morphology processing on a visible light image or an infrared image of the suspicious target.
Further, in step S4, a Support Vector Machine (SVM) method based on shallow learning is used to perform comprehensive recognition, so as to obtain the target type.
The invention has the following beneficial effects:
the invention provides a low-slow small target detection and identification system and a method, which are characterized in that a panoramic staring unit is used for alarming a low-slow small unmanned aerial vehicle target covering the space within a range of 360 degrees multiplied by 30 degrees, the distance is not more than 2km and the wingspan is not less than 0.6m multiplied by 0.6m, so that the information of a target azimuth angle, a pitch angle and the like is obtained, and a panoramic spliced image of a detection task area can be output in real time for displaying the situation of the detection task area; the alarm target is accurately detected through the photoelectric tracking unit, the characteristics of the target such as form, distance, speed, flight path and the like are identified and acquired, and tracking monitoring is carried out.
Drawings
The drawings are only for purposes of illustrating particular embodiments and are not to be construed as limiting the invention, wherein like reference numerals are used to designate like parts throughout.
FIG. 1 is a schematic diagram of the connection of the "low-slow small" target detection and identification system in the embodiment of the present invention;
FIG. 2 is a schematic diagram of a panoramic gaze unit in an example of the invention;
FIG. 3 is a schematic view of an optoelectronic turret in an embodiment of the invention
Fig. 4 is a flowchart of a target detection and identification method according to an embodiment of the present invention.
Detailed Description
The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form a part hereof, and which together with the embodiments of the invention serve to explain the principles of the invention.
The embodiment discloses a low and slow small target detection and identification system, which detects and identifies a low and slow small unmanned aerial vehicle target in a complex urban environment by integrating visible light, infrared light and laser light, and realizes early warning and monitoring of the low and slow small unmanned aerial vehicle target in a prevention and control task area.
As shown in fig. 1, the low-slow small target detection and recognition system of the present embodiment comprises a panoramic gaze fixation unit (1), a target information perception unit (2), and an image synthesis processing box (3);
the panoramic staring unit (1) is used for carrying out full-view target early warning detection on a detection task area; sending the early warning detection information to an image comprehensive processing box (3);
the target information sensing unit (2) is used for sensing distance information, visible light image information and infrared image information of a suspicious target; and sending the perception result information to an image comprehensive processing box (3);
the image comprehensive processing box (3) is used for searching a suspicious target according to the early warning detection information and calculating azimuth angle and pitch angle information of the suspicious target; controlling the target information sensing unit (2) to align to a suspicious target in an early warning area according to the azimuth angle and the pitch angle information, and sensing target information; and according to the received perception result information, extracting and comprehensively identifying the characteristic information of the suspicious target, and determining the type of the suspicious target.
Specifically, the panoramic staring unit (1) comprises a plurality of early warning subunits, each early warning subunit corresponds to one early warning sub-area, and the early warning sub-areas are spliced to form a 360-degree full-view-angle early warning area;
preferably, as shown in fig. 2, the panoramic gaze fixation unit (1) includes 11 early warning subunits, each early warning subunit is designed by adopting a modular component, each module is rationally distributed, and after correction, a spatial region covering a range of 360 ° × 30 ° (a single camera covers 44.2 ° × 30 °) is spliced;
the early warning subunit adopts a high-resolution high-frame frequency visible light detection component, the data rate is not lower than 25Hz, the early warning subunit comprises a lens, a detector, an imaging circuit, a video and communication processing circuit, an optical platform and the like, and the optical lens is preferably LEM1616 CE-H1; and a Mars5000-60UM model array image device is matched.
As shown in fig. 3, the target information sensing unit (2) includes a laser ranging sensor (201), a visible light image sensor (202), an infrared image sensor (203), and an optical bench (204);
the laser ranging sensor (201) is used for performing laser ranging on the suspicious target through laser and sending measured target distance information to the image comprehensive processing box (3);
specifically, the laser ranging sensor selects 1.064um air-cooled high repetition frequency semiconductor pump solid Nd with high peak power, high pulse repetition frequency and within a safe wavelength range: YAG laser or fiber laser, which adopts pulse laser working mode to obtain the distance information of the detected target.
The visible light image sensor (202) is used for carrying out visible light imaging on the suspicious target and sending visible light image information to the image comprehensive processing box (3);
specifically, the visible light image sensor is a CMOS sensor, a high-resolution digital image sensing device with a certain response in a near-infrared band and a FLIR model of CM3-U3-50S5M-CS is selected, a JZ10100R MP electric zoom lens is matched, the focal length is 10-1000mm, and the visible light image sensor 3 can detect a low-slow-small unmanned aerial vehicle target with a span of 50-2000m and larger than or equal to 0.6.
The infrared image sensor (203) is used for carrying out infrared imaging on a target in the early warning space region and sending infrared image information to the image comprehensive processing box (3);
specifically, the infrared image sensor uses CUBE817, and the detection of a low-slow small unmanned aerial vehicle target with a wingspan of 50-500m and larger than or equal to 0.6 can be realized.
The optical mechanical platform (204) is used for bearing the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203); driving the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203) to be aligned to a target area according to the received photoelectric detection instruction sent by the image comprehensive processing box (3);
in order to ensure the consistency of detection optical axes, the laser distance measuring sensor (201), the visible light image sensor (202) and the infrared image sensor (203) adopt a three-in-one coaxial installation mode.
The image comprehensive processing box (3) comprises an early warning image processing unit (301), a holder control unit (302), a comprehensive identification unit (304) and a target information output unit (305);
the early warning image processing unit (301) is used for processing the early warning detection information uploaded by the panoramic gaze fixation unit (1) to obtain images, searching suspicious targets and determining azimuth angle and pitch angle information of the suspicious targets;
specifically, the early warning image processing unit (301) receives detection task area image information uploaded by the panoramic gaze fixation unit (1) in real time, and 11 images uploaded by the panoramic gaze fixation unit (1) are spliced by using a Jetson TX2Arm processing board to form a detection task area situation map. When a low-slow small target enters a detection task area, the early warning image processing unit (301) obtains the serial number of an optical lens for capturing an image, processes the image by using a frame difference method, and calculates azimuth angle information and pitch angle information of the target according to the position of the target in the image and the serial number of the optical lens for capturing the image. And processing the image uploaded by the panoramic staring unit (1) and extracting the azimuth angle and pitch angle information of the suspicious target.
The holder control unit (302) is used for generating a photoelectric detection instruction according to the azimuth angle and pitch angle information of the suspicious target, and sending the photoelectric detection instruction to the optical mechanical platform (204), so that the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203) borne by the optical mechanical platform (204) point to the direction of the suspicious target;
and the comprehensive identification unit (303) is used for extracting suspicious target characteristic information from the suspicious target distance information, the visible light image information or the infrared image information sensed by the target information sensing unit (2), and performing comprehensive identification according to the suspicious target characteristic information to determine the target type.
The perception control unit (304) is used for controlling the visible light image sensor (202) and the infrared image sensor (203) to acquire images according to the distance of the suspicious target;
when the distance of the suspicious target is within a first distance preset range, starting a visible light image sensor (202) to acquire a visible light image; and when the target distance information is within a second distance preset range, simultaneously starting the visible light sensor (202) and the infrared image sensor (203), and simultaneously acquiring a visible light image and an infrared image.
Wherein the first distance preset range is 500-2000 m; the second distance preset range is less than 500 m.
Under the condition of good visibility, preferably adopting visible light image information to extract suspicious target characteristic information in the range of 0-2000 m;
under the condition of poor visibility, the visible light image sensor (202) is influenced, and the suspicious target characteristic information is preferably extracted by adopting visible light image information within the range of less than 500 m.
More specifically, the image comprehensive processing box (3) comprises an image processing circuit based on a Jetson TX2 platform, an image data acquisition and storage board, a main control board, a pan-tilt control system and the like, and is used for placing circuit boards and electrical components for data acquisition, image processing, target detection, data communication, secondary power supply and the like. The device is connected with a panoramic staring unit (1), a laser ranging sensor (201), a visible light image sensor (202) and an infrared image sensor (203) through a connector, a cable and the like, so that the work of image processing, image splicing, target form detection, target characteristic data processing, distance measurement, target accurate tracking and the like is completed.
The embodiment also discloses a low-slow small target detection and identification method, as shown in fig. 4, the specific steps are as follows:
step S1, the system is powered on, and the panoramic staring unit (1) collects the panoramic image of the detection task area in real time; the image comprehensive processing box (3) processes the acquired image and searches whether a suspicious target exists;
the laser ranging sensor (201), the visible light image sensor (202), the infrared image sensor (203) and the photoelectric turntable (204) are in a standby state.
Specifically, in step S1, the image integration processing box (3) processes the acquired image by using a frame difference method to determine whether there is a suspicious object.
Step S2, after searching the suspicious target, the image comprehensive processing box (3) calculates the azimuth angle and the pitch angle information of the suspicious target, converts the azimuth angle and the pitch angle information into a photoelectric detection instruction signal and outputs the photoelectric detection instruction signal to the optical-mechanical platform (204), and moves the optical-mechanical platform (204) to point to the direction of the suspicious target;
step S3, the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (202) detect the possible target and sense the distance information, the visible light image information and the infrared image information of the suspicious target;
specifically, step S3 includes the following substeps:
s301, carrying out laser ranging on the target by a laser ranging sensor (201) to obtain distance information of the suspicious target;
step S302, the laser ranging sensor (201) sends the distance information of the suspicious target to the image comprehensive processing box (3);
step S303, the image comprehensive processing box (3) judges the distance information, and when the target distance information is within a first distance preset range, a visible light image sensor (202) is started to collect a visible light image; and when the target distance information is within a second distance preset range, simultaneously starting the visible light sensor (202) and the infrared image sensor (203) to acquire a visible light image and an infrared image.
Wherein the first distance preset range is 500-2000 m; the second distance preset range is less than 500 m.
And step S4, extracting suspicious target characteristic information according to the sensed suspicious target distance information, visible light image information or infrared image information, and performing comprehensive identification according to the suspicious target characteristic information to determine the target type.
Specifically, in step S4, the suspicious target feature information includes target contour information, and the target contour information is obtained by performing median filtering, binarization, edge detection, and image morphology processing on a visible light image or an infrared image of the suspicious target.
Under the condition of good visibility, preferably adopting visible light image information to extract suspicious target characteristic information in the range of 0-2000 m;
under the condition of poor visibility, the visible light image sensor (202) is influenced, and the suspicious target characteristic information is preferably extracted by adopting visible light image information within the range of less than 500 m.
In step S4, comprehensive recognition is performed by using a Support Vector Machine (SVM) method based on shallow learning, and a target type is obtained.
More specifically, when a visible light image sensor and an infrared image sensor receive an image acquisition command, image acquisition is carried out on a suspicious target, and image information is transmitted to an image comprehensive processing box, the image comprehensive processing box is used for processing a target image so as to extract characteristic information of the target and comprehensively identify the target, and in order to extract the target image information, image preprocessing is required to be carried out on the image target, wherein the image preprocessing comprises a series of algorithms of image deblurring and denoising by using median filtering, image target binarization, edge detection by using a Canny operator, image morphological processing and the like, so that target contour information is obtained;
the obtained target contour information comprises the characteristics of contour circularity, rectangularity, Hu invariant moment and affine invariant moment; carrying out data normalization by using a maximum and minimum normalization method, and selecting characteristic data with the characteristic data contribution degree of more than 10% to the result by using a principal component analysis method;
preferably, a Support Vector Machine (SVM) method based on shallow learning is used for suspicious target identification, feature data with large contribution degree in data fusion are extracted to serve as positive samples, negative samples are feature data extracted by pseudo targets, an SVM classification model is trained, whether a group of data in a test set is a suspicious 'low-slow small' target or not is finally identified, multiple groups of feature data are obtained in continuous frame images, and a target type is obtained according to the multiple groups of data; the pixel coordinates of a target in the panoramic image in the panoramic staring unit are combined with the distance information to obtain the flight speed, the flight height and the flight track information of the target, the visible light image and the infrared image are subjected to image preprocessing to obtain the outline information of the target image, and the target type information is obtained by using a target identification method.
Therefore, the detection and identification of the low-slow small unmanned aerial vehicle target by the photoelectric detection system are realized.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention.
Claims (2)
1. A low-slow small target detection and recognition system is characterized by comprising a panoramic staring unit (1), a target information sensing unit (2) and an image comprehensive processing box (3);
the panoramic staring unit (1) is used for carrying out full-view target early warning detection on a detection task area; sending the early warning detection information to an image comprehensive processing box (3);
the panoramic staring unit (1) comprises 11 early warning subunits, each early warning subunit corresponds to one early warning sub-region of 44.2 degrees multiplied by 30 degrees, and the 11 early warning sub-regions are spliced to form a full-view-angle early warning region of 360 degrees multiplied by 30 degrees after layout correction; the early warning subunit adopts high-resolution high-frame frequency visible light detection components, and the optical lens of each visible light detection component is coded with a serial number;
the target information sensing unit (2) is used for sensing distance information, visible light image information and infrared image information of a suspicious target; and sending the perception result information to an image comprehensive processing box (3);
the image comprehensive processing box (3) is used for searching a suspicious target according to the early warning detection information and calculating azimuth angle and pitch angle information of the suspicious target; controlling the target information sensing unit (2) to align to a suspicious target in an early warning area according to the azimuth angle and the pitch angle information, and sensing target information; according to the received sensing result information, extracting and comprehensively identifying the characteristic information of the suspicious target, and determining the type of the suspicious target;
the target information sensing unit (2) comprises a laser ranging sensor (201), a visible light image sensor (202), an infrared image sensor (203) and an optical machine platform (204); the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203) adopt a three-in-one coaxial mounting mode;
the laser ranging sensor (201) is used for performing laser ranging on the suspicious target through laser and sending measured target distance information to the image comprehensive processing box (3);
the visible light image sensor (202) is used for carrying out visible light imaging on the suspicious target and sending visible light image information to the image comprehensive processing box (3);
the infrared image sensor (203) is used for carrying out infrared imaging on a target in the early warning space region and sending infrared image information to the image comprehensive processing box (3);
the optical mechanical platform (204) is used for bearing the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203); driving the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203) to be aligned to a target area according to the received photoelectric detection instruction sent by the image comprehensive processing box (3);
the image comprehensive processing box (3) comprises an early warning image processing unit (301), a holder control unit (302), a comprehensive identification unit (303), a perception control unit (304) and a target information output unit (305);
the early warning image processing unit (301) is used for processing the early warning detection information uploaded by the panoramic gaze fixation unit (1) to obtain images, searching suspicious targets and determining azimuth angle and pitch angle information of the suspicious targets; when a suspicious object is searched, the laser ranging sensor (201), the visible light image sensor (202), the infrared image sensor (203) and the photoelectric turntable (204) are in a standby state;
when a low-slow small target enters a detection task area, the early warning image processing unit (301) obtains the serial number of an optical lens for capturing an image, processes the image by using a frame difference method, and calculates azimuth angle information and pitch angle information of the target according to the position of the target in the image and the serial number of the optical lens for capturing the image;
the holder control unit (302) is used for generating a photoelectric detection instruction according to the azimuth angle and pitch angle information of the suspicious target, and sending the photoelectric detection instruction to the optical mechanical platform (204), so that the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203) borne by the optical mechanical platform (204) point to the direction of the suspicious target;
the comprehensive identification unit (303) is used for extracting suspicious target characteristic information from the suspicious target distance information, the visible light image information or the infrared image information sensed by the target information sensing unit (2), and performing comprehensive identification according to the suspicious target characteristic information to determine the target type;
the perception control unit (304) is used for controlling the visible light image sensor (202) and the infrared image sensor (203) to acquire images according to the distance of the suspicious target measured by the laser ranging sensor (201);
when the distance of the suspicious target is within a first preset distance range, starting a visible light image sensor (202) to acquire a visible light image; when the target distance information is within a second preset distance range, simultaneously starting a visible light sensor (202) and an infrared image sensor (203), and simultaneously acquiring a visible light image and an infrared image;
wherein the first distance preset range is 500-2000 m; the second distance preset range is less than 500 m;
extracting suspicious target characteristic information by adopting visible light image information in a range of 0-2000m under the condition of good visibility;
and under the condition of poor visibility, extracting the characteristic information of the suspicious target by adopting the visible light image information within the range of less than 500 m.
2. An object detection and recognition method using the low-slow small-object detection and recognition system according to claim 1, comprising the steps of:
step S1, the system is powered on, and the panoramic staring unit (1) collects the panoramic image of the detection task area in real time; the image comprehensive processing box (3) processes the acquired image and searches whether a suspicious target exists;
step S2, after searching the suspicious target, the image comprehensive processing box (3) calculates the azimuth angle and the pitch angle information of the suspicious target, converts the azimuth angle and the pitch angle information into a photoelectric detection instruction signal and outputs the photoelectric detection instruction signal to the optical-mechanical platform (204), and moves the optical-mechanical platform (204) to drive the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (203) to point to the direction of the suspicious target;
step S3, detecting a suspicious target by the laser ranging sensor (201), the visible light image sensor (202) and the infrared image sensor (202), and sensing distance information, visible light image information and infrared image information of the suspicious target;
step S4, extracting suspicious target characteristic information according to the sensed suspicious target distance information, visible light image information or infrared image information, and performing comprehensive identification according to the suspicious target characteristic information to determine the target type;
in the step S1, the image comprehensive processing box (3) processes the acquired image by using a frame difference method to determine whether there is a suspicious target;
the step S3 includes the steps of,
the laser ranging sensor (201) carries out laser ranging on the target to obtain distance information of the suspicious target, and the distance information is sent to the image comprehensive processing box (3);
the image comprehensive processing box (3) judges the distance information, and when the target distance information is within a first preset distance range, a visible light image sensor (202) is started to collect a visible light image; when the target distance information is within a second preset distance range, simultaneously starting a visible light sensor (202) and an infrared image sensor (203) to acquire a visible light image and an infrared image;
in step S4, the suspicious target feature information includes target contour information, and the target contour information is obtained by performing median filtering, binarization, edge detection, and image morphology processing on a visible light image or an infrared image of the suspicious target;
in step S4, comprehensive recognition is performed using a support vector machine method based on shallow learning to obtain a target type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910548691.1A CN110244314B (en) | 2019-06-24 | 2019-06-24 | Low-slow small target detection and identification system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910548691.1A CN110244314B (en) | 2019-06-24 | 2019-06-24 | Low-slow small target detection and identification system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110244314A CN110244314A (en) | 2019-09-17 |
| CN110244314B true CN110244314B (en) | 2022-04-19 |
Family
ID=67889092
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910548691.1A Active CN110244314B (en) | 2019-06-24 | 2019-06-24 | Low-slow small target detection and identification system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110244314B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111208581B (en) * | 2019-12-16 | 2022-07-05 | 长春理工大学 | Unmanned aerial vehicle multi-dimensional identification system and method |
| CN111077533A (en) * | 2019-12-27 | 2020-04-28 | 湖南傲英创视信息科技有限公司 | Multispectral wide-area panoramic photoelectric radar system and detection method thereof |
| CN111190156B (en) * | 2020-01-08 | 2022-04-22 | 中国船舶重工集团公司第七二四研究所 | Radar and photoelectric based low-slow small target and sea surface small target identification method |
| CN111223122A (en) * | 2020-01-20 | 2020-06-02 | 四川九洲电器集团有限责任公司 | Low-slow small target rapid detection and tracking device based on cloud |
| CN111325772A (en) * | 2020-02-19 | 2020-06-23 | 四川九洲电器集团有限责任公司 | Low-slow small target detection device and method based on photoelectric equipment array |
| CN111709385B (en) * | 2020-06-22 | 2023-07-25 | 电子科技大学 | Low-altitude and slow-altitude aircraft target detection system and identity recognition method thereof |
| CN114518582A (en) * | 2020-11-18 | 2022-05-20 | 北京机械设备研究所 | Target detection device and method based on panoramic detection and photoelectric tracking |
| CN112584015B (en) * | 2020-12-02 | 2022-05-17 | 达闼机器人股份有限公司 | Object detection method, device, storage medium and electronic equipment |
| CN113655488A (en) * | 2021-06-30 | 2021-11-16 | 武汉高德红外股份有限公司 | Infrared alarm system for rapid periodic scanning directional distance measurement and distance measurement method thereof |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108170139A (en) * | 2017-12-12 | 2018-06-15 | 武汉华之洋科技有限公司 | A kind of photoelectricity multitask system for unmanned boat and perform method |
| CN109737981A (en) * | 2019-01-11 | 2019-05-10 | 西安电子科技大学 | Unmanned vehicle target-seeking device and method based on multisensor |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201754214U (en) * | 2010-07-20 | 2011-03-02 | 北京四维图新科技股份有限公司 | Panorama shooting device |
| CN102890823B (en) * | 2011-07-19 | 2016-04-27 | 株式会社东芝 | Motion object outline is extracted and left ventricle image partition method and device |
| CN104113733B (en) * | 2014-07-25 | 2017-10-27 | 北京机械设备研究所 | A kind of low slow Small object TV reconnaissance probe method |
| US11022407B2 (en) * | 2015-12-15 | 2021-06-01 | Tradewinds Technology, Llc | UAV defense system |
| CN106341206A (en) * | 2016-08-26 | 2017-01-18 | 广东容祺智能科技有限公司 | Low-altitude unmanned aerial vehicle defense system |
| CN108761403A (en) * | 2018-05-31 | 2018-11-06 | 智飞智能装备科技东台有限公司 | A kind of anti-unmanned plane system of defense based on radar |
| CN109461168B (en) * | 2018-10-15 | 2021-03-16 | 腾讯科技（深圳）有限公司 | Target object identification method and device, storage medium and electronic device |
| CN109873669A (en) * | 2019-04-04 | 2019-06-11 | 北京环境特性研究所 | A kind of unmanned plane detection method and unmanned plane detection system |
-
2019
- 2019-06-24 CN CN201910548691.1A patent/CN110244314B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108170139A (en) * | 2017-12-12 | 2018-06-15 | 武汉华之洋科技有限公司 | A kind of photoelectricity multitask system for unmanned boat and perform method |
| CN109737981A (en) * | 2019-01-11 | 2019-05-10 | 西安电子科技大学 | Unmanned vehicle target-seeking device and method based on multisensor |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110244314A (en) | 2019-09-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110244314B (en) | Low-slow small target detection and identification system and method | |
| US11017228B2 (en) | Method and arrangement for condition monitoring of an installation with operating means | |
| CN102435173B (en) | System and method for quickly inspecting tunnel disease based on machine vision | |
| CN109737981B (en) | Unmanned vehicle target searching device and method based on multiple sensors | |
| CN110132226B (en) | System and method for measuring distance and azimuth angle of unmanned aerial vehicle line patrol | |
| CN111324143A (en) | Unmanned aerial vehicle autonomous patrol obstacle avoidance system, method and computer equipment | |
| CN109873669A (en) | A kind of unmanned plane detection method and unmanned plane detection system | |
| KR101633812B1 (en) | The realtime remote monitoring system of buoys on the sea | |
| CN113780246B (en) | Unmanned aerial vehicle three-dimensional track monitoring method and system and three-dimensional monitoring device | |
| CN108802742A (en) | Abnormal object monitoring method, apparatus and system | |
| CN114241177A (en) | Airport pavement apparent image detection system based on linear array scanning imaging | |
| WO2017047873A1 (en) | Scanning method using high-speed scanning apparatus | |
| CN106091942A (en) | A kind of electric force pole tower range-measurement system and safe ranging method thereof | |
| CN109708659B (en) | Distributed intelligent photoelectric low-altitude protection system | |
| CN113126649A (en) | Control system for intelligent patrol inspection unmanned aerial vehicle of power transmission line | |
| EP4296973A1 (en) | System and method for localization of anomalous phenomena in assets | |
| KR102078877B1 (en) | Photovoltaic module thermal imaging system with trio imaging device | |
| CN111063148A (en) | Remote night vision target video detection method | |
| CN111736140B (en) | Object detection method and image pickup device | |
| CN205642314U (en) | Contact wire detection device based on embedded binocular vision technique | |
| CN115035470A (en) | Low, small and slow target identification and positioning method and system based on mixed vision | |
| KR101755328B1 (en) | portable speed enforcement apparatus using dual laser technology | |
| CN208969234U (en) | A kind of multispectral radar photoelectric monitoring system | |
| CN106546171A (en) | A kind of bullet train contact wire real-time detection apparatus based on machine vision | |
| CN218332632U (en) | Railway foreign matter monitoring system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |