WO2016106956A1 - 一种红外图谱关联智能探测方法及装置 - Google Patents

一种红外图谱关联智能探测方法及装置 Download PDF

Info

Publication number
WO2016106956A1
WO2016106956A1 PCT/CN2015/072678 CN2015072678W WO2016106956A1 WO 2016106956 A1 WO2016106956 A1 WO 2016106956A1 CN 2015072678 W CN2015072678 W CN 2015072678W WO 2016106956 A1 WO2016106956 A1 WO 2016106956A1
Authority
WO
WIPO (PCT)
Prior art keywords
infrared
target
spectrum
long
image
Prior art date
Application number
PCT/CN2015/072678
Other languages
English (en)
French (fr)
Inventor
张天序
刘祥燕
戴小兵
刘立
喻洪涛
Original Assignee
华中科技大学
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 华中科技大学 filed Critical 华中科技大学
Priority to US15/104,921 priority Critical patent/US9759835B2/en
Publication of WO2016106956A1 publication Critical patent/WO2016106956A1/zh

Links

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
    • G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/28—Investigating the spectrum
    • G01J3/45—Interferometric spectrometry
    • 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
    • G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/02—Details
    • G01J3/0205—Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows
    • G01J3/0208—Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows using focussing or collimating elements, e.g. lenses or mirrors; performing aberration correction
    • 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
    • G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/02—Details
    • G01J3/0205—Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows
    • G01J3/021—Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows using plane or convex mirrors, parallel phase plates, or particular reflectors
    • 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
    • G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/02—Details
    • G01J3/0294—Multi-channel spectroscopy
    • 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
    • G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/28—Investigating the spectrum
    • G01J3/2823—Imaging spectrometer
    • 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
    • G01J5/10—Radiation pyrometry, e.g. infrared or optical thermometry using electric radiation detectors
    • G01J5/20—Radiation pyrometry, e.g. infrared or optical thermometry using electric radiation detectors using resistors, thermistors or semiconductors sensitive to radiation, e.g. photoconductive devices
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
    • G01V8/00—Prospecting or detecting by optical means
    • G01V8/10—Detecting, e.g. by using light barriers
    • G—PHYSICS
    • G02—OPTICS
    • G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
    • G02B13/00—Optical objectives specially designed for the purposes specified below
    • G02B13/14—Optical objectives specially designed for the purposes specified below for use with infrared or ultraviolet radiation
    • G—PHYSICS
    • G02—OPTICS
    • G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
    • G02B19/00—Condensers, e.g. light collectors or similar non-imaging optics
    • G02B19/0033—Condensers, e.g. light collectors or similar non-imaging optics characterised by the use
    • G02B19/009—Condensers, e.g. light collectors or similar non-imaging optics characterised by the use for use with infrared radiation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/51—Indexing; Data structures therefor; Storage structures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G06F16/5854—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using shape and object relationship
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/10—Image acquisition
    • G06V10/12—Details of acquisition arrangements; Constructional details thereof
    • G06V10/14—Optical characteristics of the device performing the acquisition or on the illumination arrangements
    • G06V10/143—Sensing or illuminating at different wavelengths
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/24—Aligning, centring, orientation detection or correction of the image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G—PHYSICS
    • G02—OPTICS
    • G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
    • G02B17/00—Systems with reflecting surfaces, with or without refracting elements
    • G02B17/02—Catoptric systems, e.g. image erecting and reversing system
    • G02B17/06—Catoptric systems, e.g. image erecting and reversing system using mirrors only, i.e. having only one curved mirror
    • G02B17/0605—Catoptric systems, e.g. image erecting and reversing system using mirrors only, i.e. having only one curved mirror using two curved mirrors
    • G02B17/061—Catoptric systems, e.g. image erecting and reversing system using mirrors only, i.e. having only one curved mirror using two curved mirrors on-axis systems with at least one of the mirrors having a central aperture
    • G—PHYSICS
    • G02—OPTICS
    • G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
    • G02B26/00—Optical devices or arrangements for the control of light using movable or deformable optical elements
    • G02B26/08—Optical devices or arrangements for the control of light using movable or deformable optical elements for controlling the direction of light
    • G02B26/10—Scanning systems
    • G02B26/101—Scanning systems with both horizontal and vertical deflecting means, e.g. raster or XY scanners
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10048—Infrared image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G06V20/194—Terrestrial scenes using hyperspectral data, i.e. more or other wavelengths than RGB

Definitions

  • the invention belongs to the technical field of image recognition and infrared detection, and more particularly to an infrared spectrum correlation intelligent detection method and device.
  • Spectral data acquisition mainly studies methods and techniques for acquiring spectral data of a target scene or a region of interest. This technology is widely used in the field of remote sensing. It provides a data foundation for studying the spectral characteristics of various target backgrounds, and then classifying, monitoring and detecting targets.
  • Infrared map correlation refers to the combination of infrared image and infrared spectrum for target detection, which can increase the target range of detection range and improve target recognition ability. Therefore, research and development of relevant spectral imaging equipment is highly valued at home and abroad.
  • map detecting devices are multi-spectral scanners and Fourier transform infrared imaging spectrometers. Multi-spectral scanners are typically mounted on an aircraft that scans the mirror for rotation so that the received instantaneous field of view is moved perpendicular to the direction of flight for scanning. Due to the forward motion of the aircraft, the multi-spectral scanner completes the two-dimensional scanning, and the ground scene is scanned point by point and measured point by point, thereby obtaining multi-spectral remote sensing image information.
  • Non-real-time detection that is more suitable for stationary targets is difficult to apply to moving targets.
  • the Fourier transform infrared imaging spectrometer can provide rich two-dimensional spatial information and third-dimensional spectral data, that is, each point of two-dimensional spatial imaging can extract spectral information.
  • This kind of device image and spectrum detection has a total of one sensor. The amount of signal processing information is very large, and it is impossible to achieve high spatial resolution and high time resolution at the same time, and the price is expensive, which is difficult for the user to accept.
  • the spectrum of stationary features and sky background does not need to be acquired in real time. It is necessary to use the spectral characteristics to automatically detect real-time detection of moving objects or time-varying objects (local areas) in the scene, such as flying. Medium aircraft, ships at sea, vehicles in motion, Fire, explosion, etc.
  • the existing "integrated map device” prototype can realize the automatic detection and spectral recognition of the above multiple moving targets and time-varying objects, but it has the following problems: (1) The device can only acquire the medium wave band ( The spectrum of 2 ⁇ m ⁇ 5 ⁇ m), while the spectral characteristics of normal temperature and low temperature targets are mainly in the long-wavelength band (8 ⁇ m ⁇ 14 ⁇ m), the device can not effectively detect such targets; (2) the device is interested in the target in the field of view The infrared image and spectrum are measured, and the target can be effectively detected by using only the infrared image.
  • the device also measures the spectrum for spectral feature recognition, which reduces the efficiency of detection and recognition; (3) the device uses a step scan tracking mirror The tracking accuracy is relatively low; (4) The device uses an infrared window to effectively protect the internal optical components, but for the use of the conventional target detection of the stationary platform with better test conditions, it is not necessary to use an infrared window to reduce the cost.
  • the object of the present invention is to provide an infrared spectrum correlation intelligent detection method and device, which aims to solve the problem that the existing infrared image detection device cannot effectively detect a target when the spatial resolution low target shape information is unavailable.
  • the infrared spectrum detecting device uses the map correlation detection when the spatial resolution high target shape information is available, and the recognition efficiency is low, and the problem of the normal temperature and the low temperature target cannot be effectively detected.
  • the invention provides an infrared spectrum correlation intelligent detection method, comprising the following steps:
  • N is an integer greater than or equal to 1;
  • step (3) Performing object recognition based on the shape information on the i-th target in the scene.
  • the process proceeds to step (4); when the recognition rate of the i-th target is If it is less than the set threshold, it proceeds to step (5); the initial value of i is 1;
  • the target spectral characteristics include a spectral peak, a peak wavelength, a number of spectral peaks and a spacing, and an area of the spectral peak;
  • the invention also provides an infrared spectrum correlation intelligent detecting device, comprising a two-dimensional scanning mirror, a multi-band infrared optical module, a long-wave infrared imaging unit, a broadband infrared spectrum measuring unit, a processing and control unit and a power module;
  • the input end of the band infrared optical module is configured to receive infrared incident light reflected by the two-dimensional scanning mirror, and the long-wave infrared imaging unit is connected to the first output end of the multi-band infrared optical module, the wide-band infrared spectrum a unit is coupled to the second output of the multi-band infrared optical module, the spectral input of the processing and control unit is coupled to the wide-band infrared spectrometer unit, the image input end of the processing and control unit is a long-wave infrared imaging unit is connected, the control output of the processing and control unit is connected to a control end of the two-dimensional scanning mirror; the output end of the power module is respectively connected
  • the infrared incident light is reflected by the two-dimensional scanning mirror to the multi-band infrared optical module, and after being concentrated, it can be directly passed to the long-wave infrared imaging unit for imaging or divided into two groups: a long-wave infrared imaging unit and a long-wave infrared infrared beam.
  • a long-wave infrared imaging unit for imaging or divided into two groups: a long-wave infrared imaging unit and a long-wave infrared infrared beam.
  • processing and control unit to receive images acquired by long-wave infrared imaging units and spectra acquired by wide-band infrared spectroscopy units, image and spectral processing, control of two-dimensional scanning mirrors The movement to achieve target tracking and recognition.
  • the two-dimensional scanning mirror includes a plane mirror and a two-dimensional servo turret
  • the plane mirror is disposed on the two-dimensional servo turret, and the plane reflection is driven by controlling the movement of the two-dimensional servo turret
  • the mirror achieves two dimensions of pitch and rotation.
  • the multi-band infrared optical module includes an infrared lens, a beam splitter, a long-wave infrared imaging lens group, a wide-band infrared spectroscopic lens group, an FPA interface, and a fiber optic interface; the optical axis of the spectroscope and the infrared lens are placed at 45 degrees.
  • the beam splitter is movable, and the spectroscope is moved out when the full field of view scan and the image information based object are recognized; when the target spectrum is acquired to identify the target based on the map feature database, the spectroscope is not removed; the long wave
  • An infrared imaging lens group is disposed on an optical axis of the transmission optical path of the beam splitter, the wide-band infrared spectroscopic lens group is disposed on an optical axis of a reflected light path of the beam splitter, and the FPA interface is disposed on the long wave
  • the FPA interface is for coupling with a long-wave infrared imaging unit;
  • the optical fiber interface is disposed on an optical axis of the wide-band infrared spectroscopic lens group, and the optical fiber interface is used for width and width
  • the band infrared spectrum unit is coupled.
  • the infrared lens is a Cassegrain type multi-band infrared lens.
  • the beam splitter is coated with a transflective film, which has a transflective effect on infrared light having a wavelength of 8 ⁇ m to 12 ⁇ m and reflects infrared light having a wavelength of 2 ⁇ m to 8 ⁇ m. effect.
  • the method of the present invention uses the spectral features of the target for target recognition in addition to the image information, and is added for identification.
  • the information dimension of different targets can improve the detection and recognition rate; since the range of the acquired target spectrum is extended from the short- and medium-wave infrared (2 ⁇ 5 ⁇ m) to the short, medium and long-wave infrared (2 ⁇ 12 ⁇ m), The high-temperature target with the main spectral features in the short and medium-wave infrared bands is identified, and the low-temperature and normal-temperature targets with the main spectral features in the long-wave infrared (8-12 ⁇ m) can be identified, and the target detection and recognition range is increased.
  • the method of the invention combines the long-wave infrared image of the target and the wide-band infrared spectrum to perform target detection and recognition, which is an improvement and improvement of the existing infrared detection method and device, and can solve the existing infrared detection system at a long distance.
  • Low detection and low resolution can not effectively detect the target and the existing infrared spectrum detection equipment can not effectively detect the problem of normal temperature and low temperature target, and can also achieve the target wide-band infrared spectrum acquisition.
  • 1 is a schematic diagram of the principle of infrared spectrum correlation detection method
  • FIG. 2 is a schematic diagram of a process of acquiring a target spectrum when a target is used
  • FIG. 3 is a schematic diagram of a process of acquiring a target spectrum when a plurality of targets are obtained
  • FIG. 4 is a schematic structural view of a smart infrared map correlation detecting device
  • FIG. 5 is a schematic diagram of a split mode optical path switching mode, wherein (a) is a mobile beam splitter optical path switching mode, and (b) a rotary beam splitter optical path switching mode;
  • FIG. 6 is a schematic diagram of an optical path when the spectroscope is cut out and cut in, wherein (a) is a schematic diagram of an optical path when the spectroscope is cut out, and (b) a schematic diagram of an optical path when the spectroscope is cut;
  • FIG. 7 is a schematic structural view of a two-dimensional scanning mirror, (a) a main view, and (b) a top view (c) is a left view;
  • Figure 8 is an example of detection using an infrared image detection mode, and (a) and (b) are two-frame long-wave infrared images of an aircraft that has just taken off;
  • Figure 9 is an example of detection using an infrared pattern correlation detection mode, (a) is a long-wave infrared image of a high-pressure sodium lamp at an airport, (b) is a spectrum of a high-pressure sodium lamp, (c) is a long-wave infrared image of a taxi aircraft, and (d) is The spectrum of the tail rotor and tail nozzle of the taxi aircraft.
  • the invention provides an infrared spectrum correlation intelligent detection method, which realizes intelligent detection of infrared image and infrared spectrum correlation, and has two modes of infrared image detection and map correlation detection.
  • the infrared image detection mode refers to: using a conventional infrared detection process, after acquiring an infrared image, the region of interest is extracted by an image processing method, and then the target is identified by using information such as a shape.
  • the infrared spectrum correlation detection mode refers to: long-wave infrared image and medium-long wave red
  • the external spectrum organic fusion is used for target detection and recognition. The target is locked in the center of the field of view, and then the infrared spectrum is acquired, and then the target recognition based on infrared spectral features is performed.
  • the target in the field of view is searched first, and then the search target is intelligently identified by the map correlation, that is, the infrared image target recognition is performed first for each target, and if the detection recognition rate is greater than or equal to the setting
  • the threshold value outputs the recognition result and saves the target image data; otherwise, the target infrared spectrum is acquired, and the target recognition based on the infrared spectrum feature is performed, and if the comparison is successful, the recognition result is output and the target map data is saved; otherwise, the target infrared spectrum is acquired.
  • Features are added to the infrared spectral feature database.
  • FIG. 1 A schematic diagram of the principle of the method of the present invention is shown in FIG. 1 and mainly includes the following steps:
  • N is an integer greater than or equal to 1;
  • step (3) Performing object recognition based on the shape information on the i-th target in the scene.
  • the process proceeds to step (4); when the recognition rate of the i-th target is If it is less than the set threshold, the process proceeds to step (5); the initial value of i is 1; wherein the threshold is an empirical value, which may be (85%-97%), preferably 90%, 95%, 97%.
  • the target spectral characteristics include a spectral peak, a peak wavelength, a number of spectral peaks and a spacing, an area of the spectral peak, and the like;
  • the alignment match is to compare the measured spectrum with the spectrum in the database to find the spectrum with the strongest similarity to the measured spectrum.
  • the similarity of the two spectral curves can be judged by the normalized distance method, and the two spectral curves with the shortest distance are considered to be the most similar.
  • the distance between spectral curves can be defined as the sum of squared or modulo of each data point. It can also be judged by the product energy of the two normalized spectral signals, that is, the sum of the products of the two normalized signals, and the largest value is considered to be the most similar.
  • the basis of the infrared pattern correlation detection mode for target recognition is to first acquire the spectrum of the target, and to obtain the target spectrum in the already searched field of view, there are mainly two stages: target tracking and target locking spectrum.
  • Target tracking refers to changing the field of view so that the measured object coincides with the center of the field of view of the infrared image at the current time.
  • the target-locked spectrum is obtained by changing the field of view so that the center of the field of view and the finger keep the target synchronous motion relatively stationary, locking the object to be measured at the center of the field of view, and dividing the infrared incident light that acquires the target radiation into two parts.
  • the spectrum and image of the target is to first acquire the spectrum of the target, and to obtain the target spectrum in the already searched field of view, there are mainly two stages: target tracking and target locking spectrum.
  • Target tracking refers to changing the field of view so that the measured object coincides with the center of the field of view of the infrared image at the current time.
  • the searched objects are prioritized according to the distance between the target and the center of the field of view, so as to ensure that as many as possible can be identified. aims.
  • FIG. 2 A schematic diagram of the target spectrum acquisition process for a target and multiple targets is shown in Figures 2 and 3.
  • the invention also provides a smart infrared map correlation detecting device, the structure of which is shown in FIG. 4, comprising a two-dimensional scanning mirror 3, a multi-band infrared optical module 4, a long-wave infrared imaging unit 5, and a broadband infrared spectrum measuring unit 7
  • the processing and control unit 8 and the power module 9 After the infrared light is incident on the inside of the system, it is reflected by the two-dimensional scanning mirror 3 and then incident on the multi-band infrared optical module 4. After being concentrated, it can be directly passed to the long-wave infrared imaging unit 5 for imaging or divided into a growing wave infrared and a wide-band infrared two.
  • the beam arrives at the long-wave infrared imaging unit 5 and the wide-band infrared spectroscopy unit 7, respectively, for imaging and spectroscopy.
  • the processing and control unit 8 receives the image acquired by the red long-wave imaging unit 5 and the spectrum acquired by the wide-band infrared spectrum unit 7, performs image and spectral processing, controls the motion of the two-dimensional scanning mirror 3, and achieves target tracking and recognition.
  • the power module 9 supplies power to the two-dimensional scanning mirror 3, the multi-band infrared optical module 4, the long-wave infrared imaging unit 5, and the wide-band infrared spectrum unit 7.
  • the two-dimensional scanning mirror 3 can be composed of a plane mirror 31 and a two-dimensional servo turret 32, and can realize two dimensions of motion of pitch and rotation.
  • the multi-band infrared optical module 4 can be composed of a Cassegrain-type multi-band infrared lens 41, a beam splitter 42, a long-wave infrared imaging lens group 43, and a wide-band infrared spectroscopic lens group 44, and multi-band infrared optics.
  • Module 4 has an FPA (Focal Plane Array) interface 45 and a fiber optic interface 46 coupled to the uncooled long-wave infrared imaging unit and the broadband infrared fiber, respectively.
  • FPA Fluor Plane Array
  • the Cassegrain-type multi-band infrared lens 41 folds the optical path, compresses the volume of the optical system, and has a good convergence effect on short, medium and long-wave infrared light.
  • the spectroscope is coated with a special transflective splitting film, which has a transflective effect on long-wave (8 ⁇ m ⁇ 12 ⁇ m) infrared, and a reflection on short, medium-wave (2 ⁇ m ⁇ 8 ⁇ m) infrared.
  • the beam splitter of the existing map device is fixed.
  • the beam splitter is movable, and is placed at 45° with the optical axis of the infrared light concentrated by the Cassegrain-type multi-band infrared lens 41, and the beam splitter has two. Position, cut into the optical path and cut out the optical path, the movement of the spectroscope can be realized by a rotating or moving mechanism, as shown in Fig. 5, wherein (a) is a mobile structure and (b) is a rotary structure.
  • the light path diagram of the beam splitter when cutting in and out is shown in Fig. 6(a) and (b). In Fig.
  • the wideband infrared spectrogram unit 7 can be a non-imaging Fourier transform unit detector coupled to a wideband infrared spectroscopic lens set by a wideband infrared fiber 6.
  • the above components may be enclosed inside the casing 1, and the side of the casing 1 is provided with side windows through which infrared incident light is incident into the interior of the system.
  • the two-dimensional scanning mirror 3 provided by the embodiment of the present invention is composed of a plane mirror 31 and a two-dimensional servo turret 32, and can realize two dimensions of motion of pitch and rotation.
  • the two-dimensional scanning mirror uses a U-shaped seat as a support to offset the lens from the rotating shaft of the motor by a distance, and there is a deviation between the motor rotation axis and the lens rotation center axis. That is, there is a deviation between the angle at which the motor rotates and the angle at which the actual tracking object moves.
  • a schematic diagram of the structure of the two-dimensional scanning mirror is shown in Fig. 7, wherein (a) is a main view, and (b) a top view is (c) a left view.
  • the geometric center of the lens is 210mm away from the mounting surface of the base.
  • the two axes of the two-dimensional servo turret are equipped with a rotating hard limit to prevent malfunction.
  • the two-dimensional turntable drives the load lens to form a two-dimensional tracking scan of the horizontal and vertical planes, wherein the horizontal scanning is ⁇ 5°, the vertical scanning is -10° to 25°, the scanning maximum speed is 16°/s, and the rotation angle accuracy is 0.013°.
  • the two-dimensional servo turntable adopts a programmable multi-axis controller (referred to as PMAC) based on DSP technology as a motion control system.
  • PMAC uses Motorola's DSP56001/56002 digital signal processor as the central processing unit to simultaneously control 2 to 8 axes for fully coordinated motion through a flexible high-level language.
  • PMAC provides basic functions such as motion control, housekeeping, and host interaction. Its speed, resolution, bandwidth and other indicators are far superior to the general motion controller. Fully meet the high-precision, high-response control requirements of two-dimensional turntables.
  • the infrared fiber 6 can use a sulfur-based glass fiber to realize the coupling of the fiber interface 46 on the multi-band infrared optical module 4 with the wide-band infrared spectrum unit 7, and can transmit short, medium, long and wide bands (2 ⁇ m ⁇ Infrared light of 12 ⁇ m), because the fiber is flexible, the fiber structure can make the system structure more compact and smaller.
  • the wide-band infrared spectroscopy unit 7 is configured to perform interference sampling on the incident light, and obtain a wide-band infrared spectrum of the target by Fourier transform; in the embodiment of the present invention, a Bruker Optics company may be used.
  • Spectral detection unit EM27 or process control spectrometry system IRCube OEM both using Michelson interferometer system, spectral resolution 2cm -1 , 4cm -1 , 8cm -1 , 16cm -1 , 32cm -1 optional , measuring spectrum MCT detectors with a range of 2 ⁇ m to 12 ⁇ m and cooled with Stirling or liquid nitrogen.
  • the long-wave infrared imaging unit 5 can adopt the UL03041 uncooled long-wave infrared detector of the French ULIS company, and the imaging band is 8 ⁇ m to 14 ⁇ m, which belongs to the thermistor focal plane, the detecting material is polysilicon, and the thermal response time is 7 ms.
  • the filling coefficient is greater than 80%, the pixel sampling frequency is 7.375MHz, the number of failed pixels is less than 1%, the power consumption is less than 4W, the frame frequency is 50HZ, the resolution is 384*288, and the equivalent noise temperature difference is 60mk.
  • the processing and control unit 8 can employ a hardware architecture that can employ an FPGA+DSP+ dedicated ASIC, SOC.
  • DSPs can use multi-core processors, and FPGAs can use Xilinx or Altera products.
  • ASIC application specific integrated circuit
  • Figures 8 and 9 show two examples of detection.
  • Figures 8(a) and (b) are two-frame long-wave infrared images of an aircraft that has just taken off. It is only necessary to use image information to identify the aircraft. It is not necessary to use the map to detect;
  • Figures 9(a) and (c) are shot at night.
  • the long-wave infrared image of the high-pressure sodium lamp and the taxiing aircraft at the airport because it occupies fewer pixels in the image, can not effectively detect and distinguish the two by image information such as shape, and then use the map correlation detection mode to detect and identify
  • Figure 9 (b) and (d) are the spectra of high-pressure sodium lamps and taxi tail flames and tail nozzles. Although they are similar in shape but have a large difference in spectrum, spectral information can be used to effectively distinguish the two.
  • the present invention can be qualitatively analyzed in conjunction with FIG. 8 and FIG. 9 to solve the above technical problems and achieve the object of our invention.

Landscapes

  • Physics & Mathematics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Health & Medical Sciences (AREA)
  • Optics & Photonics (AREA)
  • Toxicology (AREA)
  • General Engineering & Computer Science (AREA)
  • Library & Information Science (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Geophysics (AREA)
  • General Life Sciences & Earth Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Software Systems (AREA)
  • Spectrometry And Color Measurement (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)
  • Photometry And Measurement Of Optical Pulse Characteristics (AREA)

Abstract

一种红外图谱关联智能探测方法，包括：先搜索视场中目标，然后依次对搜索到的目标进行图谱关联智能识别，即对每一个目标先进行红外图像目标识别，若探测识别率大于等于设定阈值，则输出识别结果并保存目标图像数据；否则，则获取目标红外光谱，进行基于红外光谱特征的目标识别。用上述方法进行目标探测的装置，主要包括二维扫描转镜（3）、多波段红外光学模块（4）、长波红外成像单元（5）、宽波段红外测谱单元（7）和处理与控制单元（8）。该方法和装置可以用于目标的红外图像探测和红外图谱关联探测及目标红外光谱采集，相比常规红外探测设备，性价比高，能显著提高目标的探测识别率。

Description

一种红外图谱关联智能探测方法及装置 【技术领域】
本发明属于图像识别与红外探测技术领域，更具体地，涉及一种红外图谱关联智能探测方法及装置。
【背景技术】
所有温度高于绝对零度的物体都能产生红外辐射，温度越高，辐射出的能量就越大，而且物质的光谱特性曲线是唯一的。光谱数据采集主要研究采集目标场景或者感兴趣区的光谱数据的方法和技术。该技术广泛应用于遥感领域，为研究各种目标背景的光谱特性，进而对场景进行分类、监视与目标探测识别提供数据基础。
红外图谱关联是指红外图像和红外光谱相结合进行目标探测，可以增大探测范围目标的种类、提高目标识别能力。因此，国内外都非常重视研发相关光谱成像设备。目前，常用的图谱探测设备为多光谱扫描仪和傅立叶变换红外成像光谱仪。多光谱扫描仪一般安装在飞行器上，其扫描转镜旋转，使接收的瞬时视场作垂直于飞行方向的运动，从而实现扫描。由于飞行器的前向运动，多光谱扫描仪即完成二维扫描，地物景象被逐点扫过，并逐点分波段测量，从而获得多光谱的遥感图像信息。较适用于静止目标的非实时探测，对于运动目标难以适用。傅立叶变换红外成像光谱仪能够提供丰富的二维空间信息及第三维的光谱数据，即二维空间成像的每一点都可以提取光谱信息。这种设备图像、光谱探测共一个传感器，信号处理信息量非常大，无法同时实现高空间分辨率和高时间分辨率，且价格昂贵，用户难以接受。
在许多实际应用中，静止的地物和天空背景的光谱并不需要实时获取，而需要利用光谱特性对场景中运动目标或时变对象(局部区域)进行自动实时地检测识别探测识别，如飞行中飞机、海上的船舶、行驶中的车辆、 火灾、爆炸等。
现有的“图谱一体化设备”原理样机，能实现上述多个运动目标和时变对象的自动检测与光谱识别，但其存在以下几个问题：(1)该设备只能获取中波波段(2μm～5μm)的光谱，而常温和低温目标的光谱特征主要在长波波段(8μm～14μm)，对此类目标该设备不能进行有效探测；(2)该设备对视场中感兴趣的目标都测其红外图像和光谱，而对于只用红外图像就可以有效探测识别的目标该设备也测其光谱进行光谱特征识别，降低了探测识别的效率；(3)该设备采用步进扫描跟踪转镜，跟踪精度比较低；(4)该设备采用红外窗口可以有效保护内部光学部件，但对于较好试验条件的静止平台常规目标探测的使用需求，没必要采用红外窗口以降低成本。
【发明内容】
针对现有技术的缺陷，本发明的目的在于提供一种红外图谱关联智能探测方法及装置，旨在解决现有红外图像探测设备在空间分辨率低目标形状信息不可用时不能有效探测目标，现有红外图谱探测设备在空间分辨率高目标形状信息可用时用图谱关联探测识别效率低、不能有效探测常温和低温目标的问题。
本发明提供了一种红外图谱关联智能探测方法，包括下述步骤：
(1)获取目标场景的红外图像，并对红外图像进行图像处理，提取场景中的N个目标；N为大于等于1的整数；
(2)根据目标与视场中心的距离，由小到大对所述N个目标进行排序；
(3)对场景中的第i个目标进行基于形状信息的目标识别，当第i个目标的识别率大于等于设定的阈值时，则进入步骤(4)；当第i个目标的识别率小于设定的阈值时，则进入步骤(5)；i的初始值为1；
(4)i＝i+1，并判断i是否大于N，若是，则结束，若否，则返回至步骤(3)；
(5)通过改变视场范围，使得第i个目标与当前时刻红外图像的视场 中心重合；
(6)将第i个目标辐射的红外光分成两束，一束光经过成像后获得图像，另一束光经过干涉后得到干涉图，再进行傅里叶逆变换获得光谱；
(7)对所述光谱进行处理并提取目标光谱特征；目标光谱特征包括光谱峰值、峰值波长、光谱峰个数及间距、光谱峰的面积；
(8)将所述第i个目标的光谱特征与预设的数据库中的光谱特征进行比对匹配，当能匹配时，则识别出目标并获得目标的图像和光谱，并返回至步骤(4)；若不能匹配，则将所述目标光谱特征加入所述数据库中，并返回至步骤(4)。
本发明还提供了一种红外图谱关联智能探测装置，包括二维扫描转镜、多波段红外光学模块、长波红外成像单元、宽波段红外测谱单元、处理与控制单元和电源模块；所述多波段红外光学模块的输入端用于接收被二维扫描转镜反射的红外入射光，所述长波红外成像单元与所述多波段红外光学模块的第一输出端连接，所述宽波段红外测谱单元与所述多波段红外光学模块的第二输出端连接，所述处理与控制单元的光谱输入端与所述宽波段红外测谱单元连接，所述处理与控制单元的图像输入端与所述长波红外成像单元连接，所述处理与控制单元的控制输出端与二维扫描转镜的控制端连接；所述电源模块的输出端分别与所述宽波段红外测谱单元、所述长波红外成像单元、所述多波段红外光学模块和所述二维扫描转镜的电源端连接，用于分别提供电源。
工作时，红外入射光被二维扫描转镜反射至多波段红外光学模块，被汇聚后可直通到达长波红外成像单元用来成像或被分成长波红外和宽波段红外两束分别到达长波红外成像单元用于成像和宽波段红外测谱单元用于成谱；处理与控制单元接收长波红外成像单元获取的图像和宽波段红外测谱单元获取的光谱，进行图像和光谱处理，控制二维扫描转镜的运动，实现目标跟踪和识别。
更进一步地，所述二维扫描转镜包括平面反射镜和二维伺服转台，所述平面反射镜设置在所述二维伺服转台上，通过控制所述二维伺服转台运动带动所述平面反射镜实现俯仰和旋转两个维度的转动。
更进一步地，所述多波段红外光学模块包括红外镜头、分光镜、长波红外成像透镜组、宽波段红外成谱透镜组、FPA接口和光纤接口；分光镜与红外镜头的光轴成45度放置，所述分光镜可移动，当全视场扫描和基于图像信息目标识别时，所述分光镜移出；当获取目标光谱进行基于图谱特征数据库识别目标时，所述分光镜不移出；所述长波红外成像透镜组设置在所述分光镜的透射光路的光轴上，所述宽波段红外成谱透镜组设置在所述分光镜的反射光路的光轴上，所述FPA接口设置在所述长波红外成像透镜组的光轴上，所述FPA接口用于与长波红外成像单元耦合；所述光纤接口设置在所述宽波段红外成谱透镜组的光轴上，所述光纤接口用于与宽波段红外测谱单元耦合。
更进一步地，所述红外镜头为卡塞格林式多波段红外镜头。
更进一步地，所述分光镜上镀有半反半透分光膜，所述分光镜对波长为8μm～12μm的红外光有半反半透作用，且对波长为2μm～8μm的红外光有反射作用。
通过本发明所构思的以上技术方案，与现有技术相比，由于不同物质的光谱特征是不同的，本发明方法除图像信息外还将目标的光谱特征用于目标识别，增加了用于识别不同目标的信息维度，能够提高探测识别率；由于将获取的目标光谱的波段范围从短、中波红外波段(2～5μm)扩展到了短、中、长波红外波段(2～12μm)，不仅可以识别主要光谱特征位于短、中波红外波段的高温目标，还可以识别主要光谱特征位于长波红外波段(8～12μm)的低温和常温目标，增大了目标探测识别范围。本发明方法将目标的长波红外图像和宽波段红外光谱相融合进行目标探测识别，是对现有红外探测方法和设备的改进和提高，可以解决现有红外探测系统在远距 离探测、分辨率低无法有效探测目标及现有红外图谱探测设备不能有效探测常温和低温目标的问题，也可以实现目标宽波段红外光谱采集。
【附图说明】
图1为红外图谱关联探测方法原理示意图；
图2为一个目标时获取目标光谱过程示意图；
图3为多个目标时获取目标光谱过程示意图；
图4为灵巧红外图谱关联探测装置结构示意图；
图5为分光镜光路切换方式示意图，其中(a)为移动式分光镜光路切换方式，(b)旋转式分光镜光路切换方式；
图6为分光镜切出、切入时的光路示意图，其中(a)为分光镜切出时的光路示意图，(b)分光镜切入时的光路示意图；
图7为二维扫描转镜的结构示意图，(a)为主视图，(b)俯视图为(c)为左视图；
图8为采用红外图像探测模式的探测实例，(a)和(b)是刚起飞的飞机的两帧长波红外图像；
图9为采用红外图谱关联探测模式的探测实例，(a)是机场的高压钠灯的长波红外图像，(b)是高压钠灯的光谱，(c)是滑行飞机的长波红外图像，(d)是滑行飞机尾焰及尾喷管的光谱。
【具体实施方式】
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。
本发明提供了一种红外图谱关联智能探测方法，实现红外图像和红外光谱关联的智能探测，有红外图像探测和图谱关联探测两种模式。
在本发明中，红外图像探测模式是指：采用常规的红外探测流程，获取红外图像后，通过图像处理方法提取感兴趣区域，然后利用形状等信息进行目标识别。红外图谱关联探测模式是指：将长波红外图像和中长波红 外光谱有机融合进行目标检测与识别，先将目标锁定在视场中心，再获取其红外光谱，然后进行基于红外光谱特征的目标识别。
采用本发明方法进行目标识别时，先搜索视场中的目标，然后依次对搜索到的目标进行图谱关联智能识别，即对每一个目标先进行红外图像目标识别，若探测识别率大于等于设定阈值，则输出识别结果并保存目标图像数据；否则，则获取目标红外光谱，进行基于红外光谱特征的目标识别，若比对匹配成功则输出识别结果并保存目标图谱数据，否则，将目标红外光谱特征加入红外光谱特征数据库。
本发明方法的原理示意图见图1，主要包括以下步骤：
(1)获取目标场景的红外图像，并对红外图像进行图像处理，提取场景中的N个目标；N为大于等于1的整数；
(2)根据目标与视场中心的距离，由小到大对所述N个目标进行排序；
(3)对场景中的第i个目标进行基于形状信息的目标识别，当第i个目标的识别率大于等于设定的阈值时，则进入步骤(4)；当第i个目标的识别率小于设定的阈值时，则进入步骤(5)；i的初始值为1；其中，阈值为经验值，可以为(85％-97％)，优选为90％、95％、97％。
(4)i＝i+1，并判断i是否大于N，若是，则结束，若否，则返回至步骤(3)；
(5)通过改变视场范围，使得第i个目标与当前时刻红外图像的视场中心重合；
(6)将第i个目标辐射的红外光分成两束，一束光经过成像后获得图像，另一束光经过干涉后得到干涉图，再进行傅里叶逆变换获得光谱；
(7)对所述光谱进行处理并提取目标光谱特征；目标光谱特征包括光谱峰值、峰值波长、光谱峰个数及间距、光谱峰的面积等；
(8)将所述第i个目标光谱特征与预设的数据库中的光谱特征进行比对匹配，当能匹配时，则识别出目标并获得目标的图像和光谱，并返回至 步骤(4)；若不能匹配，则将所述目标光谱特征加入所述数据库中，并返回至步骤(4)。
在本发明中，比对匹配是将测量光谱和数据库中的光谱进行比对，找到与测量光谱相似性最强的光谱。两条光谱曲线的相似性可以通过归一化后的求距离法来判断，距离最短的两个光谱曲线视为最相似。光谱曲线之间的距离可定义为各数据点平方和或模的和。也可用两个归一化后光谱信号的乘积能量判断，即两个归一化信号乘积的和，数值最大的视为最相似。
在本发明中，采用红外图谱关联探测模式进行目标识别的基础是先要获取目标的光谱，获取已经搜索到的视场中的目标光谱主要有以下两个阶段：目标跟踪和目标锁定测谱。目标跟踪，是指通过改变视场范围，使得被测目标与当前时刻红外图像的视场中心重合。目标锁定测谱，是通过改变视场范围，使得视场中心和指将目标同步运动保持相对静止，将被测目标锁定在视场中心，并把获取目标辐射的红外入射光分成两部分，获取目标的光谱和图像。
在本发明中，如在视场中搜索到多个目标，则先根据目标与视场中心的距离，由小到大对搜索到的目标进行识别优先级排序，以保证可以识别尽可能多的目标。
一个目标和多个目标时目标光谱获取过程示意图见图2和图3。
图2中，在视场1时，搜索到一个运动目标，不在视场中心，先获取视场中心的光谱，作为背景谱用于光谱数据处理，然后跟踪目标，在视场2时，目标被锁定在视场中心，获取目标的光谱。
图3中，在视场1时，搜索到三个运动目标t1、t2和t3，不在视场中心，按照三个目标离视场中心的距离，从小到大排序，排好的顺序为：t1、t2、t3，先获取视场中心的光谱，作为背景谱用于光谱数据处理，然后按照t1、t2、t3的顺序依次跟踪三个目标并获取三个目标的光谱，在视场2、3、4时，将三个目标分别锁定在视场中心，获取三个目标的光谱。
本发明还提供了一种灵巧红外图谱关联探测装置，其结构如图4所示，包括二维扫描转镜3、多波段红外光学模块4、长波红外成像单元5、宽波段红外测谱单元7、处理与控制单元8、电源模块9。红外光入射到系统内部后，被二维扫描转镜3反射后入射多波段红外光学模块4，被汇聚后可直通到达长波红外成像单元5用来成像或被分成长波红外和宽波段红外两束分别到达长波红外成像单元5和宽波段红外测谱单元7，用于成像和成谱。处理与控制单元8接收红长波外成像单元5获取的图像和宽波段红外测谱单元7获取的光谱，进行图像和光谱处理，控制二维扫描转镜3的运动，实现目标跟踪和识别。电源模块9对二维扫描转镜3、多波段红外光学模块4、长波红外成像单元5、宽波段红外测谱单元7供电。
在本发明实施例中，二维扫描转镜3可以由平面反射镜31和二维伺服转台32构成，可以实现俯仰和旋转两个维度的运动。
在本发明实施例中，多波段红外光学模块4可以由卡塞格林式多波段红外镜头41、分光镜42、长波红外成像透镜组43和宽波段红外成谱透镜组44组成，多波段红外光学模块4上有FPA(Focal Plane Array，焦平面列阵)接口45和光纤接口46，分别与非制冷长波红外成像单元及宽波段红外光纤耦合。
卡塞格林式多波段红外镜头41，折叠了光路，压缩了光学系统的体积，对短、中、长波红外光具有良好的会聚作用。分光镜镀有特殊半反半透分光膜，对长波(8μm～12μm)红外有半反半透作用，对短、中波(2μm～8μm)红外有反射作用。现有图谱设备的分光镜是固定的，本发明中，分光镜是运动的，与被卡塞格林式多波段红外镜头41汇聚后的红外光的光轴成45°放置，分光镜有两个位置，切入光路和切出光路，分光镜的运动可以采用旋转或移动机构实现，如图5所示，其中(a)为移动式结构，(b)为旋转式结构。分光镜切入、切出时的光路示意图见图6(a)和(b)，图6(a)中，分光镜切出光路时，入射多波段红外光学模块的红外光，经卡塞格林 多波段红外镜头汇聚后全部直通经过FPA接口到达长波红外成像单元用于成像；图6(b)中，分光镜切入时，入射多波段红外光学模块的红外光，经卡塞格林多波段红外镜头汇聚后到达分光镜，被分光镜分成两束，分别到达长波红外成像单元和宽波段红外测谱单元，用于成像和成谱。
宽波段红外测谱单元7可以为非成像傅里叶变换单元探测器，通过宽波段红外光纤6与宽波段红外成谱透镜组相耦合。上述组件可以封闭在壳体1内部，壳体1侧面开有侧窗，红外入射光通过此窗口入射到系统内部。
本发明实施例提供的二维扫描转镜3由平面反射镜31和二维伺服转台32构成，可以实现俯仰和旋转两个维度的运动。二维扫描转镜采用U型座作为支撑，将镜片与电机的转轴偏移一个距离，电机转动轴线和镜片旋转中心轴线之间存在一个偏差。即电机旋转的角度和实际跟踪物体移动的角度之间存在一个偏差。二维扫描转镜的结构示意图见图7，其中(a)为主视图，(b)俯视图为(c)为左视图。镜片几何中心距离底座安装面210mm，二维伺服转台二个轴均配有转动硬限位，防止误动作。二维转台带动负载镜片形成水平面和竖直面二维跟踪扫描，其中水平扫描±5°，竖直扫描-10°～25°，扫描最高速度16°/s，旋转角度精度0.013°。
本实施例中，二维伺服转台采用基于DSP技术的可编程多轴控制器(简称PMAC)作为运动控制系统。PMAC使用Motorola公司的DSP56001/56002数字信号处理器作为中央处理单元，通过灵活的高级语言，能够同时控制2～8轴进行完全协调的运动。PMAC提供了运动控制、内务处理、同主机交互等基本功能，其速度、分辨率、带宽等指标均远优于一般的运动控制器。完全满足二维转台的高精度、高响应的控制要求。
本实施例中，红外光纤6可以采用硫系玻璃光纤来实现多波段红外光学模块4上的光纤接口46与宽波段红外测谱单元7的耦合，可以传输短、中、长宽波段(2μm～12μm)的红外光，由于光纤是柔性的，用光纤连接可以使系统结构更节凑，体积更小。
本实施例中，宽波段红外测谱单元7用于将入射光进行干涉采样，并通过傅里叶变换获取目标的宽波段红外光谱；本发明实施例中可以采用德国布鲁克(Bruker Optics)公司的光谱探测单元EM27或者过程控制光谱测量系统IRCube OEM，两者均采用迈克尔逊干涉仪体制，光谱分辨率2cm-1、4cm-1、8cm-1、16cm-1、32cm-1可选，测量光谱范围2μm～12μm，采用斯特林或液氮制冷的MCT探测器。
本发明实施例中，长波红外成像单元5可以采用的法国ULIS公司的UL03041非制冷长波红外探测器，成像波段是8μm～14μm，它属热敏电阻焦平面，探测材料为多晶硅，热响应时间7ms，充填系数大于80％，像素采样频率7.375MHz，失效像素数目小于1％，功耗小于4W，帧频50HZ，分辨率384*288，等效噪声温差60mk。
处理与控制单元8可以采用可采用FPGA+DSP+专用ASIC、SOC的硬件体系结构。DSP可以使用多核心处理器，FPGA可以采用Xilinx或Altera公司的产品。专用集成电路(ASIC)的使用，能进一步提高硬件设计的灵活性，减小模块体积，降低功耗。
图8和图9是给出了两个探测实例。图8(a)和(b)是刚起飞的飞机的两帧长波红外图像，只用图像信息可以识别出来是飞机，则没必要用图谱探测；图9(a)和(c)是晚上拍的机场的高压钠灯和滑行飞机的长波红外图像，因其在图像中占有较少像素，仅靠形状等图像信息无法进行有效探测和区分两者，则利用图谱关联探测模式进行探测识别，图9(b)和(d)是高压钠灯和滑行飞机尾焰及尾喷管的光谱，它虽然形状类似但是光谱有很大的区别，可以用光谱信息有效区分两者。结合图8和图9可以定性的分析出本发明能够解决上述技术问题，实现我们的发明目的。
本领域的技术人员容易理解，以上所述仅为本发明的较佳实施例而已，并不用以限制本发明，凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等，均应包含在本发明的保护范围之内。

Claims (8)

  1. 一种红外图谱关联智能探测方法，其特征在于，包括下述步骤：
    (1)获取目标场景的红外图像，并对红外图像进行图像处理，提取场景中的N个目标；N为大于等于1的整数；
    (2)根据目标与视场中心的距离，由小到大对所述N个目标进行排序；
    (3)对场景中的第i个目标进行基于形状信息的目标识别，当第i个目标的识别率大于等于设定的阈值时，则进入步骤(4)；当第i个目标的识别率小于设定的阈值时，则进入步骤(5)；i的初始值为1；
    (4)i＝i+1，并判断i是否大于N，若是，则结束，若否，则返回至步骤(3)；
    (5)通过改变视场范围，使得第i个目标与当前时刻红外图像的视场中心重合；
    (6)将第i个目标辐射的红外光分成两束，一束光经过成像后获得图像，另一束光经过干涉后得到干涉图，再进行傅里叶逆变换获得光谱；
    (7)对所述光谱进行处理并提取目标光谱特征；
    (8)将所述第i个目标的光谱特征与预设的数据库中的光谱特征进行比对匹配，当能匹配时，则识别出目标并获得目标的图像和光谱，并返回至步骤(4)；若不能匹配，则将所述目标光谱特征加入所述数据库中，并返回至步骤(4)。
  2. 如权利要求1所述的红外图谱关联智能探测方法，其特征在于，步骤(3)中所述阈值为85％～97％。
  3. 一种红外图谱关联智能探测装置，其特征在于，包括二维扫描转镜(3)、多波段红外光学模块(4)、长波红外成像单元(5)、宽波段红外测谱单元(7)、处理与控制单元(8)和电源模块(9)；
    所述多波段红外光学模块(4)的输入端用于接收被二维扫描转镜(3) 反射的红外入射光，所述长波红外成像单元(5)与所述多波段红外光学模块(4)的第一输出端连接，所述宽波段红外测谱单元(7)与所述多波段红外光学模块(4)的第二输出端连接，所述处理与控制单元(8)的光谱输入端与所述宽波段红外测谱单元(7)连接，所述处理与控制单元(8)的图像输入端与所述长波红外成像单元(5)连接，所述处理与控制单元(8)的控制输出端与二维扫描转镜(3)的控制端连接；
    所述电源模块(9)的输出端分别与所述宽波段红外测谱单元(7)、所述长波红外成像单元(5)、所述多波段红外光学模块(4)和所述二维扫描转镜(3)的电源端连接，用于分别提供电源。
  4. 如权利要求3所述的红外图谱关联智能探测装置，其特征在于，工作时，红外入射光被二维扫描转镜(3)反射至多波段红外光学模块(4)，被汇聚后可直通到达长波红外成像单元(5)用来成像或被分成长波红外和宽波段红外两束分别到达长波红外成像单元(5)用于成像和宽波段红外测谱单元(7)用于成谱；处理与控制单元(8)接收红长波外成像单元(5)获取的图像和宽波段红外测谱单元(7)获取的光谱，进行图像和光谱处理，控制二维扫描转镜(3)的运动，实现目标跟踪和识别。
  5. 如权利要求3所述的红外图谱关联智能探测装置，其特征在于，所述二维扫描转镜(3)包括平面反射镜(31)和二维伺服转台(32)，所述平面反射镜(31)设置在所述二维伺服转台(32)上，通过控制所述二维伺服转台(32)运动带动所述平面反射镜(31)实现俯仰和旋转两个维度的转动。
  6. 如权利要求3所述的红外图谱关联智能探测装置，其特征在于，所述多波段红外光学模块(4)包括红外镜头(41)、分光镜(42)、长波红外成像透镜组(43)、宽波段红外成谱透镜组(44)、FPA接口(45)和光纤接口(46)；
    分光镜(42)与红外镜头(41)的光轴成45度放置，所述分光镜(42) 可移动，当全视场扫描和基于图像信息目标识别时，所述分光镜(42)移出；当获取目标光谱进行基于图谱特征数据库识别目标时，所述分光镜(42)不移出；
    所述长波红外成像透镜组(43)设置在所述分光镜(42)的透射光路的光轴上，所述宽波段红外成谱透镜组(44)设置在所述分光镜(42)的反射光路的光轴上，所述FPA接口(45)设置在所述长波红外成像透镜组(43)的光轴上，所述FPA接口(45)用于与长波红外成像单元(5)耦合；所述光纤接口(46)设置在所述宽波段红外成谱透镜组(44)的光轴上，所述光纤接口(46)用于与宽波段红外测谱单元(7)耦合。
  7. 如权利要求5所述的红外图谱关联智能探测装置，其特征在于，所述红外镜头(41)为卡塞格林式多波段红外镜头。
  8. 如权利要求5所述的红外图谱关联智能探测装置，其特征在于，所述分光镜(42)上镀有半反半透分光膜，所述分光镜(42)对波长为8μm～12μm的红外光有半反半透作用，且对波长为2μm～8μm的红外光有反射作用。
PCT/CN2015/072678 2014-12-30 2015-02-10 一种红外图谱关联智能探测方法及装置 WO2016106956A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US15/104,921 US9759835B2 (en) 2014-12-30 2015-02-10 Infrared image-spectrum associated intelligent detection method and apparatus

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410843190.3A CN104501959B (zh) 2014-12-30 2014-12-30 一种红外图谱关联智能探测方法及装置
CN2014108431903 2014-12-30

Publications (1)

Publication Number Publication Date
WO2016106956A1 true WO2016106956A1 (zh) 2016-07-07

Family

ID=52943381

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/072678 WO2016106956A1 (zh) 2014-12-30 2015-02-10 一种红外图谱关联智能探测方法及装置

Country Status (3)

Country Link
US (1) US9759835B2 (zh)
CN (1) CN104501959B (zh)
WO (1) WO2016106956A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160371851A1 (en) * 2014-12-30 2016-12-22 Huazhong University Of Science And Technology Infrared image-spectrum associated intelligent detection method and apparatus
CN113744349A (zh) * 2021-08-31 2021-12-03 湖南航天远望科技有限公司 一种红外光谱图像测量对准方法、装置及介质
CN115170974A (zh) * 2022-09-08 2022-10-11 深圳市勘察研究院有限公司 基于ai智能检测溶洞连通性的方法及装置

Families Citing this family (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160206179A1 (en) * 2015-01-15 2016-07-21 National Taiwan University Assistive robot endoscopic system with intuitive maneuverability for laparoscopic surgery and method thereof
WO2017013863A1 (ja) * 2015-07-17 2017-01-26 日本電気株式会社 照射システム、照射方法およびプログラム記憶媒体
CN105676305B (zh) * 2015-12-31 2017-05-31 南京华图信息技术有限公司 一种共口径多视场图谱协同探测系统与方法
CN106706133B (zh) * 2016-12-31 2017-09-29 华中科技大学 一种点斑状目标姿态估计方法及系统
CN108957573B (zh) * 2017-05-19 2024-04-02 北京英特威视科技有限公司 危险源检测装置和方法
CN107563370B (zh) * 2017-07-07 2020-04-03 西北工业大学 一种基于视觉注意机制的海上红外目标检测方法
CN108038856B (zh) * 2017-12-22 2020-08-04 杭州电子科技大学 基于改进多尺度分形增强的红外小目标检测方法
CN108152863B (zh) * 2017-12-29 2019-10-25 华中科技大学 可大视场搜索的图谱协同探测系统及搜索方法
CN108415097B (zh) * 2017-12-29 2019-07-19 华中科技大学 一种多波段红外成像的图谱协同探测系统和方法
CN108198228B (zh) * 2017-12-29 2020-09-08 华中科技大学 一种天基红外高光谱图像智能压缩方法
CN108335330B (zh) * 2017-12-31 2019-05-14 华中科技大学 一种图谱协同实时处理系统
CN109299691B (zh) * 2018-09-25 2022-02-18 路庄 一种失火发生情况分析方法及装置
CN109341868B (zh) * 2018-12-03 2020-11-13 北京遥感设备研究所 一种灵敏度自动测试系统及方法
CN109489814A (zh) * 2018-12-05 2019-03-19 武汉纺织大学 图谱融合智能目标探测方法及其光路切换装置
CN109781259B (zh) * 2018-12-29 2020-05-19 华中科技大学 一种图谱关联的精准测量空中运动小目标红外光谱的方法
CN109781258A (zh) * 2018-12-29 2019-05-21 华中科技大学 一种图谱关联获取海面目标红外光谱的方法
US11726034B2 (en) 2019-03-07 2023-08-15 Missouri State University IR spectra matching methods
CN110109189B (zh) * 2019-05-17 2021-11-30 中国人民解放军92942部队 海上目标红外控制方法及装置
CN110470395A (zh) * 2019-07-10 2019-11-19 中国科学院上海技术物理研究所 一种成像光谱仪前置二维指向与定标装置
CN114764760A (zh) * 2020-12-30 2022-07-19 珠海汉升电子科技有限公司 基于边界搜索的红外图像多目标温度越限告警方法
CN112945381B (zh) * 2021-02-02 2022-08-16 上海机电工程研究所 一种基于视域选通的图谱分时探测系统及方法
CN113504181A (zh) * 2021-08-04 2021-10-15 梁宵 基于傅里叶红外光谱技术的气云监测装置及监测方法
CN114374779B (zh) * 2021-12-16 2023-06-20 中国科学院上海高等研究院 一种全光场成像相机及其成像方法及全光场成像装置
CN114354514B (zh) * 2021-12-24 2024-02-09 南昌大学 一种非接触式多模态材料感知与识别装置

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101303291A (zh) * 2008-06-05 2008-11-12 苏州大学 基于数字微镜器件的多通道多目标超光谱成像方法及系统
JP2009139352A (ja) * 2007-12-11 2009-06-25 Shimadzu Corp フーリエ変換型赤外分光光度計
CN102564589A (zh) * 2011-12-20 2012-07-11 华中科技大学 一种多波段动目标光谱特征探测识别方法和装置
US8330087B2 (en) * 2007-10-16 2012-12-11 Cambridge Research & Instrumentation, Inc. Spectral imaging system with dynamic optical correction
CN103776540A (zh) * 2013-12-30 2014-05-07 华中科技大学 一种多波段共光路图谱联合遥感测量系统及方法
CN103777348A (zh) * 2013-12-30 2014-05-07 华中科技大学 一种多波段灵巧红外光学系统

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7119337B1 (en) * 1997-08-04 2006-10-10 Ion Optics, Inc. Infrared radiation sources, sensors and source combinations, and methods of manufacture
CA2742273A1 (en) * 2008-11-04 2010-05-14 William Marsh Rice University Image mapping spectrometers
US10244190B2 (en) * 2009-03-02 2019-03-26 Flir Systems, Inc. Compact multi-spectrum imaging with fusion
US9635285B2 (en) * 2009-03-02 2017-04-25 Flir Systems, Inc. Infrared imaging enhancement with fusion
CN103646232B (zh) * 2013-09-30 2016-08-17 华中科技大学 飞行器地面运动目标红外图像识别装置
CN104484577B (zh) * 2014-12-30 2017-06-16 华中科技大学 一种基于山脊能量校正的山地中带状地下目标的探测方法
CN104501956B (zh) * 2014-12-30 2016-07-13 华中科技大学 一种超宽波段图谱关联探测装置与探测方法
CN104637073B (zh) * 2014-12-30 2017-09-15 华中科技大学 一种基于太阳照射阴影补偿的带状地下结构探测方法
CN104501959B (zh) * 2014-12-30 2016-08-17 华中科技大学 一种红外图谱关联智能探测方法及装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8330087B2 (en) * 2007-10-16 2012-12-11 Cambridge Research & Instrumentation, Inc. Spectral imaging system with dynamic optical correction
JP2009139352A (ja) * 2007-12-11 2009-06-25 Shimadzu Corp フーリエ変換型赤外分光光度計
CN101303291A (zh) * 2008-06-05 2008-11-12 苏州大学 基于数字微镜器件的多通道多目标超光谱成像方法及系统
CN102564589A (zh) * 2011-12-20 2012-07-11 华中科技大学 一种多波段动目标光谱特征探测识别方法和装置
CN103776540A (zh) * 2013-12-30 2014-05-07 华中科技大学 一种多波段共光路图谱联合遥感测量系统及方法
CN103777348A (zh) * 2013-12-30 2014-05-07 华中科技大学 一种多波段灵巧红外光学系统

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160371851A1 (en) * 2014-12-30 2016-12-22 Huazhong University Of Science And Technology Infrared image-spectrum associated intelligent detection method and apparatus
US9759835B2 (en) * 2014-12-30 2017-09-12 Huazhong University Of Science And Technology Infrared image-spectrum associated intelligent detection method and apparatus
CN113744349A (zh) * 2021-08-31 2021-12-03 湖南航天远望科技有限公司 一种红外光谱图像测量对准方法、装置及介质
CN115170974A (zh) * 2022-09-08 2022-10-11 深圳市勘察研究院有限公司 基于ai智能检测溶洞连通性的方法及装置

Also Published As

Publication number Publication date
CN104501959B (zh) 2016-08-17
US20160371851A1 (en) 2016-12-22
US9759835B2 (en) 2017-09-12
CN104501959A (zh) 2015-04-08

Similar Documents

Publication Publication Date Title
WO2016106956A1 (zh) 一种红外图谱关联智能探测方法及装置
WO2015101062A1 (zh) 一种多波段共光路图谱联合遥感测量系统及方法
WO2016106957A1 (zh) 一种超宽波段图谱关联探测装置与探测方法
WO2016106953A1 (zh) 一种动平台红外图谱关联探测系统及方法
WO2017041335A1 (zh) 一种全光学波段图谱协同探测动目标的装置和方法
CN101738619B (zh) 双波段红外光学系统
US5345304A (en) Integrated LADAR/FLIR sensor
CN108415097B (zh) 一种多波段红外成像的图谱协同探测系统和方法
WO2013091286A1 (zh) 一种多波段动目标光谱特征探测识别方法和装置
CN103076092B (zh) 一种提高光谱分辨率的干涉成像光谱装置及方法
CN204439211U (zh) 一种超宽波段图谱关联探测装置
CN103777348B (zh) 一种多波段灵巧红外光学系统
CN103323124A (zh) 红外成像光谱仪及其对快速移动目标的超光谱成像方法
CN104897282B (zh) 一种实用型热红外高光谱成像仪载荷系统
CN109655157A (zh) 一种可见光-红外图谱探测装置及方法
CN102519596B (zh) 地球静止轨道高分辨率干涉光谱成像系统
CN208588406U (zh) 一种空间目标三维信息实时探测系统
CN203658669U (zh) 一种多波段灵巧红外光学系统
CN105823558A (zh) 变间隙法布里-珀罗干涉式长波红外双模光谱成像系统
CN109931969A (zh) 一种空间目标三维信息实时探测系统
CN109030358B (zh) 基于同轴腔微波谐振原理的微弱红外信号检测系统与方法
CN110702612B (zh) Offner式偏振光谱一体式空间目标观测装置及空间目标观测方法
CN104165694A (zh) 一种空间调制静态干涉光谱成像仪干涉图光学拼接方法
Hongbo et al. Design and development of a single-photon laser and infrared common aperture optical system
Zhu et al. A novel design of a high performance passive millimeter-wave imager for security

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 15104921

Country of ref document: US

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15874649

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: 15874649

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 15874649

Country of ref document: EP

Kind code of ref document: A1