CN116128931B - Full-spectrum rapid positioning system and method based on mask space regulation and control - Google Patents
Full-spectrum rapid positioning system and method based on mask space regulation and control Download PDFInfo
- Publication number
- CN116128931B CN116128931B CN202310395013.2A CN202310395013A CN116128931B CN 116128931 B CN116128931 B CN 116128931B CN 202310395013 A CN202310395013 A CN 202310395013A CN 116128931 B CN116128931 B CN 116128931B
- Authority
- CN
- China
- Prior art keywords
- mask
- moving object
- self
- light
- luminous
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
-
- 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
- G06T2207/10052—Images from lightfield camera
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P70/00—Climate change mitigation technologies in the production process for final industrial or consumer products
- Y02P70/50—Manufacturing or production processes characterised by the final manufactured product
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Geometry (AREA)
- Testing Of Optical Devices Or Fibers (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention relates to the technical field of optical precision testing, and discloses a full-spectrum rapid positioning system and method based on space regulation and control of a mask plate, wherein the method comprises the following steps: the lens collects the light field of the self-luminous moving object to be positioned, the lens images the light field onto the mask, the mask modulates two-dimensional information of the light field, the modulated light is sent to the lens, the lens converges the light modulated by the mask, the light is converged on a back focus of the lens, a detector positioned on the back focus of the lens detects the converged light in a full spectrum, the predicted centroid position of the self-luminous moving object to be positioned is calculated according to the detection result in the full spectrum, and the predicted centroid position of the self-luminous moving object to be positioned is corrected by utilizing a corresponding relation table of the actual position of the sample self-luminous moving object and the predicted centroid position, so that the corrected centroid position of the object is obtained, and the movement track of the object is obtained. The invention realizes real-time positioning and tracking of the high-speed moving object on the basis of not obtaining the target scene image.
Description
Technical Field
The invention relates to the technical field of optical precision testing, in particular to a full-spectrum rapid positioning system and method based on space regulation and control of a mask plate.
Background
The real-time tracking technology of fast moving target is the process of acquiring the spatial characteristic information of moving object and estimating the track of the object in moving scene. The conventional dynamic target tracking method generally adopts an area array detector with high spatial resolution and high frame frequency to capture a sequence image of a moving target, and then utilizes an image processing algorithm to extract the position and spatial distribution information of the target from the captured sequence image. The accuracy of tracking a fast moving object by the method depends on the quality of an image sequence, and the higher the resolution of an image is, the higher the accuracy of an acquired object motion trail is. On the other hand, the image acquired by the camera is an average of the motion states of the object during the exposure time, so that the rapid motion of the object can cause the image acquired by the camera to generate motion blur.
So far, the imaging speed of some professional high-speed cameras can reach tens of thousands of frames per second, and the problems are well solved. The throughput of image data generated by long-time operation of a high-speed camera is very huge, and huge pressure is generated on data storage, data transmission, image analysis and the like, so that continuous tracking of a moving object is difficult to realize. Also, high speed cameras are either not operable in some non-visible bands or are extremely expensive to manufacture.
The fast moving object tracking method based on image analysis needs to acquire a sequence image of a scene, and the image of the scene contains not only the spatial information of a moving object but also the background information of the scene. The acquisition of the trajectory of a moving object requires only spatial information of the object, accounting for only a small part of the scene information. In other words, most of the background information of the scene information belongs to redundant information. Therefore, the method for acquiring the motion trail of the object can generate waste of information resources. And the traditional method still has a serious problem that the algorithm requirement is high. The current mature image processing algorithm cannot improve the time resolution of tracking moving objects, and the cost of processing a large amount of image data is more expensive.
Disclosure of Invention
In order to solve the defects in the prior art, the invention provides a full-spectrum rapid positioning system and method based on space regulation and control of a mask plate; the method breaks through the technical bottlenecks in the aspects of tracking principle, model and the like, realizes real-time positioning tracking of the high-speed moving object on the basis of not obtaining the target scene image, and provides a new research thought and technical approach for tracking and detecting the high-speed moving object. Meanwhile, the research result of the invention has positive scientific significance and application value for promoting the development of high-speed moving object tracking technology and enriching the research connotation in the field of 'no-image' tracking and detection.
In a first aspect, the invention provides a full spectrum rapid positioning system based on space regulation of a mask plate;
full spectrum quick positioning system based on mask space regulation and control includes: four light paths; each light path includes: the lens, the mask, the lens and the detector are sequentially arranged, and the center points of the lens, the mask, the lens and the detector on each light path are all positioned on the same straight line; the mask plates of the four light paths are distributed in a central symmetry way by taking the center of the view field as a center point;
under each light path, a lens collects a light field of the self-luminous moving object to be positioned, the lens images the light field onto a mask plate, the mask plate modulates two-dimensional information of the light field, the modulated light is sent to the lens, the lens converges the light modulated by the mask plate onto a rear focal point of the lens, a detector positioned at the rear focal point of the lens detects the converged light in a full spectrum, a computer calculates a predicted centroid position of the self-luminous moving object to be positioned according to the detection result in the full spectrum, and corrects the predicted centroid position of the self-luminous moving object to be positioned by utilizing a corresponding relation table of the actual position and the predicted centroid position of a sample self-luminous moving object to obtain a corrected centroid position of the self-luminous moving object to be positioned, so as to obtain a movement track of the self-luminous moving object to be positioned.
In a second aspect, the invention provides a full spectrum rapid positioning method based on space regulation and control of a mask plate;
the full-spectrum rapid positioning method based on the mask space regulation and control adopts the full-spectrum rapid positioning system based on the mask space regulation and control of the first aspect, and comprises the following steps:
acquiring the actual position of a sample when the self-luminous moving object moves; in the moving process of the sample self-luminous moving object, performing space regulation and control on target light field information by using a mask, performing full-spectrum detection on the space-regulated light field by using a detector, calibrating a regulation and control effect to obtain a predicted centroid position of the sample self-luminous moving object, and constructing a corresponding relation table of the actual position of the sample self-luminous moving object and the predicted centroid position;
acquiring a light intensity value of the self-luminous moving object to be positioned after being modulated by a mask plate in the moving process of the self-luminous moving object to be positioned, and processing the light intensity value by adopting a centroid algorithm to acquire a predicted centroid position of the self-luminous moving object to be positioned at the current moment;
and carrying out error correction on the predicted centroid position of the self-luminous moving object to be positioned by adopting a corresponding relation table of the actual position and the predicted centroid position of the sample self-luminous moving object to obtain a corrected centroid position of the self-luminous moving object to be positioned, and further obtaining a motion track of the self-luminous moving object to be positioned.
Compared with the prior art, the invention has the beneficial effects that:
(1) Compared with the traditional target tracking method, the method directly acquires the centroid position information of the target without an intermediate link for acquiring the image information of the target. The redundancy of information acquisition in the traditional optical positioning tracking system is avoided, and the information acquisition efficiency is improved. The invention can acquire the mass center position of the object in one detection process without acquiring the image of a scene, the frame rate of the moving object only depends on the speed of acquiring data by the photoelectric detector and is not influenced by motion blur, and the high frame rate real-time tracking of the moving object can be realized.
(2) The performance advantage of the single-point photoelectric detector is fully exerted. Based on the tracking mode of mask space regulation, only the light intensity value fed back after object modulation is focused during detection, and the information transmission and processing pressure is reduced, so that the occupied memory is smaller and the detection speed is higher. Not only is the photodetector array has the advantages of higher quantum efficiency, lower dark noise, shorter response time and the like compared with the array detector, and has wide application range in weak light detection and invisible light wave bands.
(3) And a full-spectrum focal point detector is adopted to realize full-spectrum detection. The target surface of the full-spectrum focal point detector is formed by splicing ultraviolet band, visible light band and infrared band point detector units, and can be detected in the full-spectrum band without the spectrum band limitation of the traditional photoelectric detector when the target surface is arranged on the back focal surface of the collecting lens.
(4) The mask plate is used as a modulation device, and has the advantages of wide spectrum, high speed and low cost. The mask plate of the quartz substrate has higher transmittance (the ratio of the light intensity value transmitted through the mask plate to the light intensity value irradiated to the surface of the mask plate is used for representing the light transmission capacity of the mask plate) in the ultraviolet region to the near infrared region, and the chromium coating of the mask plate is almost opaque in the ultraviolet region, the visible region and even the near infrared short wave region, so that the wide-spectrum space regulation and control of the ultraviolet region, the visible region and the infrared region can be simultaneously realized in theory. In addition, continuous real-time spatial regulation and control of the light field can be realized through the mask plate, the modulation speed is unlimited, and the positioning speed of a moving object is greatly improved. And the manufacturing cost of the mask plate for light field regulation is low, which is favorable for popularization of the method.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention.
FIG. 1 is a schematic diagram of a system of the present invention;
FIG. 2 is a schematic view of the spatial distribution of four reticles in the present invention;
FIG. 3 is a schematic representation of a reticle design of the present invention;
FIG. 4 is a schematic diagram of a full spectrum focal point detector of the present invention;
FIG. 5 is a schematic diagram of a system calibration process according to the present invention;
FIG. 6 is a schematic diagram showing the correspondence between the actual position of a sample self-luminous moving object and the predicted centroid position;
fig. 7 is a schematic diagram showing the comparison between the centroid position and the actual position of a high-speed moving object according to the present invention.
Detailed Description
Example 1
The embodiment provides a full-spectrum rapid positioning system based on space regulation and control of a mask plate;
as shown in fig. 1, the full spectrum rapid positioning system based on reticle space regulation comprises: four light paths; each light path includes: the lens, the mask, the lens and the detector are sequentially arranged, and the center points of the lens, the mask, the lens and the detector on each light path are all positioned on the same straight line; the mask plates of the four light paths are distributed in a central symmetry way by taking the center of the view field as a center point;
under each light path, a lens collects a light field of the self-luminous moving object to be positioned, the lens images the light field onto a mask plate, the mask plate modulates two-dimensional information of the light field, the modulated light is sent to the lens, the lens converges the light modulated by the mask plate onto a rear focal point of the lens, a detector positioned at the rear focal point of the lens detects the converged light in a full spectrum, a computer calculates a predicted centroid position of the self-luminous moving object to be positioned according to the detection result in the full spectrum, and corrects the predicted centroid position of the self-luminous moving object to be positioned by utilizing a corresponding relation table of the actual position and the predicted centroid position of a sample self-luminous moving object to obtain a corrected centroid position of the self-luminous moving object to be positioned, so as to obtain a movement track of the self-luminous moving object to be positioned.
Further, the preparation process of the mask plate comprises the following steps:
constructing a gray scale map, and up-sampling the gray scale map to obtain an up-sampled high-resolution gray scale map;
processing the up-sampling high-resolution gray level map to obtain a binary distribution modulation mode mask pattern;
covering a chromium coating on a quartz substrate, and etching a mask pattern on the chromium coating;
and cutting out the area covered by the mask pattern to obtain the mask.
Further, the construction of the gray scale map designs the spatial resolution as the gray scale value of the pixels in the gray scale map by taking the element values in the matrix as the gray scale values according to the matrix distributionGray scale map of pixels.
Further, the gray level image is up-sampled, and bicubic interpolation algorithm is adopted for the up-samplingThe gray-scale map of the pixel is up-sampled, one original pixel is composed of +.>Sub-pixel representation, resulting in spatial resolution +.>Up-sampled gray scale of pixels.
Further, the processing of the up-sampling high-resolution gray level image to obtain a binary distribution modulation mode mask pattern is to process the up-sampling gray level image by using a Floyd-Steinberg error diffusion dithering algorithm to obtain a binary image, and the binary image is used as the mask pattern.
It should be appreciated that continuous light field modulation over the full spectrum can be achieved using the binary image as a reticle pattern.
Further, the mask plate is divided into four mask plates, namely a first mask plate, a second mask plate, a third mask plate and a fourth mask plate;
the first mask plateAnd a third mask->Are axisymmetrically arranged along the vertical direction;
the second mask plateAnd a fourth mask->Is axisymmetrically arranged along the horizontal direction;
the pixel values of the first mask plate gradually increase from left to right, the pixel values of the second mask plate gradually increase from bottom to top, the pixel values of the third mask plate gradually increase from right to left, and the pixel values of the fourth mask plate gradually increase from top to bottom, so that the complementation of the areas of the mask plates in the same direction is ensured.
The beneficial effects of the technical scheme are as follows: through the structural light modulation of the four light path masks, continuous and simultaneous coded modulation of self-luminous object images is realized, and the modulation speed is unlimited and motion blur is avoided.
Further, the mask plate carries out two-dimensional information modulation on the light field, namely light field modulation is realized by the mask plate on the light field image acquired by the lens.
Further, the target surface of the detector is formed by splicing an ultraviolet band point detector, a visible light band point detector, a middle infrared band point detector and a near infrared band point detector; the detector is arranged on the back focal plane of the collecting lens and is used for detecting the full spectrum of the light modulated by the mask plate.
Further, the corresponding relation table of the actual position of the sample self-luminous moving object and the predicted centroid position comprises the following specific construction processes:
imaging the target light fields in the four light paths onto corresponding masks through respective lenses, modulating two-dimensional information of the target light fields by the masks, and converging light modulated by the masks to a rear focus of the lens through the lens;
the detector is positioned at the back focus of the lens and is used for detecting the full spectrum of the light modulated by the mask plate;
establishing a two-dimensional Cartesian coordinate system on the mask, and controlling a sample self-luminous moving object to uniformly move Q positions along the direction from low to high of the pixel value of the mask; q is a positive integer; q is greater than 2;
according to the Q groups of experimental data acquired by the detector, calculating a predicted centroid position of the sample self-luminous moving object at the ith position by adopting a centroid algorithm;
and constructing a mapping relation table of the actual position of the sample self-luminous moving object and the predicted centroid position according to the predicted centroid position and the actual position of the ith position of the sample self-luminous moving object.
Further, the correcting the predicted centroid position of the self-luminous moving object to be positioned by using the corresponding relation table of the actual position and the predicted centroid position of the sample self-luminous moving object to obtain the corrected centroid position of the self-luminous moving object to be positioned includes:
according to the predicted centroid position of the self-luminous moving object to be positioned, searching the actual position of the sample self-luminous moving object corresponding to the predicted centroid position of the self-luminous moving object to be positioned from the relation table, and taking the actual position of the sample self-luminous moving object obtained by table searching as the corrected centroid position of the self-luminous moving object to be positioned.
In order to realize full-spectrum real-time tracking detection of fast moving objects, a solution idea is mainly sought from the perspective of never obtaining scene images. And a modulating pattern model is established by adopting a spatial dithering method according to the actual condition of the mask plate by adopting the thought of spatial regulation of the mask plate. According to the principle of a multi-optical path system, a photoelectric detection acquisition system with four optical paths is established, and full-band real-time tracking of a moving object is realized. And designing a full-spectrum focal point detector covering ultraviolet light, visible light and infrared light, and carrying out full-spectrum detection on the light field subjected to space regulation and control of the mask plate by using a focus behind each collecting lens. And then, developing a system calibration process, establishing a mapping relation between the actual position of the sample self-luminous moving object and the predicted centroid position, developing a system compensation algorithm, and improving the accuracy of tracking the moving object by the system. In principle, the advantages of the system structure, instantaneity, wide spectrum and tracking precision of fast moving object tracking based on mask modulation are verified, finally experimental research is carried out, and the working performances of the system scheme such as co-phase detection precision, efficiency and detection stability are verified through experiments.
Fig. 1 is a schematic system diagram of the invention. The imaging unit, the structural light modulation unit and the data acquisition and processing unit are organically integrated together to form a compact and light high-speed moving object tracking and detecting system. Light emitted by a moving object in a scene is imaged onto a mask plate through a lens, two-dimensional information modulation is carried out on an image formed by a target object through an optical mask plate, and a transmitted light signal is subjected to total light intensity value acquisition through a collecting lens by a full-spectrum focal point detector. In the system, the distribution of the four light paths is shown in fig. 2, and the masks of the four light paths are symmetrically distributed by taking the center of a field of view as the center, so that the consistency of mask modulation areas in complementary light paths is ensured.
The proposed reticle modulation pattern is a gray scale pattern, and is limited by the processing technique, and the reticle carries a binary pattern generated by a spatial dithering algorithm. The specific process is shown in FIG. 3, for original spatial resolutionThe rate is as followsThe gray-scale map of pixels is first up-sampled by bicubic interpolation algorithm, one original pixel is composed of +.>Up-sampled pixel representation, resulting in spatial resolution +.>Is used for the up-sampling of the gray-scale map. And then applying the Floyd-Steinberg error diffusion dithering method and upsampling the gray level diagram to obtain a binary distribution modulation mode, so as to realize continuous light field regulation and control in the whole spectrum range.
As shown in fig. 4, a full-spectrum focal point detector covering ultraviolet light, visible light and infrared light is designed, and the number of typical splicing units of the target surface of the full-spectrum focal point detector is four, and the full-spectrum focal point detector is formed by splicing ultraviolet, visible light, near infrared and middle infrared band point detector units. The number of the splicing units can be adjusted according to the requirement, and the coverage wave band can be increased. The full-spectrum focal point detector is arranged on the back focal plane of the collecting lens, and the full-spectrum detection is carried out on the light modulated by the mask plate by utilizing the space broadening of the back focal point.
Around the system framework, a fast moving object tracking system comprises the following coding, detection and reconstruction processes:
the coding process, i.e. the modulation process of the light field. An object is imaged onto a reticle. The object is imaged onto the reticle through the lens, and the image of the target object falls within the encoded region of the reticle. Since the encoded region of the reticle is square, the side dimension of the object image should be no greater than the lateral width of the encoded region of the reticle. The four light paths of the system are symmetrically distributed by taking the center of the field of view as the center. In the system, the offset of the optical path from the center of the field of view causes the object to be imaged onto the reticle at a location offset from the center of the reticle.
To improve the accuracy of tracking the object, the first mask plates are respectively used forAnd a third mask->Symmetrically distributed along the y direction, and a second mask plate +.>And a fourth mask->Symmetrically distributed along x direction>And representing the pixel coordinates of the mask modulation pattern, thereby ensuring that the areas of the mask in the same direction where the object is located are complementary. The four-light path mask plate is used for carrying out the structural light modulation, the system realizes the continuous and simultaneous code modulation of the object image, the modulation speed is not upper bound and there is no motion blur.
And detecting and photoelectrically converting the optical signal. The optical signal is collected by using a lens, the lens is selected, so that the light intensity information after being periodically modulated by an object is converged on the target surface of the full-spectrum focal point detector, the detector detects at high speed in the full-spectrum range, the collected light intensity signal is converted into a current signal and output, the signal is subjected to analog-to-digital conversion, the input current signal is sampled at equal intervals by the analog-to-digital conversion equipment at a fixed sampling rate, and the analog signal is converted into a digital signal, so that the detection of the optical signal in the full-spectrum range is realized.
The reconstruction process comprises the steps of system calibration, data processing and calculation. And (3) establishing a mapping relation table of the actual position and the measured centroid position of the self-luminous moving object of the known sample through system calibration, and correcting the centroid position coordinate of the self-luminous moving object to be positioned through a table look-up method. Firstly, a two-dimensional Cartesian coordinate system is established on a mask plate, and self-luminous objects are controlled by a computer to modulate patterns along the mask plate respectivelyAnd->Uniformly moving from black to white81 positions; 81 groups of experimental data acquired by the detector are input into a computer through analog-to-digital conversion, and the measured centroid position of the actual position of each sample self-luminous moving object is calculated according to a centroid algorithm; and establishing a mapping relation table of the actual position of the sample self-luminous moving object and the measured centroid position by adopting an interpolation algorithm according to the measured centroid position corresponding to the actual position of each sample self-luminous moving object, as shown in fig. 6. FIG. 5 is a schematic diagram of a system calibration process according to the present invention; fig. 6 is a schematic diagram of a correspondence between an actual position and a predicted centroid position of a sample self-luminous moving object according to the present invention, wherein X in fig. 6 represents a correspondence between a predicted centroid position in a horizontal direction and an actual position, and Y in fig. 6 represents a correspondence between a predicted centroid position in a vertical direction and an actual position.
For the unmeasured area, the mask modulation effect is evaluated by adopting a linear differential algorithm, the corresponding relation between the actual positions of two adjacent points and the measured centroid position is measured in the calibration process, the centroid position corresponding to the actual positions between the two points is obtained by adopting an interpolation algorithm, the linear interpolation algorithm regards the change of the corresponding relation between the two points as linear change, the corresponding actual position is obtained by the measured centroid position, and the centroid position is corrected by adopting the system calibration result. And (5) bringing the data acquired simultaneously into an integral formula to preliminarily obtain the original mass center position of the object. And combining the mass center position of the object with the system calibration mapping relation, and obtaining the corrected mass center position of the object through an algorithm. The corrected object centroid position is compared with the actual position of the object as shown in fig. 7.
The mask modulation pattern designed according to the tracking principle has the spatial resolution ofWhile in the invention, an optical mask is used as a modulation device, the gray scale modulation difficulty is high, and the gray scale modulation pattern is difficult to bear. Therefore, the modulation pattern is converted from the gray scale to the binary image using the spatial dithering method, and the process is as shown in fig. 3. For the original spatial resolution +.>The gray level map of a pixel is first up-sampled by bicubic interpolation algorithm, one original pixel is composed ofUp-sampled pixel representation, resulting in spatial resolution +.>Is used for the up-sampling of the gray-scale map. Bicubic interpolation is a more complex way of interpolation that creates smoother image edges than bilinear interpolation. In this approach, the value of the function f at point (x, y) can be obtained by a weighted average of the nearest sixteen sample points in a rectangular grid, where two polynomial interpolation cubic functions are required, one for each direction. The upsampled gray pattern is then converted into a binary distributed modulation pattern by dithering to represent a larger gray scale range with fewer gray scales by the Floyd-Steinberg error diffusion dithering method. The Floyd-Steinberg method is actually an error dithering method, which dithers the color information of the present pixel to surrounding pixels by unevenly distributed quantization errors, meaning that the center pixel is rounded to 0 or 1 first, and then the residual is added to its surrounding pixels.
Aiming at the important requirements and the technical bottleneck problem of high-speed moving object tracking and monitoring, the invention researches a mask-based full-band tracking method of the high-speed moving object, attempts to break through the technical bottleneck in the aspects of tracking principle, model and the like, realizes real-time positioning tracking of the high-speed moving object on the basis of not acquiring a target scene image, and provides a new research thought and technical approach for tracking and detecting the high-speed moving object.
By adopting the idea of space regulation of the mask plates, two-dimensional information encoding is carried out on the light field through a plurality of mask plates, a four-light-path high-speed moving object system model is constructed, and the feasibility of the moving object tracking detection method based on the structural light modulation is demonstrated through theoretical analysis. By designing the photoelectric detector, the infrared, visible and ultraviolet band detector units are integrated, so that the tracking and positioning of the whole band are realized. And then, designing a mask modulation pattern based on a space optical dithering method, providing a system calibration scheme in consideration of the influence of systematic errors such as lens vignetting effect, mask modulation pattern design, optical paths and the like on a system, establishing a corresponding relation table by measuring the predicted centroid position and the actual position of a self-luminous object, and evaluating an uncalibrated area through an interpolation algorithm. The centroid position of the self-luminous moving object to be positioned is obtained through a centroid algorithm, and is combined with a corresponding relation table of the actual position of the sample self-luminous moving object and the predicted centroid position, so that the corrected centroid position is obtained, the influence of a system error is reduced, and the tracking precision is greatly improved. The invention verifies the advantages of the system structure, the detection speed, the detection precision and the spectrum range based on the structural light modulation in principle. Finally, experimental study is carried out, and the real-time performance, the accuracy and other working performances of the system scheme are verified through experiments.
The mask plate is used as a modulation device, so that the mask plate can realize wide-spectrum light field modulation, the limit of the DMD modulation speed in the prior art is broken through, and the positioning speed is only dependent on the speed of the detector for collecting the light intensity value and is far greater than the modulation speed 22kHz of the DMD.
The full-spectrum focal point detector is used, the traditional photoelectric detector is limited and limited in spectrum and can only detect in a specific wave band range, and the full-spectrum focal point detector integrates ultraviolet band, visible light band and infrared wave band point detector units and can detect in the full-spectrum range.
The multi-light path detector system is used for detecting the light intensity value modulated by the scene at the same time, so that the positioning of the moving object can be realized in one detection, the influence of motion blur is avoided, and the positioning speed of the moving object is greatly improved.
In the system calibration process, the limitation and light path deviation of a multi-light path system, the vignetting effect of a lens and the design error of a mask plate have great influence on a positioning result, and a corresponding relation table of the predicted centroid position and the actual position of a self-luminous object is established through the system calibration process, so that the predicted centroid position is corrected, and the positioning precision is improved.
Reticle design, the reticle carrying a modulation pattern formed byAnd a binary image generated by a spatial dithering algorithm. The binary image is obtained by the spatial resolution ofThe spatial resolution of the gray level map of the pixel obtained by up-sampling and error dithering algorithm isA binary image of pixels, an original pixel is composed of +.>The number of pixels is expressed, so that the change of the gray level is divided into larger gray levels, the accuracy of space regulation of the mask plate is improved, and the accuracy of positioning of a moving object is improved.
Example two
The embodiment provides a full-spectrum rapid positioning method based on space regulation and control of a mask plate;
a full-spectrum rapid positioning method based on mask space regulation is provided, which adopts the full-spectrum rapid positioning system based on mask space regulation in the first embodiment, and comprises the following steps:
s201: acquiring the actual position of a sample when the self-luminous moving object moves; in the moving process of the sample self-luminous moving object, performing space regulation and control on target light field information by using a mask, performing full-spectrum detection on the space-regulated light field by using a detector, calibrating a regulation and control effect to obtain a predicted centroid position of the sample self-luminous moving object, and constructing a corresponding relation table of the actual position of the sample self-luminous moving object and the predicted centroid position;
s202: acquiring a light intensity value of the self-luminous moving object to be positioned after being modulated by a mask plate in the moving process of the self-luminous moving object to be positioned, and processing the light intensity value by adopting a centroid algorithm to acquire a predicted centroid position of the self-luminous moving object to be positioned at the current moment;
s203: and carrying out error correction on the predicted centroid position of the self-luminous moving object to be positioned by adopting a corresponding relation table of the actual position and the predicted centroid position of the sample self-luminous moving object to obtain a corrected centroid position of the self-luminous moving object to be positioned, and further obtaining a motion track of the self-luminous moving object to be positioned.
Further, the step S201: acquiring the actual position of a sample when the self-luminous moving object moves; in the motion process of a sample self-luminous moving object, performing space regulation and control on target light field information by using a mask, performing full-spectrum detection on a space regulated and controlled light field by using a detector to obtain a predicted centroid position of the sample self-luminous moving object, and constructing a corresponding relation table of the actual position and the predicted centroid position of the sample self-luminous moving object, wherein the method specifically comprises the following steps:
s201-1: imaging the target light fields in the four light paths onto corresponding masks through respective lenses, modulating two-dimensional information of the target light fields by the masks, and converging light modulated by the masks to a rear focus of the lens through the lens;
s201-2: the detector is positioned at the back focus of the lens and is used for detecting the full spectrum of the light modulated by the mask plate;
s201-3: establishing a two-dimensional Cartesian coordinate system on the mask, and controlling a sample self-luminous moving object to uniformly move Q positions along the direction from low to high of the pixel value of the mask; q is a positive integer; q is greater than 2;
s201-4: according to the Q groups of experimental data acquired by the detector, calculating a predicted centroid position of the sample self-luminous moving object at the ith position by adopting a centroid algorithm;
s201-5: and constructing a mapping relation table of the actual position of the sample self-luminous moving object and the predicted centroid position according to the predicted centroid position and the actual position of the ith position of the sample self-luminous moving object.
And the detector receives the light intensity values, sequentially records the light intensity values, modulates the light signals, and finally restores the spatial information of the self-luminous moving object through a centroid algorithm and a mapping relation table of the actual position and the predicted centroid position of the sample self-luminous moving object.
Further, S201-4: calculating a predicted centroid position of the sample self-luminous moving object at the ith position by adopting a centroid algorithm, wherein the method specifically comprises the following steps of:
adjustment of the jth reticle at the ith positionLight intensity value collected by detector after pattern modulationExpressed as:
wherein, the liquid crystal display device comprises a liquid crystal display device,for the gray distribution function of the target scene, +.>For the matrix distribution of the j-th reticle modulation pattern at the i-th position +.>Embody +.>The result of two-dimensional integration, j is the mode index,/>Respectively representing the abscissa and the ordinate of the pixel point;
the modulation pattern matrix distribution corresponding to the mask plate is respectively as follows:
；
；
；
wherein, the liquid crystal display device comprises a liquid crystal display device,representing the corresponding modulation pattern matrix distribution of the first reticle,/->Representing the modulation pattern matrix distribution corresponding to the second mask plate; />Is an intermediate variable;
the four light paths are symmetrically distributed by taking the center of the field of view as the center and are matched with the mask plateAnd->The modulation function of a symmetrically distributed reticle is expressed as:
；
；
wherein, the liquid crystal display device comprises a liquid crystal display device,representing the matrix distribution of the modulation pattern corresponding to the third reticle,/->Representing the modulation pattern matrix distribution corresponding to the fourth mask plate;
if the sample self-luminous moving object moves along the first reticle pixel from the low to the high direction, the predicted centroid position of the sample self-luminous moving object is expressed as:
；
；
wherein, the liquid crystal display device comprises a liquid crystal display device,representation->Direction centroid position->Representation->Centroid position of direction, ++>Representing the intensity value received by the detector after modulation of the first reticle, < >>Representing the intensity value received by the detector after modulation of the second reticle, < >>The representation is an intermediate variable;
，/>is the total light intensity value of the scene.
Four light paths are symmetrically distributed by taking the center of a field of view as the center, and the mask plateAnd->、/>Andsymmetrically distributed in the transverse direction and the axial direction respectively, and taking the sum of two mask modulation light intensity values which are symmetrically distributed as the total light intensity value of a scene, the centroid position is expressed as:
；
；
wherein, the liquid crystal display device comprises a liquid crystal display device,representation->Direction centroid position->Representation->Centroid position of direction, ++>Representing the intensity value received by the detector after modulation of the first reticle, < >>Representing the intensity value received by the detector after modulation of the second reticle, < >>Representing the intensity value received by the detector after the third reticle modulation, < >>Representing the intensity value received by the detector after the fourth reticle modulation.
It should be appreciated that for a gray scale image, we can build a two-dimensional cartesian coordinate system based on the coordinates of the pixels, and then a pair of gray scale maps can be represented by a two-dimensional gray scale density function, so that the centroid position can be obtained by a two-dimensional integration method. For a target scene, a mask plate carrying a modulation pattern is adopted to carry out structural light modulation on the light field, and the modulated light intensity value is collected by a detector, so that the integration process of the scene is realized.
It is to be understood that the four light paths are set to perform structural light modulation on the scene at the same time, so that the centroid position of the moving object can be obtained in one detection process, the influence of motion blur is eliminated, and the real-time tracking of the moving object is realized. In consideration of the influence of lens vignetting effect and light path deviation, the modulation effect of the mask plate has errors, the ideal effect cannot be achieved, and the light path deviation leads to the deviation of the image on the mask plate from the center position, so that four light paths in the system are symmetrically distributed by taking the center of a view field as the center, the total light intensity of a scene is respectively represented by the sum of filter integral values of the symmetrical distribution in the transverse direction and the axial direction, the same modulation effect of the mask plate of the symmetrical light paths is ensured, and the accuracy of the centroid position is greatly improved. The mask is used as a light field regulating device, so that continuous real-time modulation of a light field can be realized, the modulation speed is unlimited, and the spectrum range is wide.
Further, the step S202: in the motion process of the self-luminous moving object to be positioned, the light intensity value of the self-luminous moving object to be positioned after being modulated by the mask plate is obtained, and the specific calculation step is the same as the specific calculation step of S201-4, and is not repeated here.
Further, the step S202: the light intensity value is processed by adopting a centroid algorithm to obtain the predicted centroid position of the self-luminous moving object to be positioned at the current moment, and the specific calculation step is the same as the specific calculation step of S201-5 and is not repeated here.
Further, the step S203: performing error correction on the predicted centroid position of the self-luminous moving object to be positioned by adopting the corresponding relation between the actual position and the predicted centroid position of the sample self-luminous moving object to obtain the corrected centroid position of the self-luminous moving object to be positioned, and further obtaining the motion trail of the self-luminous moving object to be positioned, specifically comprising the following steps:
according to a table look-up method, according to the predicted centroid position of the self-luminous moving object to be positioned, the actual position of the sample self-luminous moving object is found out from the corresponding relation between the actual position of the sample self-luminous moving object and the predicted centroid position, and the actual position of the sample self-luminous moving object is taken as the corrected centroid position of the self-luminous moving object to be positioned, so that the movement track of the self-luminous moving object to be positioned is obtained.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (7)
1. Full spectrum quick positioning system based on mask space regulation and control, characterized by including: four light paths; each light path includes: the lens, the mask, the lens and the detector are sequentially arranged, and the center points of the lens, the mask, the lens and the detector on each light path are all positioned on the same straight line; the mask plates of the four light paths are distributed in a central symmetry way by taking the center of the view field as a center point;
the preparation process of the mask plate comprises the following steps: constructing a gray scale map, and up-sampling the gray scale map to obtain an up-sampled high-resolution gray scale map; processing the up-sampling high-resolution gray level map to obtain a binary distribution modulation mode mask pattern; covering a chromium coating on a quartz substrate, and etching a mask pattern on the chromium coating; cutting out the area covered by the mask pattern to obtain a mask;
the target surface of the detector is formed by splicing an ultraviolet band point detector, a visible light band point detector, a middle infrared band point detector and a near infrared band point detector; the detector is arranged on the back focal plane of the collecting lens and is used for detecting the full spectrum of the light modulated by the mask plate;
under each light path, a lens collects a light field of the self-luminous moving object to be positioned, the lens images the light field onto a mask plate, the mask plate modulates two-dimensional information of the light field, the modulated light is sent to the lens, the lens converges the light modulated by the mask plate onto a rear focal point of the lens, a detector positioned at the rear focal point of the lens detects the converged light in a full spectrum, a computer calculates a predicted centroid position of the self-luminous moving object to be positioned according to the detection result in the full spectrum, and corrects the predicted centroid position of the self-luminous moving object to be positioned by utilizing a corresponding relation table of the actual position and the predicted centroid position of a sample self-luminous moving object to obtain a corrected centroid position of the self-luminous moving object to be positioned, so as to obtain a movement track of the self-luminous moving object to be positioned.
2. The rapid positioning system of full spectrum based on space control of mask plate as set forth in claim 1, wherein the construction of gray scale image designs spatial resolution as gray scale value of pixels in gray scale image by taking element value in matrix as gray scale value of pixels in gray scale image according to matrix distributionGray scale map of pixels.
3. The rapid full-spectrum positioning system based on mask space regulation and control as set forth in claim 1, wherein said up-sampling of gray level map adopts bicubic interpolation algorithm to perform the up-sampling of gray level mapThe gray-scale map of the pixel is up-sampled, one original pixel is composed of +.>Sub-pixel representation, resulting in spatial resolution +.>Up-sampled gray scale of pixels.
4. The full-spectrum rapid positioning system based on mask space regulation and control according to claim 1, wherein the processing of the up-sampling high-resolution gray scale pattern to obtain a binary distribution modulation mode mask pattern is to process the up-sampling gray scale pattern by adopting an error diffusion dithering algorithm to obtain a binary pattern, and the binary pattern is used as the mask pattern.
5. The full-spectrum rapid positioning system based on mask space regulation and control according to claim 1, wherein the mask is divided into four masks, namely a first mask, a second mask, a third mask and a fourth mask;
the first mask plateAnd a third mask->Are axisymmetrically arranged along the vertical direction;
the second mask plateAnd a fourth mask->Is axisymmetrically arranged along the horizontal direction;
the pixel values of the first mask plate gradually increase from left to right, the pixel values of the second mask plate gradually increase from bottom to top, the pixel values of the third mask plate gradually increase from right to left, and the pixel values of the fourth mask plate gradually increase from top to bottom, so that the complementation of the areas of the mask plates in the same direction is ensured.
6. The full-spectrum rapid positioning system based on mask space regulation and control of claim 1, wherein the corresponding relation table of the actual position and the predicted centroid position of the sample self-luminous moving object comprises the following specific construction processes:
imaging the target light fields in the four light paths onto corresponding masks through respective lenses, modulating two-dimensional information of the target light fields by the masks, and converging light modulated by the masks to a rear focus of the lens through the lens;
the detector is positioned at the back focus of the lens and is used for detecting the full spectrum of the light modulated by the mask plate;
establishing a two-dimensional Cartesian coordinate system on the mask, and controlling a sample self-luminous moving object to uniformly move Q positions along the direction from low to high of the pixel value of the mask; q is a positive integer; q is greater than 2;
according to the Q groups of experimental data acquired by the detector, calculating a predicted centroid position of the sample self-luminous moving object at the ith position by adopting a centroid algorithm;
and constructing a mapping relation table of the actual position of the sample self-luminous moving object and the predicted centroid position according to the predicted centroid position and the actual position of the ith position of the sample self-luminous moving object.
7. The full-spectrum rapid positioning method based on the mask space regulation is characterized by comprising the following steps of:
acquiring the actual position of a sample when the self-luminous moving object moves; in the moving process of the sample self-luminous moving object, performing space regulation and control on target light field information by using a mask, performing full-spectrum detection on the space-regulated light field by using a detector, calibrating a regulation and control effect to obtain a predicted centroid position of the sample self-luminous moving object, and constructing a corresponding relation table of the actual position of the sample self-luminous moving object and the predicted centroid position;
acquiring a light intensity value of the self-luminous moving object to be positioned after being modulated by a mask plate in the moving process of the self-luminous moving object to be positioned, and processing the light intensity value by adopting a centroid algorithm to acquire a predicted centroid position of the self-luminous moving object to be positioned at the current moment;
and carrying out error correction on the predicted centroid position of the self-luminous moving object to be positioned by adopting a corresponding relation table of the actual position and the predicted centroid position of the sample self-luminous moving object to obtain a corrected centroid position of the self-luminous moving object to be positioned, and further obtaining a motion track of the self-luminous moving object to be positioned.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310395013.2A CN116128931B (en) | 2023-04-14 | 2023-04-14 | Full-spectrum rapid positioning system and method based on mask space regulation and control |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310395013.2A CN116128931B (en) | 2023-04-14 | 2023-04-14 | Full-spectrum rapid positioning system and method based on mask space regulation and control |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116128931A CN116128931A (en) | 2023-05-16 |
| CN116128931B true CN116128931B (en) | 2023-07-18 |
Family
ID=86310343
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310395013.2A Active CN116128931B (en) | 2023-04-14 | 2023-04-14 | Full-spectrum rapid positioning system and method based on mask space regulation and control |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116128931B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101021694A (en) * | 2007-03-06 | 2007-08-22 | 上海微电子装备有限公司 | Aligning system and aligning method based on image technique |
| CN106404713A (en) * | 2016-11-23 | 2017-02-15 | 重庆大学 | Double-detector micro near-infrared spectrometer with whole spectral band of 800-2,500nm |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150234102A1 (en) * | 2012-08-20 | 2015-08-20 | Drexel University | Dynamically focusable multispectral light field imaging |
| WO2014190027A1 (en) * | 2013-05-22 | 2014-11-27 | Massachusetts Institute Of Technology | Methods, systems, and apparatus for imaging spectroscopy |
| US10942116B2 (en) * | 2017-10-09 | 2021-03-09 | Photothermal Spectroscopy Corp. | Method and apparatus for enhanced photo-thermal imaging and spectroscopy |
| CN109000591B (en) * | 2018-06-26 | 2020-07-24 | 中国科学院苏州生物医学工程技术研究所 | Eccentricity difference measuring instrument |
| CN109104582B (en) * | 2018-08-14 | 2020-08-18 | 西安工业大学 | Real-time super-resolution polarization infrared photoelectric imaging method |
| CN110262044B (en) * | 2019-06-18 | 2020-06-12 | 天津大学 | Non-diffraction two-dimensional optical lattice period adjusting system based on zoom lens |
| CN111198037A (en) * | 2020-01-12 | 2020-05-26 | 杭州电子科技大学 | Rapid multispectral imaging device |
| CN111579521B (en) * | 2020-05-06 | 2021-10-15 | 中国科学院沈阳自动化研究所 | Terahertz compression imaging optimization method and system based on data selection |
| CN113033723B (en) * | 2021-03-08 | 2023-06-16 | 山东大学 | Annular mask, light field regulation and control method, single-pixel imaging method and system |
| CN114859377B (en) * | 2022-04-22 | 2023-06-09 | 中国科学院合肥物质科学研究院 | Method and equipment for capturing single-pixel imaging of moving target in real time |
| CN115561979A (en) * | 2022-09-30 | 2023-01-03 | 华南理工大学 | Moire alignment device and method for projection lithography |
-
2023
- 2023-04-14 CN CN202310395013.2A patent/CN116128931B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101021694A (en) * | 2007-03-06 | 2007-08-22 | 上海微电子装备有限公司 | Aligning system and aligning method based on image technique |
| CN106404713A (en) * | 2016-11-23 | 2017-02-15 | 重庆大学 | Double-detector micro near-infrared spectrometer with whole spectral band of 800-2,500nm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116128931A (en) | 2023-05-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107092752B (en) | A kind of optical camera simulation imaging method and system based on ray tracing | |
| JP5519460B2 (en) | Apparatus and method for high dynamic range imaging using spatially varying exposures | |
| CN105589210B (en) | A kind of digitlization synthetic aperture imaging method based on pupil modulation | |
| US9438793B2 (en) | Image forming apparatus, image forming method and image sensor | |
| CN111707413B (en) | Centroid detection method based on single-pixel detector | |
| CN110471082A (en) | Single pixel laser calculates imaging device and method | |
| CN112985600B (en) | Spectrum coding imaging system and method based on diffraction | |
| US8462179B2 (en) | Scene independent method for image formation in lenslet array imagers | |
| CN116128931B (en) | Full-spectrum rapid positioning system and method based on mask space regulation and control | |
| CN110455418B (en) | Super-structure lens array and method for correcting non-uniformity of infrared detector | |
| CN113670456B (en) | Wavefront restoration method realized by Hartmann wavefront sensor with adjustable spatial resolution | |
| CN113188660B (en) | Novel snapshot type polarization spectrum imaging system with adjustable multidimensional parameters | |
| CN114659634A (en) | Miniature snapshot type compressed spectrum imaging detection device and detection method | |
| CN113884234A (en) | Complementary single-pixel centroid detection system and method | |
| CN112113661B (en) | Deep learning type snapshot spectrum imaging device and detection method thereof | |
| CN109557070A (en) | A kind of Raman Imaging System based on space encoding light | |
| CN113033723A (en) | Annular mask, light field regulation and control method, single-pixel imaging method and system | |
| CN109781260B (en) | Ultra-compact snapshot type polarization spectrum imaging detection device and detection method | |
| CN102158659A (en) | A method and an apparatus for difference measurement of an image | |
| CN114186664B (en) | Mode wavefront restoration method based on neural network | |
| CN112985601B (en) | Long-focus spectral coding imaging system and method based on diffraction | |
| CN105473984A (en) | System architecture for a constant footprint, constant GSD, constant spatial resolution line scanner | |
| CN105761212A (en) | Hadamard transform high sensitivity imaging method based on DMD encoding | |
| CN114323310B (en) | High-resolution Hartmann wavefront sensor | |
| CN113890997B (en) | High dynamic range compressed sensing imaging system and method based on random dithering |
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 |