Multispectral object detection technology has important application prospects in the fields of autonomous driving and so on. Conventional multispectral object detection algorithm rely solely on deep neural networks to learn multispectral image sample information, lacking the guidance of prior knowledge, and not fully utilizing infrared, visible, and other spectral information, resulting in decreased accuracy of object detection in complex scenes. To address this problem, this paper proposes an object detection algorithm based on infrared visible sample augmentation and illumination guidance. The algorithm adopts the MMDetection framework and extracts multispectral object features based on a designed sample augmentation method based on the fusion of positive and negative samples in multispectral images. Based on a designed adaptive weight allocation method guided by illumination, it enhances the algorithm's adaptability to the lighting environment. Finally, through the design of a multi-task loss function, it achieves high-precision and robust object detection in complex scenes. Experimental results on datasets such as FLIR and M3FD show that the proposed algorithm has significant advantages over comparative algorithms such as CFR_3 and GAFF in terms of average detection precision.
KEYWORDS: Modulation, Data transmission, Transmitters, Transceivers, Receivers, Demodulation, Electrodes, Signal attenuation, Analog to digital converters
Conductive intracardiac communication (CIC) can be utilized for the synchronization of multi-chamber leadless pacemakers (LLPMs) to overcome complications arising from traditional pacemaker lead connections. Current CIC methods are mainly based on pulse modulation (PM) and On-Off Keying (OOK), which suffer from high power consumption and poor interference resilience, with transmitter output bit energy (Ebit) exceeding 5 pJ. Addressing these concerns, we proposed a method of conductive intracardiac communication based on Gaussian Minimum Shift Keying (GMSK). In the transmitter, a power-optimized GMSK modulation method is employed for signal transmission. In the receiver, a variable gain amplifier is utilized for CIC signal reception, and signal recovery is achieved through low-power demodulation and bit synchronization methods. A prototype transceiver was designed for measuring the bit error ratio (BER) and transmitter output Ebit of GMSK and OOK methods in in-vitro experiments using porcine hearts as channels. The transceiver was connected to the right atrium and right ventricle via stainless steel needle electrodes. At data rates of 75-500 kbps and a BER of 1e-4, the average Ebit for OOK ranged from 0.06-0.61 pJ, whereas for GMSK, it was lower, ranging from 0.02-0.35 pJ. This study demonstrates that LLPM achieving reliable CIC using GMSK at transmission powers below 35 pW is feasible in practical channel conditions. The proposed GMSK-based CIC method is more preferable to PM and OOK due to its higher reliability and lower transmission power requirements.
Event stream has been used in various vision tasks due to the low latency and high dynamic range of event camera. However, because of the temporal dynamic of event stream, convolution neural networks (CNNs) are difficult to effectively extract features from event streams to achieve object tracking tasks. Besides, SNNs is suitable for processing data with temporal information because of its spiking delivery mechanism and membrane potential accumulation over time. In this work, we propose a Hybrid Neural Network (HNNet) to achieve effective event-based single object tracking tasks by combining the advantages of SNNs and Swin-Transformer. For higher feature expression ability of SNNs, we adopt the Swin-Transformer to extract features from sparse event stream. Then we use these features to modulate the threshold of SNNs neurons. What’s more, for improving tracking performance for both special and temporal features, a cross-modality fusion module is designed to fuse the two features extracted by the Swin-Transformer and SNNs. We conduct extensive experiments on three public event-based datasets (FE240, FE108, and VisEvent) and our tracker outperforms other trackers maximum at 1.1% and 6.8% in terms of area under curve (AUC) scores and precision rate respectively.
Infrared target detection and tracking technology has been widely used in the fields of transportation, medical, safety and military affairs, etc. However, there stills exists some challenges in infrared target detection and tracking, such as dim small target, complex background, target occlusion and appearance changes, etc. On the other hand, as the most effective bio-intelligence system, Human Visual System (HVS) has significant advantages in image processing. In this paper, several brain-inspired models (including lateral inhibition, receptive field, synchronous burst, visual attention, and cognitive memory) and Deep Neural Networks (DNNs) have been studied. Furthermore, the relevant mathematical models are established, the corresponding algorithms are proposed, and the comparison experiments are conducted. In summary, applying the brain-inspired models and DNNs to the infrared target detection and tracking is beneficial to achieve the accurate infrared target detection and robust tracking under complex conditions.
Quantification of image clutter plays an important role in predicting target acquisition performances of a photoelectric imaging system due to the strong effect of optoelectronic image clutter. Accuracy in predicting the targeting performance of previous reported clutter metrics was relatively low because of disadvantages, such as lack of ability to accurately quantify the image with complex clutters and threshold selection problem. To address this problem, a novel multidirectionaldifference-Hash-based image clutter metric is proposed in this paper. Initially, an image similarity measure method based on multidirectional difference hash is established. Then, this method is applied to the quantification of image clutter, and an MDHash-based image clutter metric is obtained. Experimental results show that the proposed clutter metric correlates effectively with probability of detection, false alarm rate, and search time of observers.
Traditional lectures do hardly satisfy or benefit all of graduate students because of their diversity background, which might lead to inactive in the class, especially when the courses are professionally concentrated. Jigsaw cooperative learning (JCL), a student-centered learning method which engages all students and is beneficial in facilitating relationship-building, might be an alternative solution. A trial was made in an optical course “Optical Interferometric Measurement (OIM)”. However, the first trial we made is hardly counted as success. After thorough analysis on all factors that might cause the frustration, a new course plan is formed detailed. In this paper, we would like share our trials made in the course, both the unsuccessful one and the one work in progress.
Optical Manufacturing Technology is a compulsory course for postgraduate and advanced undergraduate majors in optical engineering and similar majors. Modern fabrication and testing equipment are extremely expensive, which cost millions or even tens of millions, and having high requirements for the use environment. Most schools that do not focus on optical fabrication rarely purchase these devices for educational purposes only. Relevant professional teachers are also not familiar with the working principle and use process of the new fabrication and testing equipment such as Computer Controlled Optical Surfacing (CCOS), Magnetorheological Finishing (MRF), Ion Beam Figuring (IBF), laser tracker and LuphoScan interferometer. This has caused the teaching content of the course to stay at the stage of singleaxis machining, knife-edge tester, so that students' knowledge of modern equipment is almost blank. While teaching with micro-course videos, the relevant equipment use process can be visually presented to students, and relevant technical experts can also explain the working principle and scope of application to students more professionally. According to the survey, 82% of students like this micro-course video teaching, acknowledging that it helps to better understanding modern optical fabrication and testing equipment, expand students' knowledge, and increase students' advantages in employment competition.
Infrared detecting and display device(IR-DDD)is a newly developed optical up-conversion device that integrates the light-emitting diode(LED)onto the infrared ( IR ) photo-detector, in order to convert IR light into the carriers photo-generated in detection materials and inject them into LED to emit visible light. This IR-DDD can achieve the direct up-conversion from IR ray to visible light, showing the considerable potential in night-vision application. This paper attempts a review of its working principle and current research progresses.
Pulse Coupled Neural Network (PCNN) is improved by Adaptive Lateral Inhibition (ALI), while a method of infrared (IR) dim small target segmentation based on ALI-PCNN model is proposed in this paper. Firstly, the feeding input signal is modulated by lateral inhibition network to suppress background. Then, the linking input is modulated by ALI, and linking weight matrix is generated adaptively by calculating ALI coefficient of each pixel. Finally, the binary image is generated through the nonlinear modulation and the pulse generator in PCNN. The experimental results show that the segmentation effect as well as the values of contrast across region and uniformity across region of the proposed method are better than the OTSU method, maximum entropy method, the methods based on conventional PCNN and visual attention, and the proposed method has excellent performance in extracting IR dim small target from complex background.
KEYWORDS: Signal processing, Analog electronics, Video, Digital signal processing, Nonuniformity corrections, Video processing, Infrared imaging, Imaging systems, Field programmable gate arrays, Infrared radiation
Compared with the infrared focal plane array (IRFPA) based on amorphous silicon (α-si), IRFPA based on vanadium oxide (VOx) has the advantages of big temperature coefficient of resistance, low noise and so on. In this paper, the design of the micro uncooled infrared imaging system based on VOx IRFPA is introduced. Firstly, the hardware design of the proposed system is discussed, which includes the system structure, VOx IRFPA module, driving module and the signal processing module based on FPGA. Secondly, the designs of the system configuration program as well as the consistency correction of the proposed system are discussed. Finally, some experiments were carried out to verify the validity of the models and the whole infrared imaging system, which indicated that our work will lay a foundation for the implement of micro and low-cost infrared imaging system.
In our previous works, we presented a zoom system and image stabilization design based on deformable mirrors (DMs). According to the high bandwidth and free edge characteristics of the piezoelectric deformable mirror (PDM), we tested the system’s image-stable capability. We found the PDM could realize some tilt displacements while keeping a certain stable surface shape, it could obtain higher image stabilizing precision when integrated with the traditional mechanical image stabilization systems. In the design of the image stabilization system, the PDM’s tilt displacement range is a key factor for consideration. So in this paper, we carried out a tilt displacement range testing experiment by using the OKO’s 37-channel PDM. We measured and analyzed the variation of the tilt displacements in optical image stabilization process, and calculated the maximum tilt angle as the PDM surface shape was stabilized. We built an experimental platform consisting of a fixed target, an imaging system based on PDM, and a CCD camera. We used the ZYGO interferometer as an evaluation instrument to measure the surface shape stability. When the PDM surface had a tilt displacement, the image point of the fixed target on the camera sensor shifted correspondingly. The tilt angle of the PDM could be obtained by calculating this shift. The results showed that the maximum tilt angle of the PDM was 0.2mrad. The paper also analyzed the experiment errors when concerning about the off-axis error of the PDM deflection center.
A novel space-variant image sensor based on optical method is proposed. Firstly, the mathematical models of the proposed image sensor and its non-uniform lens array are developed and verified. Secondly, the relationships among the parameters of the non-uniform lens have been simulated and discussed. Thirdly, experiments are carried out for verifying the characteristic of rotation and scaling invariance of the proposed image sensor. Finally, some conclusions are deduced, which will help to result in a space-variant image sensor with the characteristics of high sensitivity, high speed and big planar array, etc.
People attach more and more importance to bionic compound eye due to its advantages such as small volume, large field of view and sensitivity to high-speed moving objects. Small field of view and large volume are the disadvantages of traditional image sensor and in order to avoid these defects, this paper intends to build a set of compound eye system based on insect compound eye structure and visual processing mechanism. In the center of this system is the primary sensor which has high resolution ratio. The primary sensor is surrounded by the other six sensors which have low resolution ratio. Based on this system, this paper will study the target image fusion and extraction method by using plane compound eye structure. This paper designs a control module which can combine the distinguishing features of high resolution image with local features of low resolution image so as to conduct target detection, recognition and location. Compared with traditional ways, the way of high resolution in the center and low resolution around makes this system own the advantages of high resolution and large field of view and enables the system to detect the object quickly and recognize the object accurately.
The recharging of Active Medical Implant (AMI) is an important issue for its future application. In this paper, a method for recharging active medical implant using wearable incoherent light source has been proposed. Firstly, the models of the recharging method are developed. Secondly, the recharging processes of the proposed method have been simulated by using Monte Carlo (MC) method. Finally, some important conclusions have been reached. The results indicate that the proposed method will help to result in a convenient, safe and low-cost recharging method of AMI, which will promote the application of this kind of implantable device.
A novel retina-like image sensor based on space-variant lens array is proposed, in which a space-variant lens array is used for performing log-polar mapping. Firstly, the mathematical models are developed and verified. Secondly, the relationships among the parameters of the space-variant lens have been simulated and discussed. Finally, some conclusions are deduced, which will help to result in a retina-like image sensor with the characteristics of high speed, large resolution, high sensitivity and big planar array, etc.
KEYWORDS: Clouds, Image segmentation, 3D modeling, Visualization, 3D image processing, 3D visualizations, Detection and tracking algorithms, Data modeling, Reconstruction algorithms, Laser applications
Traditional visualization algorithms based on three-dimensional (3D) laser point cloud data consist of two steps: stripe point cloud data into different target objects and establish the 3D surface models of the target objects to realize visualization using interpolation point or surface fitting method. However, some disadvantages, such as low efficiency, loss of image details, exist in most of these algorithms. In order to cope with these problems, a 3D visualization algorithm based on space-slice is proposed in this paper, which includes two steps: data classification and image reconstruction. In the first step, edge detection method is used to check the parametric continuity and extract edges to classify data into different target regions preliminarily. In the second stage, the divided data is split further into space-slice according to coordinates. Based on space-slice of the point cloud data, one-dimensional interpolation methods is adopted to get the curves connected by each group of cloud point data smoother. In the end, these interpolation points obtained from each group are made by the use of getting the fitting surface. As expected, visual morphology of the objects is obtained. The simulation experiment results compared with real scenes show that the final visual images have explicit details and the overall visual result is natural.
KEYWORDS: Image sensors, LCDs, Field programmable gate arrays, Image processing, Sensors, Data conversion, Signal processing, Digital electronics, Optical data conversion, Fluctuations and noise
Retina-like image sensor is based on the non-uniformity of the human eyes and the log-polar coordinate theory. It has advantages of high-quality data compression and redundant information elimination. However, retina-like image sensors based on the CMOS craft have drawbacks such as high cost, low sensitivity and signal outputting efficiency and updating inconvenience. Therefore, this paper proposes a retina-like image sensor based on space-variant lens array, focusing on the circuit design to provide circuit support to the whole system. The circuit includes the following parts: (1) A photo-detector array with a lens array to convert optical signals to electrical signals; (2) a strobe circuit for time-gating of the pixels and parallel paths for high-speed transmission of the data; (3) a high-precision digital potentiometer for the I-V conversion, ratio normalization and sensitivity adjustment, a programmable gain amplifier for automatic generation control(AGC), and a A/D converter for the A/D conversion in every path; (4) the digital data is displayed on LCD and stored temporarily in DDR2 SDRAM; (5) a USB port to transfer the data to PC; (6) the whole system is controlled by FPGA. This circuit has advantages as lower cost, larger pixels, updating convenience and higher signal outputting efficiency. Experiments have proved that the grayscale output of every pixel basically matches the target and a non-uniform image of the target is ideally achieved in real time. The circuit can provide adequate technical support to retina-like image sensors based on space-variant lens array.
In order to improve the definition of the infrared image, and make it more accessible for human eyes, an infrared image enhancement algorithm based on Riemann-Liouville (R-L) fractional calculus and human visual properties is proposed in this paper. Combining the mathematical model of human retinal receptive field with R-L fractional calculus theory, an R-L fractional order Rodieck enhancement mask is designed. The mask is used to enhance the textures and edges of the image. Then, the grayscales of the enhancement result are adjusted according to the grayscale resolution capabilities of human eyes. It further improves the contrast of infrared images. Experimental results show that the proposed algorithm can effectively enhance the texture details and contrast of infrared images. Compared with histogram equalization method and multi-Retinex method, the enhancement result of the proposed algorithm has better visual effect, and it is more accessible for human eyes.
With the development of 3-D imaging techniques, three dimensional point cloud partition becomes one of the key
research fields. In this paper, two data partition algorithms are proposed. Each algorithm includes two parts: data
re-organization and data classification. Two methods for data re-organization are proposed: dimension reduction and
triangle mesh reconstruction. The algorithm of data classification is based on edge detection of depth data. The edge
detection algorithms of gray images are improved for depth data partition. As to the triangulation method, the data
partition is realized by region growing. The simulation result shows that the two methods can achieve point cloud data
partition of standard template and real scene. The result of standard template shows the total error rates of the two
algorithms are both less than 3%.
KEYWORDS: Data acquisition, Field programmable gate arrays, Data conversion, Signal processing, Data transmission, Laser development, Data processing, Radar signal processing, Telecommunications, Control systems
In this paper, a high-speed data acquisition system based on the technology of USB2.0 (Universal Serial Bus) is
designed, in which USB master logic is implemented in an FPGA (Field Programmable Gate Array). Firstly, the
hardware of data acquisition system is discussed, which includes chip selection, data acquisition and transmission circuit
and power conversion circuit. Secondly, the corresponding software including USB firmware program, USB device
driver and application program as well as its modifications have been described. The designed hardware and software
will help to achieve a data acquisition system with the characterstics of high speed and high accuracy, etc.
In the paper, a design scheme of driving circuit and collimating optical system used for 3D (three-dimensional) imaging device has been proposed. The driving circuit based on power MOSFET for high-power pulsed laser diode has the characteristics of short pulse-width and high output current. According to semiconductor laser’s far field divergence characteristic, the aspheric collimation part has been designed by using optical design software ZEMAX. Far field beam tracing and collimation results are simulated. The laser driver’s output current and pulse width are about 144A and 13.9ns respectively. The RMS of divergent angle of simulation in ZEMAX is 0.318mrad and the spot is more uniform.
This paper proposes a novel Micro-Opto-Electro-Mechanical-System (MOEMS) based beam steering method mainly
designed for far operating distance applications. The proposed system incorporates a MOEMS analog mirror developed
by Texas Instruments Inc. as core component and a beam expanding telescope. The analog mirror can change angle
within a ±5° range of mechanical rotation in two axes. The expanding telescope could reduce beam dispersion caused by
diffraction of the mirror's boundary. Expressions of steering equation, steering range and angular resolution are deduced.
Steering trace and far field steering beam patterns are simulated. Analyses on main performance of the system are
proposed.
KEYWORDS: Finite element methods, Waveguides, Electrodes, Signal attenuation, Telecommunications, Data modeling, Electromagnetism, Skin, 3D modeling, Head
The simulation based on the finite-element (FE) method plays an important role in the investigation of the intra-body
communication (IBC). In this paper, the method for modeling the whole human body based on the finite-element method
is proposed, while a finite-element model of the whole human body used for the simulations of the waveguide intra-body
communication has been developed. Finally, the simulations of the waveguide IBC with different signal transmission
paths have been achieved by using the developed finite-element model. Moreover, both the potential distributions and
the signal attenuations of the simulation results are discussed in detail, which indicate that the proposed method and
model offer the significant advantages in the theoretical analysis and the system design of the waveguide intra-body
communication.
Intra-body Body Communication (IBC) is a communication technology in which human body is used as a signal
transmission medium. Due to its unique characters, IBC technology is proposed as a novel and promising technology for
personal area network (PAN), computer network access, implant biomedical monitoring, human energy transmission,
etc. In this paper, investigation has been done in the computer simulation of the electrostatic coupling IBC by using the
developed finite-element models, in which (1) the incidence and reflection of electronic signal in the upper arm model
were analyzed by using the theory of electromagnetic wave, (2) the finite-element models of electrostatic coupling IBC
were developed by using the electromagnetic analysis package of ANSYS software, (3) the signal attenuation of
electrostatic coupling IBC were simulated under the conditions of different signal frequency, electrodes direction,
electrodes size and transmission distance. Finally, some important conclusions are deduced on the basis of simulation
results.
A novel multi-object detection and tracking technology based on hexagonal opto-electronic detector is proposed, in which
(1) a new hexagonal detector, which is composed of 6 linear CCDs, has been firstly developed to achieve the field of
view of 360 degree, (2) to achieve the detection and tracking of multi-object with high speed, the object recognition
criterions of Object Signal Width Criterion (OSWC) and Horizontal Scale Ratio Criterion (HSRC) are proposed. In this
paper, Simulated Experiments have been carried out to verify the validity of the proposed technology, which show that
the detection and tracking of multi-object can be achieved with high speed by using the proposed hexagonal detector and
the criterions of OSWC and HSRC, indicating that the technology offers significant advantages in Photo-electric
Detection, Computer Vision, Virtual Reality, Augment Reality, etc.
A system designed for measuring large-size parallelism and perpendicularity of large-size workpieces by laser alignment was developed. This paper mainly focuses on three questions. First, the building of long distance alignment optical beam based on laser diode, single mode optical fiber and phase plate diffraction technique. Second, the building of datum planes that parallel or perpendicular to each other by using pentagonal prism. Third, a new image processing algorithm called orthogonal projection method. The algorithm converts two-dimension image into two one-dimension images, which can speed up the data processing and restrain noise. The performance of this system was tested and verified in National CIMS Engineering Technology Research Center in China. The experimental results show that the relative measurement accuracy of the alignment system attains micron (μm) (0.3×10-6L) in 10 meters measurement range.
Real-time measurement of the variation of the concentration and temperature of the solution in liquid-phase by the use of contactless Mach-Zehnder interferometry was presented in this paper. The relationships between the temperature, concentration and refractive index of solution were investigated. The feasibility of the measurement of the temperature and concentration variation separately when they altered simultaneously was discussed. Consider the practical application, Koster prisms were chosen to built the Mach-Zehnder interferometer instead of common mirrors and beamsplitters. The use of Koester prisms made the optical system compact and increased the system stability. Utilizing two wavelength interference measurement we can work out the variation of temperature and concentration from the variation of the refractive index data using fitted second order polynomial was drew. The experiment of measuring the solution concentration and temperature by use of the undersaturation aqueous solution of KDP was done.
KEYWORDS: Eye, Image processing, Image analysis, Virtual reality, Detection and tracking algorithms, Cameras, Light sources, Inspection, Control systems design, Head
According to the research result in neural biology, human eyes can obtain high resolution only at the center of view of field. In the research of Virtual Reality helmet, we design to detect the gazing direction of human eyes in real time and feed it back to the control system to improve the resolution of the graph at the center of field of view. In the case of current display instruments, this method can both give attention to the view field of virtual scene and resolution, and improve the immersion of virtual system greatly. Therefore, detecting the gazing direction of human eyes rapidly and exactly is the basis of realizing the design scheme of this novel VR helmet. In this paper, the conventional method of gazing direction detection that based on Purklinje spot is introduced firstly. In order to overcome the disadvantage of the method based on Purklinje spot, this paper proposed a method based on image processing to realize the detection and determination of the gazing direction. The locations of pupils and shapes of eye sockets change with the gazing directions. With the aid of these changes, analyzing the images of eyes captured by the cameras, gazing direction of human eyes can be determined finally. In this paper, experiments have been done to validate the efficiency of this method by analyzing the images. The algorithm can carry out the detection of gazing direction base on normal eye image directly, and it eliminates the need of special hardware. Experiment results show that the method is easy to implement and have high precision.
A novel frame shift and integral method for the enhancement of moving image sequence acquired by MAVs is introduced. In this paper, we first describe the method of frame shift and integral, then the relationship between the shift value and flight velocity, flight height of MAVs is discussed, and the model to calculate the shift value of moving images is established. Moreover, the solution to remove the ladder stripe caused the opening section of shifted image is given. Finally, the imaging in the flight of MAVs is simulated. The experimental results show that 1) The calculated values of the shift value accord with the actual values, 2) With the increasing of the number of integrated frames, the SNR (signal noise ratio) of moving image is increased obviously.
The integration time limits the performance of CCD imaging-system, especially when it is used in MAVs (Micro Air Vehicles), so it is important to regulate the integration time of CCD in order to improve the quality ofmoving images when MAVs is flying. In this paper, we analyzed the relationship between image smear and the flying status of MAVs, and the degradation of the image quality caused by the variety of the integration time of CCD. We have proposed a method to ascertain the integration time of CCD imaging-system on the basis of analyzing the relationship between the integration time and the quality of moving images, and constructed a kind of experiment system to validate the method. The experiment result indicates that the method is valid for ascertaining the integration time of micro CCD imaging-system used in MAVs.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.