Multichannel imaging in remote sensing

The detailed, high-resolution images captured by drones pose challenges to target detection algorithms with complex scenes and small-sized targets. Moreover, targets in unmanned aerial vehicle images are usually affected by factors such as viewing perspective, occlusion, and light, which increase the difficulty of target detection. In response to the above issues, we propose an improved SDMSEAF-YOLOv8 for target detection based on YOLOv8, combined with a Bi-directional Feature Pyramid Network, to improve the sensing ability of the model for multiscale targets. A Space-to-depth layer replaces the traditional strided convolution layer to enhance the extraction of fine-grained information and small-sized target features. A Multi-Separated and Enhancement Attention module enhances the feature learning ability of the occluded target region, thus reducing missed and false detections. Four detection heads are employed for tiny target detection, each responsible for different size ranges, so as to improve the accuracy and robustness of small target detection. The conventional non-maximum suppression algorithm is improved, so as to reduce the problem of missed detections under a densely occluded scene by setting the attenuation function to adjust the confidence of the treated box based on the overlap between it and the highest-scoring box. Experiments demonstrate that the accuracy of SDMSEAF-YOLOv8 exceeds that of state-of-the-art models on the VisDrone2019-DET-val dataset, with a mAP of 42.9% at 640-pixel resolution, 14.8% over the baseline YOLOv8-x algorithm model, and 6.0% over the known state-of-the-art Fine-Grained Target Focusing Network model and with twice as fast detection.

Multispectral, hyperspectral, polarimetric, and other types of multichannel imaging spectrometers are coming into common use for a variety of applications, including remote sensing, material identification, forensics, and medical diagnosis. These instruments are often bulky and intolerant of field abuse, so designing compact, reliable, portable, and robust devices is a priority. In contrast to most engineering designs, animals have been building compact and robust multichannel imaging systems for millennia—their eyes. Biological sensors arise by evolution, of course, and are not designed ‘for’ a particular use; they exist because the creatures that were blessed with useful mutations were better able to survive and reproduce than their competitors. While this is an inefficient process for perfecting a sensor, it brings unexpected innovations and novel concepts into visual system design—concepts that may be useful in the inspiration of new engineered solutions to problematic challenges, like the ones mentioned above. Here, we review a diversity of multichannel visual systems from both vertebrate and invertebrate animals, considering the receptor molecules and cells, spectral sensitivity and its tuning, and some aspects of the higher-level processing systems used to shape spectral (and polarizational) channels in vision. The eyes of mantis shrimps are presented as potential models for biomimetic multichannel imaging systems. We end with a description of a bioinspired, newly developed multichannel spectral/polarimetric imaging system based on mantis shrimp vision that is highly adaptable to field application.

To more comprehensively understand the target detection technology based on laser imaging and therefore learn the realization means, constitutional features and action effects of this technology on the whole, it is necessary to analyze the detection system before studying each technique. The target detection technology based on a laser imaging system is the equipments using for target detection using a laser imaging system integrating many techniques. They include laser receiving and transmitting, dynamic measurement of carrier attitude, dynamic differential global positioning system (GPS) with high precision, data processing, target detection and location, target classification and identification, and photoelectric transceiver and calibration. This chapter mainly introduces the components, functions, and working principles of the target detection using laser imaging systems.

Imaging spectroscopic technique has been used for the mineral and rock geological mapping and alteration information extraction successfully with many reasonable results, but it is mainly used in arid and semi-arid land with low vegetation covering. In the case of the high vegetation covering, the outcrop of the altered rocks is small and distributes sparsely, the altered rocks is difficult to be identified directly. The target detection technique using imaging spectroscopic data should be introduced to the extraction of small geological targets under high vegetation covering area. In the paper, we take Ding-Ma gold deposit as the study area which located in Zhenan country, Shanxi province, the spectral features of the targets and the backgrounds are studied and analyzed using the field reflectance spectra, in addition to the study of the principle of the algorithms, some target detection algorithms which is appropriate to the small geological target detection are introduced. At last, the small altered rock targets under the covering of vegetation in forest are detected and discriminated in imaging spectroscopy data with the methods of spectral angle mapper (SAM), Constrained Energy Minimization (CEM) and Adaptive Cosine Estimator (ACE). The detection results are reasonable and indicate the ability of target detection algorithms in geological target detection in the forest area.

Multichannel passive millimeter-wave (PMMW) imaging systems have advantage in operating speed, however traditional data processing methods for single channel systems are not applicable for such kind of system. This paper describes an 8-mm multichannel PMMW imaging system, points out the specific problems encountered in multichannel systems, and proposes an experimental multichannel calibration approach in microwave unreflected chamber. In the last superresolution algorithms considering imaging process and some related issues are discussed. The imaging result shows the proposed processing approaches are effective.

Present day airborne radars have to perform multiple tasks like detection, tracking, and identification of targets in the presence of advanced EW countermeasures like RF jamming, and must be able to separate true targets from noise and clutter conditions caused by complex physical conditions, such as mountainous terrain, and environmental conditions. Although conventional signal processing methods for target detection have achieved a great level of maturity, they still continue to perform poorly under low SNR conditions and model imperfections due to statistical variations. Meanwhile, performing target detection using neural network-based methods has also been gathering increased attention. We show that various Convolutional Neural Network(CNN) and Recurrent Neural Network(RNN) based techniques improve target detection at low SNR conditions. The CNN-based architecture proposed in this paper performs single shot radar target detection on RD map, and localization of the target. Various enhancements to the technique is also proposed. The proposed method shows improvement over conventional and state-of-the-art deep learning based methods. Furthermore, we evaluate the proposed method on real airborne data and prove its effectiveness.

The strong target contamination has a great influence on the performance of the conventional clutter suppression method. In this paper, a robust approach to clutter suppression and ground moving target detection has been proposed for multi-channel SAR system. Firstly, the modulus normalized vector corresponding to the pixel data vector is constructed and its covariance matrix has been formed. Secondly, the ground clutter is suppressed using the new adaptive clutter suppression weight. Lastly, target detection and localization are carried out. The effectiveness and robustness of the proposed approach are confirmed by performance analysis and real measured data processing.

We have recently conducted a series of laboratory and field test to demonstrate the utility of combining active illumination with hyperspectral imaging for the detection of concealed targets in natural terrain. The active illuminator, developed at MIT Lincoln Laboratory, is a novel microlaser-pumped fiber Raman source that provides high- brightness, subnanosecond-pulse-length output spanning the visible through near-IR spectral range. The hyperspectral- imaging system is comprised of a compact, grating-based spectrometer that uses a gateable, intensified CCD array as the detector element. The illuminator and hyperspectral imaging system are mounted on a small platform that is itself mounted on a tripod and scanned in azimuth to build an image scene of up to several hundred spectral bands. The system has been deployed under a variety of environmental conditions, including night-time illumination, and on a variety of target scenes, including exposed and concealed plastic and metallic mine-like targets. Targets have been detected and identified on the basis of spectral reflectance, fluorescence signatures, degree of polarization, and range-to-target information. The combination of laser-like broadband illumination and hyperspectral imaging offers great promise in concealed or obscured target detection. On-going developments include the incorporation of broadband illuminators in the 1 to 2 micrometers and 3 to 5 micrometers spectral bands, with corresponding increases in spectral coverage of the imaging and detection systems.

We present a statistically-based method for the enhancement and detection of mines and mine-like targets, in multi-channel imagery. Standard approaches to such multi-channel image processing take advantage of the correlation across channels within a pixel, but typically do not exploit the spatial dependency between pixels. This work aims to construct appropriate spatial statistical models for multi-channel mine imagery and apply these models to allow both image enhancement as well as direct and improved detection of anomalies (i.e., targets) in such data. We base the method on a Markov Random Field (MRF) model that incorporates a priori information about both the target's and the background's spatial characteristics. In particular, we find a Maximum A Posterior (MAP) detector of mine targets in background under the prior assumption target pixels are locally spatially dependent. We implement our algorithm on polarimetric and thermal data obtained from the Remote Minefield Detection System (REMIDS), with favorable results compared to a Maximum Likelihood (ML) detector that performs detections on a pixel-by-pixel basis, i.e. without spatial correlation.

: The goal of this project was to investigate the design, fabrication, and characterization, of multispectral quantum dots-in-a-double well (DDWELL) infrared detectors and investigate the integration of a Surface Plasmon (SP) assisted resonant cavity with the infrared detectors. Multicolor detector capabilities are highly desirable for advanced IR imaging systems, since they provide enhanced target discrimination and identification, combined with lower false-alarm rates. Systems that collect data in separate IR spectral bands can discriminate both absolute temperature as well as unique signatures of objects in the scene. By providing this new dimension of contrast, multiband detection also offers advanced color processing algorithms to further improve sensitivity above that of single-color devices. This is extremely important for identifying temperature differences between missile targets, warheads, and decoys. Multispectral IR focal plane arrays (FPAs) are highly beneficial for a variety of applications such as missile warnings guidance, precision strike, airborne surveillance, target detection, and recognition. Military surveillance, target detection, and target tracking can be undertaken using single-color FPAs if the targets are easy to identify. However, in the presence of clutter, or when the target and/or background are uncertain, or in situations where the target and/or background may change during engagement, single-color system design involves compromises that can degrade overall capability. It is well established that in order to reduce clutter and enhance the desired features contrast, one will require the use of multispectral focal plane arrays. In such cases, multicolor imaging can greatly improve overall system performance. The focus of this one year effort was to undertake a trade study to evaluate these technologies and a rigorous investigation of the performance parameters of these technologies.

Given recent advancements of modern hyperspectral (HS) sensors, the potential for information extraction has increased drastically given the continual improvements in spatial and spectral resolution. As a result, more sophisticated feature extraction and target detection (TD) algorithms are needed to improve the performance of the image analyst, whether computer-based or human. In this paper, a novel TD algorithm based on Projection Pursuit (PP) is proposed and implemented. PP is a well-known technique for dimensionality reduction in multi-band data sets without loss of any critical information. This technique highlights different features of interest in an image, thus improving and simplifying subsequent anomaly detection. The new target detection technique is based on a hybrid of PP and Reed_Xiaoli (RX) anomaly detector. In this study, the combining of PP with the RX detector (PPRX) adds some extra value to the standard RX detection technique and leads the development of a TD method that can be applied on hyperspectral/multispectral (MS) data sets. This novel technique, after being trained by using the Projection Index (PI) and a priori information of target of interest, utilizes RX detector to evaluate each potential projection. The main drawback of previously introduced PP methods such as those based on Information Divergence and Kurtosis/Skewness is that these techniques are sensitive to statistical outliers and cannot be used to highlight a specific target of interest. This study uses three data sets: (1) 4-band IKONOS multispectral data (2) 210-band HYDICE, and (3) 200-band simulated hyperspectral data set.

Camera arrays can enhance the signal-to-noise ratio (SNR) between dim targets and backgrounds through multi-view synthesis. This is crucial for the detection of dim targets. To this end, we design and develop an infrared camera array system with a large baseline. The multi-view synthesis of camera arrays relies heavily on the calibration accuracy of relative poses in the sub-cameras. However, the sub-cameras within a camera array lack strict geometric constraints. Therefore, most current calibration methods still consider the camera array as multiple pinhole cameras for calibration. Moreover, when detecting distant targets, the camera array usually needs to adjust the focal length to maintain a larger depth of field (DoF), so that the distant targets are located on the camera’s focal plane. This means that the calibration scene should be selected within this DoF range to obtain clear images. Nevertheless, the small parallax between the distant sub-aperture views limits the calibration. To address these issues, we propose a calibration model for camera arrays in distant scenes. In this model, we first extend the parallax by employing dual-array frames (i.e., recording a scene at two spatial locations). Secondly, we investigate the linear constraints between the dual-array frames, to maintain the minimum degrees of freedom of the model. We develop a real-world light field dataset called NUDT-Dual-Array using an infrared camera array to evaluate our method. Experimental results on our self-developed datasets demonstrate the effectiveness of our method. Using the calibrated model, we improve the SNR of distant dim targets, which ultimately enhances the detection and perception of dim targets.

In recent years, target detection in maritime surveillance has become a popular subject. Horizon under maritime environment is one of the most important semantic boundaries for segmenting the image into sea and sky. With the help of horizon, seeking scope can be restricted to a definite small area, which significantly reduces the computational complexity. Meanwhile, horizon is used to automatically adjust the attitude of air vehicles as well. Although a plenty of algorithms for horizon detection have been created, they are prone to low visibility and occlusions. In this paper, a novel robust algorithm based on probability distribution and physical characteristics is proposed. At first, the probabilities are distributed to each vertically divided region by weighted textures, and the sea-sky region is located from definite probabilistic intervals. Based on a Canny edge detector, Hough transform extracts a series of candidate horizons. Finally, a novel voting method is applied. The line with the maximum value will be regarded as the true horizon. The proposed algorithm can not only precisely detect the horizon from fine images but also from blurred image, even with splashed camera. The performance of the proposed algorithm is verified by experiments on 65 videos containing more than 20 000 frames under different sceneries. The results illustrate that our algorithm can detect the horizon line with minor errors in comparison with other five state-of-the-art algorithms.

We report on numerical and experimental characterization of the performance of a fiber link optimized for the delivery of sub-100-fs laser pulses at 1550 nm over several meters of fiber. We investigate the power handling capacity of the link, and demonstrate all-fiber delivery of 1-nJ pulses over a distance of 5.3 m. The fiber link consists of dispersion-compensating fiber (DCF) and standard single-mode fiber. The optical pulses at different positions in the fiber link are measured using frequency-resolved optical gating (FROG). The results are compared with numerical simulations of the pulse propagation based on the generalized nonlinear Schrödinger equation. The high input power capacity of the fiber link allows the splitting and distribution of femtosecond pulses to an array of fibers with applications in multi-channel fiber-coupled terahertz time-domain spectroscopy and imaging systems. We demonstrate THz pulse generation and detection using a distributed fiber link with 32 channels and 2.6 nJ input pulse energy.

Wide field-of-view (FOV) and high-resolution (HR) imaging systems have become indispensable information acquisition equipment in many applications, such as video surveillance, target detection and remotely sensed imagery. However, due to the constraints of spatial sampling and detector processing level, the ability of remote sensing to obtain high spatial resolution is limited, especially in the wide FOV imaging. To solve these problems, we propose a multi-scale feature extraction (MSFE) network to realize super-resolution imaging in a low-light- level (LLL) environment. In order to perform data fusion and information extraction for low resolution (LR) images, the network extracts high-frequency detail information from different dimensions by combining the channel attention mechanism module and skip connection module. In this way, redundant low-frequency signals can pass through the network tail-ends, furthermore, the more important high-frequency components calculation can be focused. The qualitative and quantitative analysis results show that the proposed method achieves the most advanced performance compared with other state-of-the-art methods, which shows the superiority of the design framework and the effectiveness of presenting modules.