New background estimation and suppression algorithm via Zernike-facet model

Recent technical developments have significantly increased the signal-to-noise ratio (SNR) of arterial spin labeled (ASL) perfusion MRI. Despite this, typical ASL acquisitions still employ large voxel sizes. The purpose of this work was to implement and evaluate two ASL sequences optimized for whole-brain high-resolution perfusion imaging, combining pseudo-continuous ASL (pCASL), background suppression (BS) and 3D segmented readouts, with different in-plane k-space trajectories. Identical labeling and BS pulses were implemented for both sequences. Two segmented 3D readout schemes with different in-plane trajectories were compared: Cartesian (3D GRASE) and spiral (3D RARE Stack-Of-Spirals). High-resolution perfusion images (2 × 2 × 4 mm(3) ) were acquired in 15 young healthy volunteers with the two ASL sequences at 3 T. The quality of the perfusion maps was evaluated in terms of SNR and gray-to-white matter contrast. Point-spread-function simulations were carried out to assess the impact of readout differences on the effective resolution. The combination of pCASL, in-plane segmented 3D readouts and BS provided high-SNR high-resolution ASL perfusion images of the whole brain. Although both sequences produced excellent image quality, the 3D RARE Stack-Of-Spirals readout yielded higher temporal and spatial SNR than 3D GRASE (spatial SNR = 8.5 ± 2.8 and 3.7 ± 1.4; temporal SNR = 27.4 ± 12.5 and 15.6 ± 7.6, respectively) and decreased through-plane blurring due to its inherent oversampling of the central k-space region, its reduced effective TE and shorter total readout time, at the expense of a slight increase in the effective in-plane voxel size.

To present a constrained slice-dependent (CSD) background-suppression (BS) scheme in 2D arterial spin labeling (ASL) using simultaneous multislice acquisition with blipped-CAIPIRINHA (controlled aliasing in parallel imaging results in higher acceleration). Background suppression for 2D acquisition is challenging because of the multiple nulling points required for sequential slice readout. Constrained slice-dependent BS exploits the simultaneous multislice technique to reduce the readout duration, and uses slice-dependent premodulation pulses to achieve BS across slice groups. The proposed scheme was evaluated by in vivo brain experiments at 3 Tesla with multiband acceleration factors up to four. The utility of CSD BS was demonstrated through comparison against standard 2D acquisitions as well as 3D-BS pseudo-continuous ASL (pCASL). An average of 95% background signal reduction was achieved with CSD BS. As a result, the temporal signal-to-noise ratio (SNR) increased 48.2/39.9/36.9/36.0% and spatial SNR increased 132.5/80.0/63.5/54.2 in CSD-BS MB-1/2/3/4 scans, respectively. Whole-brain coverage was achievable with CSD-BS pCASL with MB-4, which yielded comparable spatial SNR as 3D BS pCASL. The proposed CSD-BS scheme for 2D-SMS pCASL offers a promising approach for effective suppression of background signals across a wide range of T1 to achieve whole-brain perfusion imaging. Magn Reson Med 79:394-400, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Optimization-based methods have gained prominence in infrared small target detection (ISTD) due to their effectiveness in handling sparse targets. However, existing approaches often struggle with spatio-temporal modeling, resulting in suboptimal background suppression and target misclassification. Moreover, stringent recovery tolerances limit computational efficiency in sequential infrared processing. To overcome these challenges, we introduce a novel approach that enhances both background estimation and target detection by leveraging spatio-temporal features. Specifically, we integrate a four-dimensional (4-D) spatio-temporal tensor with a fully connected tensor network (FCTN) completion strategy, utilizing the sum of nuclear norms to maximize intermode correlations while minimizing clutter. For target extraction, we employ a dual Gaussian-core spatial saliency filter (SSF) and a temporal difference filter (TDF), enabling precise discrimination between true targets and static interference. The resulting model—Spatial-Temporal Prior-Assisted Fully Connected Tensor Network (STPA-FCTN)—balances background suppression with target preservation. Furthermore, we systematically analyze recovery convergence in STPA-FCTN and identify an optimal tolerance of 10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-3</sup> using the proposed open-source TensorISTD benchmark. Extensive evaluations on public datasets confirm the superior numerical and visual performance of our method over state-of-the-art techniques. Code and benchmark resources are available at https://github.com/fengyiwu98/TensorISTD.

The ability to accurately predict natural backgrounds over a wide range of potential image acquisition conditions is key to wide-area change detection. The lack of robustness in pure image-based techniques has led to the development of site modeling and model-supported exploitation approaches. Instead of using multiple images to build a site model, we consider the possibility of using them directly to estimate the background of a new image for change detection. Our method uses a nonlinear mean-squared estimation technique to compute image backgrounds from multiple reference images. In comparing linear and nonlinear estimators, we find that although the performance of both improves as the number of reference images increases, the nonlinear estimator yields significantly better background estimates and is much less sensitive to registration errors between images than the linear estimator. We show that as the number of references increases, the pixels within a region become physically significant; i.e., they correspond to parts of the scene with similar physical properties. Synthetic aperture radar (SAR) background modeling and change detection results including a crosssensor change detection example in which SAR reference images are used to predict the background of an electro-optical (EO) image are presented. Significant performance gains over linear change detection (up to two orders of magnitude reduction in the false alarm rate) are demonstrated in each case over a 60-deg range of SAR sensor aspect angles.

Background statistics estimation is the key point for the statistics model based detectors. The background statistics obtained globally from the whole image may be inaccurate due to target contamination of the background information. To solve this problem, this paper proposed a local density based background estimation algorithm (LDBE) based on the definition of local density based anomaly score (LDAS). LDAS is a new metric that utilizes the distance between spectral to calculate each pixel's probability of background. LDBE uses LDAS as a criterion to determine whether a pixel is part of the background or not. By applying this algorithm, the background statistics can be estimated more accurately with the non-background pixels eliminated. The experimental results on real hyperspectral datasets suggest that the proposed background estimation algorithm can greatly improve the performance of statistical model based target detectors.

To address issues such as significant scale differences, complex pose variations, strong background interference, and similar category characteristics of pests in the images obtained from field traps, this study proposes a pest recognition method based on a two-stage “segmentation–detection” approach to improve the accuracy of field pest situation monitoring. In the first stage, an improved segmentation model, BAE-UNet (Background-Aware and Edge-Enhanced U-Net), is adopted. Based on the classic U-Net framework, a Background-Aware Contextual Module (BACM), a Spatial-Channel Refinement and Attention Module (SCRA), and a Multi-Scale Edge-Aware Spatial Attention Module (MESA) are introduced. These modules respectively optimize multi-scale feature extraction, background suppression, and boundary refinement, effectively removing complex background information and accurately extracting pest body regions. In the second stage, the segmented pest body images are input into the YOLOv8 model to achieve precise pest detection and classification. Experimental results show that BAE-UNet performs excellently in the segmentation task, achieving an mIoU of 0.930, a Dice coefficient of 0.951, and a Boundary F1 of 0.943, significantly outperforming both the baseline U-Net and mainstream models such as DeepLabV3+. After segmentation preprocessing, the detection performance of YOLOv8 is also significantly improved. The precision, recall, mAP50, and mAP50–95 increase from 0.748, 0.796, 0.818, and 0.525 to 0.958, 0.971, 0.977, and 0.882, respectively. The results verify that the proposed two-stage recognition method can effectively suppress background interference, enhance the stability and generalization ability of the model in complex natural scenes, and provide an efficient and feasible technical approach for intelligent pest trap image recognition and pest situation monitoring.

The monitoring of space debris is important for spacecraft such as satellites operating in orbit, but the background in star images taken by ground-based telescopes is relatively complex, including stray light caused by diffuse reflections from celestial bodies such as the Earth or Moon, interference from clouds in the atmosphere, etc. This has a serious impact on the monitoring of dim and small space debris targets. In order to solve the interference problem posed by a complex background, and improve the signal-to-noise ratio between the target and the background, in this paper, we propose a novel star image enhancement algorithm, MBS-Net, based on background suppression. Specifically, the network contains three parts, namely the background information estimation stage, multi-level U-Net cascade module, and recursive feature fusion stage. In addition, we propose a new multi-scale convolutional block, which can laterally fuse multi-scale perceptual field information, which has fewer parameters and fitting capability compared to ordinary convolution. For training, we combine simulation and real data, and use parameters obtained on the simulation data as pre-training parameters by way of parameter migration. Experiments show that the algorithm proposed in this paper achieves competitive performance in all evaluation metrics on multiple real ground-based datasets.

To compare single-shot turbo spin-echo (TSE) diffusion-weighted whole-body imaging with background suppression (DWIBS) and echo-planar imaging (EPI) DWIBS to determine the feasibility of direct-coronal TSE-DWIBS. All measurements were performed using a 3.0 T MRI scanner. In the phantom study, we compared the contrast ratios (CRs) of tumor-mimicking phantom (tumor) to muscle-mimicking phantom (muscle) and water to muscle and the signal-to-noise ratio (SNR) between TSE-DWIBS and EPI-DWIBS. In the volunteer study, 10 healthy volunteers were whole-body scanned with direct-coronal TSE-DWIBS, direct-coronal EPI-DWIBS (corEPI-DWIBS), and transverse EPI-DWIBS (traEPI-DWIBS). Two radiologists assessed the image distortion, uniformity of fat suppression, overall artifacts, and overall image quality in maximum intensity projection (MIP) images from each DWIBS image using a 5-point scale. In the phantom study, the CR of tumor to muscle was found to be lower for TSE-DWIBS (10.57 ± 0.45) than for EPI-DWIBS (15.38 ± 0.27), and the CR of water to muscle was higher for TSE-DWIBS (9.61 ± 0.66) than for EPI-DWIBS (2.52 ± 0.60). The volunteer study revealed good inter observer agreement between TSE-DWIBS and EPI-DWIBS with respect to image distortion, uniformity of fat suppression, overall artifacts, and overall image quality, with weighted Cohen's κ coefficients of 0.91, 0.74, 0.87, and 0.72, respectively. Qualitative analysis scores for image distortion, uniformity of fat suppression, overall artifacts, and overall image quality were significantly higher for TSE-DWIBS than for corEPI-DWIBS or traEPI-DWIBS (p < 0.05). Direct-coronal TSE-DWIBS is robust against magnetic field inhomogeneity. High-quality images without distortion or fat suppression inhomogeneity were obtained. Many studies on DWIBS have been previously reported; however, these studies used EPI read-out. To the best of our knowledge, no studies using TSE read-out have been reported. In this study, we examined the feasibility of TSE-DWIBS with lesser artifacts than EPI-DWIBS.

The images collected by CCD devices have the characteristics of low SNR (Signal-to-Noise Ratio) and complex background. In order to reduce the difficulty of target extraction, this paper uses real star images to train a background suppression network based on CBDNet [1], a denoising network structure. The experimental result shows that the network can effectively suppress the background and improve the SNR of the star points while retaining the detailed information of the star points.

Camera-based respiratory triggering improves the image quality of 3D magnetic resonance cholangiopancreatography

Equivalent circuit model (ECM) based approaches are widely used by Li-ion battery management systems. In the ECM approach, the voltage drop within a battery is modelled using resistive and capacitive electrical components. Numerous approaches were reported to estimate the parameters of ECM based on the voltage and current measurements from the battery. Battery ECM parameter estimation is done based on both voltage and current measurements. The measurement noise in these applications results in a doubly noisy observation model and makes theoretical analysis challenging. In this paper, theoretical performance bounds are derived for ECM parameter estimation in batteries where the effect of both voltage and current measurement noises are objectively analyzed. The performance bound is derived in the form of Cramer-Rao lower bound (CRLB) by considering the doubly noisy observation model that represent ECM parameter estimation in batteries. It is shown that traditional least square estimation approach becomes biased and inefficient at low signal to noise ratio (SNR) levels. A new approach, based on the total least squares (TLS) method, is developed for low SNR conditions. It is shown through simulation experiments that the TLS approach can be efficient for ECM parameter estimation. The proposed approach is validated through data collected from cylindrical Li-ion battery cells.

Background initialization is the task of computing a background model by processing a set of preliminary frames in a video scene. The initial background estimation serves as bootstrap model for video segmentation of foreground objects, although the background estimation could be refined and updated in steady state operation of video processing systems. In this paper we approach the background modeling problem with a weightless neural network called WiSARD $$^{rp}$$ . The proposed approach is straightforward, since the computation is pixel–based and it exploits a dedicated neural network to model the pixel background by using the same training rule.

Segmenting objects from cluttered backgrounds in single-channel images, such as marine radar echoes, medical images, and remote sensing images, poses significant challenges due to limited texture, color information, and diverse target types. This paper proposes a novel solution: the Onet, an O-shaped assembly of twin U-Net deep neural networks, designed for unsupervised binary semantic segmentation. The Onet, trained with an intensity-complementary image pair and without the need for annotated labels, maximizes the Jensen-Shannon divergence (JSD) between the densely localized features and the class probability maps. By leveraging the symmetry of U-Net, Onet subtly strengthens the dependence between dense local features, global features, and class probability maps during the training process. The design of the complementary input pair aligns with the theoretical requirement that optimizing JSD needs the class probability of negative samples to accurately estimate the marginal distribution. Compared to the current leading unsupervised segmentation methods, the Onet demonstrates superior performance in target segmentation in marine radar frames and cloud segmentation in remote sensing images. Notably, we found that Onet's foreground prediction significantly enhances the signal-to-noise ratio (SNR) of targets amidst marine radar clutter. Onet's source code is publicly accessible at https://github.com/joeyee/Onet.

Shore-based polarized infrared images of ship target and background were obtained in the MAPTIP measurement series, using vertical and horizontal polarization filters. These images have been analyzed to yield degree of polarization and polarization contrast enhancement for targets with sea and sky background Strong vertical polarization is observed in the sea surface emission near the Brewster Angle in the 8 - 12 micrometer (LWIR) band, whereas horizontal polarization due to reflection is usually dominant in the 3 - 5 micrometer (MWIR) band. Ship targets at broadside aspect show a degree of polarization less than 5%. An enhancement of contrast by up to 30% in the LWIR (less in the MWIR) is obtained by polarization suppression of sea background. The contrast enhancement in the MWIR is counteracted by the horizontal polarization of reflected sunlight.

Detection of moving objects is a fundamental task in video based surveillance and security applications. Many detection systems use background estimation methods to model the observed environment. In outdoor surveillance, moving backgrounds (waving trees, clutter) and illumination changes (weather changes, reflections, etc.) are the major challenges for background modelling and the development of a single model that fulfils all these requirements is usually not possible. In this paper we present a background estimation technique for motion detection in non-static backgrounds that overcomes this problem. We introduce an enhanced background estimation architecture with a long-term model and a short-term model. Our system showed that fusion of the detections of these two complementary approaches, improves the quality and reliability of the detection results.