Adaptive Scale Patch-Based Contrast Measure for Dim and Small Infrared Target Detection

Dim and small infrared target fast detection guided by visual saliency

Real-time small infrared (IR) target detection is critical to the performance of the situational awareness system in high-altitude aircraft. However, current IR target detection systems are generally hardware-unfriendly and have difficulty in achieving a robust performance in datasets with clouds occupying a large proportion of the image background. In this paper, we present new results by using an efficient method that extracts the candidate targets in the pre-processing stage and fuses the local scale, blob-based contrast map and gradient map in the detection stage. We also developed mid-wave infrared (MWIR) and long-wave infrared (LWIR) cameras for data collection experiments and algorithm evaluations. Experimental results using both publicly available datasets and image sequences acquired by our cameras clearly demonstrated that the proposed method achieves high detection accuracy with the mean AUC being at least 22.3% higher than comparable methods, and the computational cost beating the other methods by a large margin.

Designing and learning a lightweight network for infrared small target detection via dilated pyramid and semantic distillation

Infrared (IR) small-target detection plays a vital role in various applications of both civil and military areas, such as marine rescue, forest fire prevention, and precise guidance. However, challenges stemming from the small size of IR targets and noise interference in complex backgrounds often limit the accuracy of IR small-target detection algorithms. Owing to the rapid development of deep learning, numerous convolutional neural networks (CNNs)-based methods have emerged in recent years, but they are inevitable to encounter the risk of target loss in deep layers due to the use of pooling layers. To address this problem, we present a wavelet transform-based augmented perception network, namely WTAPNet, in this article. First, an enhancement and enlarge (EE) module is designed to improve the network’s perceptual capability for IR small targets by magnifying image resolution and augmenting target features at the same time. Then, a discrete wavelet transform-based downsampling (DWTD) module and an inverse wavelet transform-based fusion (IWTF) module are proposed. These two modules collaboratively work to extract and fuse multiscale features, which simultaneously reduces information loss. Finally, a bottom-up path fusion strategy is exploited to highlight and preserve small-target features, which associate the lowest level with the highest level features and interconnect predictions from different hierarchical levels. Experimental results on the NUDT-SIRST dataset and the SIRST dataset demonstrate the superiority of our WTAPNet over other state-of-the-art IR small-target detection methods in terms of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$F1$ </tex-math></inline-formula>-measure, recall, and other indicators. Our codes are publicly available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/MinjieWan/WTAPNet</uri>.

Small infrared target detection using absolute average difference weighted by cumulative directional derivatives

Specific to the problem of infrared small-target detection in complex backgrounds, a multimechanism deep feature enhancement network model (MDFENet) was proposed. A lightweight multimechanism attention collaborative fusion module was proposed to efficiently fuse low-level features and high-level features to solve the problem that small infrared targets are easy to annihilate in the deep layer of the network. Based on the analysis of the background and target data, a normalized loss function was proposed, which integrates the segmentation threshold selection into the network and normalizes the probability of the network output to simulate a step function and reflect relative differences. Aiming at the sparseness of infrared target features, we used the subpixel convolution method to upsample the features to obtain high-resolution feature images while expanding the size of the feature map. We conducted detailed comparison and ablation experiments, comparing MDFENet with ALCNet, APGCNet, and other state-of-the-art networks to verify the effectiveness and efficiency of the network. Results show that the MDFENet algorithm achieves the optimal result in the balance of detection efficiency and lightweightedness on two datasets.

Entropy-based window selection for detecting dim and small infrared targets

The detection of small infrared targets lacking texture and shape information in the presence of complex background clutter is a challenge that has attracted considerable research attention in recent years. Typical deep learning-based target detection methods are designed with deeper network structures, which may lose targets in the deeper layers and cannot directly be used for small infrared target detection. Therefore, we designed the attention fusion feature pyramid network (AFFPN) specifically for small infrared target detection. Specifically, it consists of feature extraction and feature fusion modules. In the feature extraction stage, the global contextual prior information of small targets is first considered in the deep layer of the network using the atrous spatial pyramid pooling module. Subsequently, the spatial location and semantic information features of small infrared targets in the shallow and deep layers are adaptively enhanced by the designed attention fusion module to improve the feature representation capability of the network for targets. Finally, high-performance detection is achieved through the multilayer feature fusion mechanism. Moreover, we performed a comprehensive ablation study to evaluate the effectiveness of each component. The results demonstrate that the proposed method performs better than state-of-the-art methods on a publicly available dataset. Furthermore, AFFPN was deployed on an NVIDIA Jetson AGX Xavier development board and achieved real-time target detection, further advancing practical research and applications in the field of unmanned aerial vehicle infrared search and tracking.

To address the issues of low model accuracy caused by complex ground environments and uneven target scales and high computational complexity in unmanned aerial vehicle (UAV) aerial infrared image target detection, this study proposes a lightweight UAV aerial infrared small target detection algorithm called PHSI-RTDETR. Initially, an improved backbone feature extraction network is designed using the lightweight RPConv-Block module proposed in this paper, which effectively captures small target features, significantly reducing the model complexity and computational burden while improving accuracy. Subsequently, the HiLo attention mechanism is combined with an intra-scale feature interaction module to form an AIFI-HiLo module, which is integrated into a hybrid encoder to enhance the focus of the model on dense targets, reducing the rates of missed and false detections. Moreover, the slimneck-SSFF architecture is introduced as the cross-scale feature fusion architecture of the model, utilizing GSConv and VoVGSCSP modules to enhance adaptability to infrared targets of various scales, producing more semantic information while reducing network computations. Finally, the original GIoU loss is replaced with the Inner-GIoU loss, which uses a scaling factor to control auxiliary bounding boxes to speed up convergence and improve detection accuracy for small targets. The experimental results show that, compared to RT-DETR, PHSI-RTDETR reduces model parameters by 30.55% and floating-point operations by 17.10%. Moreover, detection precision and speed are increased by 3.81% and 13.39%, respectively, and mAP50, impressively, reaches 82.58%, demonstrating the great potential of this model for drone infrared small target detection.

It is always a difficult challenge to detect small infrared targets under complex background and ensure high detection rate and low false alarm rate. In this paper, a detection method based on morphological Top-hat filtering and local energy difference is proposed. This method utilizes the advantage that morphological Top-hat filter can suppress the backgTround, and then segmented the image based on local energy difference to identify small targets. In this paper, the specific detection process of this method is given, and a number of small infrared target grayscale images under complex background are selected for the simulation experiment. The experimental results show that this method has a high detection rate and a low false alarm rate, and is of certain reference value for the detection of small infrared targets under complex background.

IRST (Infrared Search and Track) has been applied to many military or civil fields such as precise guidance, aerospace, early warning. As a key technique, small target detection based on infrared image plays an important role. However, infrared targets have their own characteristics, such as target size variation, which make the detection work quite difficult. In practical application, the target size may vary due to many reasons, such as optic angle of sensors, imaging distance, environment and so on. For conventional detection methods, it is difficult to detect such size-varying targets, especially when the backgrounds have strong clutters. This paper presents a novel method to detect size-varying infrared targets in a cluttered background. It is easy to find that the target region is salient in infrared images. It means that target region have a signature of discontinuity with its neighboring regions and concentrates in a relatively small region, which can be considered as a homogeneous compact region, and the background is consistent with its neighboring regions. Motivated by the saliency feature and gradient feature, we introduce minimum target intensity (MTI) to measure the dissimilarity between different scales, and use mean gradient to restrict the target scale in a reasonable range. They are integrated to be multiscale MTI filter. The proposed detection method is designed based on multiscale MTI filter. Firstly, salient region is got by morphological low-pass filtering, where the potential target exists in. Secondly, the candidate target regions are extracted by multiscale minimum target intensity filter, which can effectively give the optimal target size. At last, signal-to-clutter ratio (SCR) is used to segment targets, which is computed based on optimal scale of candidate targets. The experimental results indicate that the proposed method can achieve both higher detection precision and robustness in complex background.

The infrared (IR) small target detection algorithm with a high detection rate, low false alarm rate, and high real-time performance has significant application value in the field of IR remote sensing. IR small targets in complex backgrounds have low contrast and low signal-to-noise ratio (SNR). Therefore, small target detection is more difficult. Traditional IR small target detection is generally implemented by local contrast methods (LCM), nonlocal autocorrelation methods (NAM), and adaptive segmentation. In this letter, a robust infrared small target detection network (RISTDnet) is proposed based on deep learning. In RISTDnet, a feature extraction framework combining handcrafted feature methods and convolutional neural networks is constructed, a mapping network between feature maps and the likelihood of small targets in the image is established, and a threshold is applied on the likelihood map to segment real targets. Experimental results show that the RISTDnet can detect small targets with different sizes and low SNRs in complex backgrounds and have better effectiveness and robustness against existing algorithms.

Infrared small target detection is one of the most important parts of infrared search and tracking (IRST) system. Generally, the small and dim target is of low signal-to-noise ratio and buried in the complicated background and heavy noise, which makes it extremely difficult to be detected with low false alarm rates. To solve this problem, we propose a small target detection method based on multiscale local contrast measure. Different from conventional methods, we novelly measure the local contrast from two aspects: local dissimilarity and local brightness difference. First, we present a new dissimilarity measure called the local energy factor (LEF) to describe the dissimilarity between the small targets and their surrounding backgrounds. Second, the feature of the brightness difference between the small targets and the backgrounds is utilized. Afterward, the local contrast is measured by taking both features of the above into account. Finally, an adaptive segmentation method is applied to extract the small targets from the backgrounds. Extensive experiments on real test data set demonstrate that our approach outperforms the state-of-the-art approaches.

Detecting infrared small targets quickly and accurately in complex backgrounds has always been a challenging task. Data-driven methods have achieved good results because of their powerful feature extraction capabilities. Many algorithms use ResNet or VGG as their backbone, but because of the small size and inconspicuous features, pooling layers in their networks could lead to the loss of targets in deep layers. Even though dense network structure is proposed to alleviate this issue, its excessive dense connections makes it difficult to achieve real-time detection. To meet the requirements of both accurate performance and real-time detection, we propose MDAFNet. We design a Moderately Dense Adaptive Feature Fusion (MDAF) module that contains only three feature layers as the backbone of the network. This module connects all internal features with each other and utilizes a weighted sum of different layers as the output, promoting feature reuse and maintaining infrared small target features in the deep layers of the network. We also design a Coarse-to-Fine detection head and introduce auxiliary loss to enable the network to predict target contours with greater precision. Moreover, we propose a new data augmentation method that effectively enhances the generalization performance of network. Experimental results demonstrate that our network achieves excellent performance in detection accuracy and meets the requirements for real-time detection on RTX3080 gpu.

A pixel-level local contrast measure for infrared small target detection