9 January 2023 in IEEE Transactions on Geoscience and Remote Sensing
Infrared small target detection plays an important role in military and civilian fields while it is difficult to be solved by deep learning technologies due to scarcity of data and strong inter-class imbalance. To relieve scarcity of data, we build a massive dataset IRDST, which contains 142,727 frames. Also, we propose a Receptive-Field and Direction Induced Attention Network (RDIAN), which is designed by utilizing characteristics of target size and gray-scale to solve inter-class imbalance between targets and background. Using convolutional layers with different receptive-fields in feature extraction, target features in different local regions are captured, which enhances diversity of target features. By using multi-direction guided attention mechanism, targets are enhanced in low-level feature maps. Experimental results with comparison methods and ablation study demonstrate effective detection performance of our model. Dataset and code will be available at http://xzbai.buaa.edu.cn/datasets.html.
1 February 2022 in IEEE Transactions on Fuzzy Systems
Infrared ship segmentation is extensively applied in military fields. Due to noise and intensity inhomogeneity, segmentation of infrared ship is a challenging task. The fuzzy c-means (FCM) clustering algorithm is widely used in image segmentation. However, traditional FCM is sensitive to noise and unable to obtain desirable segmentation results for infrared ship images. In this paper, a novel probability induced intuitionistic fuzzy c-means clustering algorithm is proposed to address the problem. First, the target probability information is incorporated into intuitionistic FCM to induce and refine membership which is affected by interferences. Second, by making use of neighborhood information in the form of regularization term, the proposed method could suppress intensity inhomogeneity as well as maintain image details. Experimental results demonstrate that the proposed method could achieve better results than other 12 comparing algorithms for infrared ship segmentation.
11 August 2022 in IEEE Transactions on Intelligent Transportation Systems
Semantic segmentation in urban scenes is widely used in applications of intelligent transportation systems (ITS). In urban scenes, thermal infrared (TIR) images can be captured in weak illumination conditions or in the presence of obscuration (e.g., light fog, smoke). Therefore, TIR images have great potential to endow automated intelligent vehicles or assist navigation systems. However, TIR imaging is blurry and low-contrast due to the absorption by atmospheric gases and heat transfer effect. Hence, TIR semantic segmentation in urban scenes has rarely been explored even though it has a wide range of scenarios in ITS. To overcome this limitation, we analyze the light transport of TIR light. Our analysis reveals that contours are the reliable features shared by TIR and Visible Spectrum (VS) light. Inspired by this, we attempt to transfer joint features from VS domain to TIR domain. Thus, we propose a curriculum domain adaptation method to guide the TIR urban scene semantic segmentation task from VS domain through contours. Moreover, to evaluate the proposed model, we build TIR-SS: an open-for-request dataset consisting of TIR images and pixel level annotations of 8 classes in urban scenes. Qualitative and quantitative experimental results on the dataset indicate that the proposed domain adaptation method outperforms related methods on this TIR semantic segmentation task.
16 March 2016 in International Journal of General Systems
In this paper, the minimum spanning tree problem is investigated on the graph with fuzzy edge weights. The notion of fuzzy -minimum spanning tree is presented based on the credibility measure, and then the solutions of the fuzzy -minimum spanning tree problem are discussed under different assumptions. First, we respectively, assume that all the edge weights are triangular fuzzy numbers and trapezoidal fuzzy numbers and prove that the fuzzy -minimum spanning tree problem can be transformed to a classical problem on a crisp graph in these two cases, which can be solved by classical algorithms such as the Kruskal algorithm and the Prim algorithm in polynomial time. Subsequently, as for the case that the edge weights are general fuzzy numbers, a fuzzy simulation-based genetic algorithm using Prüfer number representation is designed for solving the fuzzy -minimum spanning tree problem. Some numerical examples are also provided for illustrating the effectiveness of the proposed solutions.
2 December 2014 in Journal of Intelligent Manufacturing