- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
Robust visible-infrared image matching by exploiting dominant edge orientations
摘要： Finding the correspondences between visible and infrared images is a challenging task due to the image spectral inconsistency which leads to large differences of gradient distributions between these images. To alleviate this problem, we propose a novel feature descriptor for visible and infrared image matching based on Log-Gabor filters. The descriptor employs multi-orientation and multi-scale Log-Gabor filters to encode the edge information statistically. Furthermore, the descriptor provides rotation invariance by estimating the dominant orientation which is based on accumulated edge orientations. The experimental results demonstrate the effectiveness of the proposed rotation invariant descriptor and the better performance for matching visible and longwave infrared images as compared with state-of-the-art descriptors.
关键词： Log-Gabor filters,rotation invariant descriptor,edge orientations,visible-infrared image matching
更新于2025-09-23 15:21:01
-
An Experimental Study on Infrared Radiation Characteristics of Sandstone Samples Under Uniaxial Loading
摘要： The failure of surrounding strata in the excavation faces is the root cause of some disasters, including roof caving, coal pillar instability, and water inrushes, during the extraction of coal resource. Identifying the abnormal phenomena and precursor information before rock failure forms an important theoretical foundations of achieving safety operation in coal mines. The infrared radiation (IR) on the surface of rock will change during loading process (Luong 1990; Freund et al. 2007; Gong 2013). By monitoring the IR, the deformation and failure characteristics of rock can be obtained (Sheinin and Blokhin 2012; Mineo and Pappalardo 2016; Sun et al. 2017; Fiorucci et al. 2018), reliable information for predicting rock failure precursors may be obtained, as well (Wu et al. 2015; Salami et al. 2017). In addition, the water content affects the IR characteristics during the failure of rock (Liu et al. 2010). To quantitatively characterize the IR from the rock surface during loading process, Wu (1998) first proposed the index of average infrared radiation temperature (AIRT). Later, Liu et al. (2015), Yang et al. (2017) and Ma et al. (2016) proposed new indexes for quantitative analysis, e.g., variance, entropy, characteristic roughness, euclidean distance, and VDIIT. The previous experimental results revealed that these new indexes could be used to explain a lot of IR phenomena during rock failure (Pappalardo 2017; Zhang et al. 2018). For example, the author have investigated the inherent relation between stress and IR during rock loading (Ma et al. 2017), and found that stress had significant, universal, and hysteretic control effects on IR. However, there are still difficulties during the acquisition of IR mutation data and the estimation of mutation amplitude under loading conditions. Therefore, new appropriate indexes should be proposed for more accurate determination of the quantitative relation between stress and IR. To further investigate the relation between stress and IR, two mutation coefficients, stress rate (SR) and variance of differential infrared image temperature (VDIIT), were adopted in this study. The related IR observation experiments were conducted using samples with different water content, so as to obtain the quantitative relation between stress and IR under uniaxial loading tests.
关键词： Stress rate (SR),Infrared radiation (IR),Control effect,Variance of differential infrared image temperature (VDIIT),Mutation coefficient
更新于2025-09-19 17:15:36
-
Deep infrared pedestrian classification based on automatic image matting
摘要： Infrared pedestrian classification plays an important role in advanced driver assistance systems. However, it encounters great difficulties when the pedestrian images are superimposed on a cluttered background. Many researchers design very deep neural networks to classify pedestrian from cluttered background. However, a very deep neural network associated with a high computational cost. The suppression of cluttered background can boost the performance of deep neural networks without increasing their depth, while it has received little attention in the past. This study presents an automatic image matting approach for infrared pedestrians that suppresses the cluttered background and provides consistent input to deep learning. The domain expertise in pedestrian classification is applied to automatically and softly extract foreground objects from images with cluttered backgrounds. This study generates trimaps, which must be generated manually in conventional approaches, according to the estimated positions of pedestrian's head and upper body without the need for any user interaction. We implement image matting by adopting the global matting approach and taking the generated trimap as an input. The representation of pedestrian is discovered by a deep learning approach from the resulting alpha mattes in which cluttered background is suppressed, and foreground is enhanced. The experimental results show that the proposed approach improves the infrared pedestrian classification performance of the state-of-the-art deep learning approaches at a negligible computational cost.
关键词： infrared image,pedestrian classification,image matting,deep learning
更新于2025-09-19 17:15:36
-
[IEEE 2019 18th International Conference on Optical Communications and Networks (ICOCN) - Huangshan, China (2019.8.5-2019.8.8)] 2019 18th International Conference on Optical Communications and Networks (ICOCN) - Study of Infrared Image Denoising Algorithm Based on Steering Kernel Regression Image Guided Filter
摘要： Infrared image has the disadvantages of narrow dynamic range, low contrast and easy to be polluted by noise. For this reason, this paper proposes a new infrared image denoising method based on steering kernel regression image guided filter. In this method, the shape and size of the steering kernel are firstly determined according to the local gradient information of the image, and then the filtering weight is obtained by the steering kernel regression, which is used to modify the analysis window of image guided filtering. The experimental results show that the proposed method is better than the guided filter and the steering kernel regression algorithm. Compared with the traditional infrared image denoising algorithm, this method can retain the image details, and acquire better visual experience.
关键词： Denoising,steering kernel regression,Infrared image,Image Guided Filter
更新于2025-09-16 10:30:52
-
The infrared moving target extraction and fast video reconstruction algorithm
摘要： Due to the problems of targets submerged to the background, which could easily produce ghosts, and hard complete extraction of dim targets in the surveillance video, we propose the moving target extraction and fast video reconstruction algorithm in accord with visual principle. The sample selection strategy of VIBE algorithm is improved to alleviate the errors of pixel classification. The infrared imaging features are fused to suppress the artifact. A regional growth mechanism is established to extract and store moving targets and pure background regions, and according to the characteristics of video surveillance, it is the first to establish the mapping mechanism of target, background and video to propose the fast video reconstruction algorithm. The experiment shows that the algorithm can extract the moving target completely, establish the pure background in a variety of complex conditions, and greatly reduce the storage room of the surveillance video.
关键词： background reconstruction,Infrared image,moving target extraction,VIBE
更新于2025-09-10 09:29:36
-
[IEEE 2018 IEEE 3rd Optoelectronics Global Conference (OGC) - Shenzhen, China (2018.9.4-2018.9.7)] 2018 IEEE 3rd Optoelectronics Global Conference (OGC) - A Directional-Progressive Search Method for Infrared Small Target Detection
摘要： Infrared small target detection plays a key role in infrared precision guidance and infrared early-warning system. It has been a difficult problem for researchers to study on how to detect targets accurately at a long distance as early as possible. Most of the existing algorithms can detect small targets in simple backgrounds, but they would fail on the detection when the background clutters are chaotic and the signal to clutter ratio (SCR) is low. Therefore, we propose a new infrared small target detection method which called a directional-progressive search (DPS) method. Our method derives from a fact that a small target is an isotropic Gaussian distribution at a long distance, while clutters show different characteristics in different directions. Based on this difference, we decompose the original image into first-order sub-images with different directions by using a first-order directional derivative (FODD) filter. Then zero-crossing points are detected in each direction step by step to distinguish small targets and background clutters. After screening progressively, the positions where existing zero-crossing points in every sub- image can be confirmed as targets. Experimental results show that our method acquires higher detection rates and lower false alarm rates compared with other methods. At the same time, our method can still keep better performance under various complex backgrounds. The robustness of our method is strong.
关键词： facet model,target detection,infrared image,zero-crossing point
更新于2025-09-10 09:29:36
-
Robust Infrared Small Target Detection Using Multiscale Gray and Variance Difference Measures
摘要： As a long-standing problem, infrared small target detection is challenging due to the dimness of targets and the complexity of background. Considering the limitation of traditional approaches, we propose an accurate and robust method for infrared small target detection using multiscale gray and variance difference measures. A multiscale adaptive gray difference measure is first used to enhance small targets and improve detection accuracy. Then, a multiscale variance difference measure is proposed to alleviate the impact of background fluctuation and improve the robustness of our method. By integrating these two approaches, targets can be extracted accurately using a threshold-adaptive segmentation. Extensive experiments have been conducted on datasets with various scenes. Results have demonstrated the effectiveness and outperformance of our method as compared to the state-of-the-art methods.
关键词： point target,Infrared image,multiscale gray difference measure,multiscale variance difference measure
更新于2025-09-10 09:29:36
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia, Spain (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - An Adaptation of Cnn for Small Target Detection in the Infrared
摘要： Due to the low signal to noise ratio and limited spatial resolution, small target detection in an infrared image is a challenging task. Existing methods often have high false alarm rates and low probabilities of detection when infrared small targets submerge in the background clutter. In this paper, the Convolutional Neural Network (CNN) is adapted to extract the hidden features of small targets from infrared imagery with a proposed technique for a large amount of training data generation. The Point Spread Function (PSF) is employed to model the small target data and generate positive samples. The random background image patches are selected as the negative samples. In this way, the detection problem is skillfully converted into a problem of pattern classification using CNN. Extensive synthetic and real small targets were tested to evaluate the performance of this novel small target detection framework. The experimental results indicate that the proposed algorithm is simple and effective with satisfactory detection accuracy.
关键词： Infrared image (IR),Convolutional Neural Network (CNN),Point Spread Function (PSF),small target detection
更新于2025-09-10 09:29:36
-
[IEEE 2018 IEEE 3rd Optoelectronics Global Conference (OGC) - Shenzhen, China (2018.9.4-2018.9.7)] 2018 IEEE 3rd Optoelectronics Global Conference (OGC) - Infrared Target Detection Based on Local Contrast Method and LK Optical Flow
摘要： A robust and effective small dim object detection algorithm is the key to the success of an infrared tracking system. To help solve practical tracking problems, a detecting algorithm based on local contrast method (LCM) and Lucas– Kanade method (LK) is put forward. Firstly, the local contrast map of the input image is obtained using the local contrast measure which measures the dissimilarity between the current location and its neighborhoods. In this way, target signal enhancement and background clutter suppression are achieved simultaneously. Secondly, an adaptive threshold is applied to extract the suspected object regions. Finally, the central points of obtained regions are used as characteristic points, then LK optical flow algorithm to calculate optical flow at these points, and through the instantaneous velocity calculation and selection targets are detected. The experimental result shows that this method works perfectly and can effectively detect infrared targets under complex backgrounds.
关键词： target detection,infrared image,lucas–kanade method (LK),local contrast method (LCM)
更新于2025-09-10 09:29:36
-
Generation of enhanced information image using curvelet-transform-based image fusion for improving situation awareness of observer during surveillance
摘要： Image fusion has been widely used to combine multispectral information into an enhanced information image. The application of such enhanced information content in the field of surveillance for improving situation awareness of observer is highly recommended. When a single sensor information is used for surveillance like visible camera output during poor ambient lighting conditions, ‘hot-target’ details are not visible to the observer. The use of visible-infrared fused image is recommended during surveillance in poor ambient lighting conditions to visualise background scene details and ‘hot-target’ details simultaneously. A wrapping-based curvelet transform method is proposed for fusion of infrared and visible images. Curvelet transform is used because of its advantages over wavelet transform limitations like directional insensitivity, isotropic basis and inability to resolve curves. The approximation coefficients are fused using the principal component analysis rule while detailed coefficients are fused using absolute maximum rule. The reconstructed fused image is compared with results of other fusion approaches proposed in literature. The performance of proposed wrapping-based curvelet fusion method is found visually and statistically better in comparison to other fused image outputs. The fused image obtained using proposed method retains background details as well as hot target presence with fidelity.
关键词： Curvelet transform,infrared image,situation awareness,visible image,principal component analysis,image fusion
更新于2025-09-09 09:28:46