- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
Real-time detection of moving cast shadows using foreground luminance statistics
摘要： This paper presents a robust real-time method for detection of moving cast shadows which employs the assumption of higher interdependence of luminance values for the shadow pixels in larger regions compared to the object pixels. First, a fast modified image differencing technique is used to separate foreground pixels from the background. Next, for a moving window of fixed width scanning the foreground regions, a new measure called Modified Correlation is introduced. The new measure is determined by first computing the correlation between the luminance values of the moving window and luminance values of its neighbouring windows; this correlation is then divided by a robust-to-noise range measured based on the luminance values of the moving window. The modified correlation exhibits abrupt rising transitions as it enters the shadow region from the object region, transitions which can be used to separate object pixels from shadow pixels. Thus, the new method is very effective at suppressing moving cast shadows, while avoiding limiting structures, unrealistic assumptions, the need for a-priori knowledge, and manual selection of critical parameters. An average shadow detection rate of 85.4% and an average shadow discrimination rate of 99.5% over multiple different sequences, higher than those of available methods in the literature, confirm the efficacy of the method. The robustness of the method is examined under different lighting conditions, different target-environment combinations, and sequences with object-shadow occlusion. The proposed method is computationally efficient and suitable for real-time situations.
关键词： Moving cast shadow,Image difference,Correlation,Sliding window,Foreground mask,Real-time detection
更新于2025-09-23 15:23:52
-
[IEEE 2018 China International SAR Symposium (CISS) - Shanghai (2018.10.10-2018.10.12)] 2018 China International SAR Symposium (CISS) - A Fast Target Detection Method for SAR Image Based on Electromagnetic Characteristics
摘要： Target detection for remote sensing images which contain optical images and radar images has attracted lots of relative researchers. With the development of deep learning, target detection for optical images has been developing towards high accuracy and real-time detection. High resolution optical images reflect geometric features of the object. Unlike optical images, SAR images reflect the electromagnetic characteristics of the target, so the SAR image detection which uses optical image detection algorithm will lead to weak detection performance. This paper studies a fast target detection algorithm for SAR images which fused electromagnetic characteristics and geometric features through support vector machine. The algorithm is based on the Faster R-CNN framework enabling nearly cost-free target detection.
关键词： real-time detection,scattering center model,electromagnetic characteristics,Faster R-CNN,target detection
更新于2025-09-23 15:22:29
-
Xenobiotic Contamination of Water by Plastics and Pesticides Revealed Through Real-time, Ultrasensitive and Reliable Surface Enhanced Raman Scattering
摘要： Uncontrolled utilization and consequent ubiquitous percolation of carcinogenic and xenobiotic contaminants, such as plasticizers and pesticides, into ecosystem has created an immediate demand for robust analytical detection techniques to identify their presence in water. Addressing this demand, we uncover the presence of xenobiotic contaminants such as Bisphenol A (BPA), Triclosan (TC), and Dimethoate (DM) through a robust, ultrasensitive and reliable Surface Enhanced Raman Scattering (SERS) platform. Thereby, conclusive real-time evidence of degradation of polyethylene terephthalate (PET) leading to release of BPA in water is presented. Worryingly, the release of BPA occurs at ambient temperature (40 0C) and within realistic timescales (12 hours) that are regularly encountered during the handling, transport and storage of PET-based water containers. Complementary mass-spectrometric, surface-specific atomic force microscopy and surface selective X-ray Photoelectron spectroscopy confirms the nanoscale surface degradation of PET through loss of C=O and C-O surface functionalities. Such ultra-sensitive (ppm-level), spectroscopic detection is enabled by the bottom-up assemblies of metal nanoparticles (Soret Colloids, SCs) acting as SERS platform to provide high analytical enhancement factor (108) with high reliability (relative standard deviation, RSD <5%). Effective and rapid detection (30 s) of several other potential xenobiotic contaminants such as Triclosan (TC) and Dimethoate (DM) over a wide range of concentrations (10-5 to 10-1 M) has also been demonstrated. Finally, non-destructive real-time spectroscopic “sniffing” of organophosphorous pesticides from the surface of fruits is achieved, illustrating the multi-phasic versatility of this label-free, non-lithography-based SERS platform.
关键词： plastic degradation,Soret colloids,water and food contamination,real-time detection,nanoparticle assembly,surface enhanced Raman scattering,Xenobiotics
更新于2025-09-23 15:21:01
-
A Real-Time Method to Detect the Deformation Behavior during Laser Solid Forming of Thin-Wall Structure
摘要： Laser solid forming (LSF) is a promising additive manufacturing technology. In the LSF process, deformation behaviors dictate the accuracy of the produced parts. In this study, by using a laser displacement detector based on laser triangulation principle, an accurate and effective real-time detection method was established to monitor the real-time deformation behavior of the key position during the LSF of a thin-wall structure. The results confirmed that increasing thin-wall length results in increasing final deformation of the edge. The displacement fluctuation range and value in the middle of thin wall are both smaller than that of the positions near the end, while the entire displacement changing direction in the middle is opposite to that of the end positions. When the deposition process is paused, the deformation of the thin wall during the cooling stage will deviate the position of the deposited thin wall, resulting in the dislocation between the subsequent deposited part and that before the pause, which affect the dimensional accuracy of the thin wall structure. This non-contact real-time detection method also confirmed the ability to monitor the initiation of cracking during the LSF process, and a potential to be used for the on-line feedback control of deformation of detected key position of deposited structure.
关键词： laser solid forming,real-time detection,accuracy,crack,additive manufacturing,deformation behaviors
更新于2025-09-23 15:19:57
-
A Fast Recursive Collaboration Representation Anomaly Detector for Hyperspectral Image
摘要： Even though collaboration representation-based detector (CRD) performs well for hyperspectral image (HSI) anomaly detection, its computational cost is too high for the widely demanded real-time applications. To reduce the computational complexity, a recursive CRD is proposed in this letter. By constructing two elementary transformation matrices in accordance with the location of the pixels, a recursive update approach is derived by a matrix inversion lemma to speed up the detector. Experimental results on two real HSI data sets show that the proposed method saves over 30% processing time without accuracy loss.
关键词： collaboration representation,Anomaly detection,real-time detection,hyperspectral image (HSI)
更新于2025-09-10 09:29:36