Search

The specific chemical and physical evolutions of electrode materials under operating conditions should be understood to optimize their electrochemical performances. The in-situ/operando techniques including Raman spectrum, transmission electron microscope, X-ray diffraction, X-ray absorption spectrum, and magnetization are powerful tools, which can provide the real-time surficial/interfacial changes of electrodes, the transformation of crystal lattice structures, the adjustment of electronic states and even the influence of magnetic properties under operating conditions. In this Review, the advantages and limitations of these in-situ/operando techniques in investigating the inner energy storage mechanisms of various type electrode materials are analyzed. The representative research results such as the ion dependent storage mechanism, step-alloying processes and space charge storage theory are highlighted. In addition, the challenges and opportunities of in-situ/operando characterizations are proposed as well.

Ultrathin core-shell V3 S4 @C nanosheets assembled into hierarchical nanotubes (V3 S4 @C NS-HNTs) are synthesized by a self-template strategy and evaluated as general anodes for alkali-ion batteries. Structural/physicochemical characterizations and DFT calculations bring insights into the intrinsic relationship between crystal structures and electrochemical mechanisms of the V3 S4 @C NS-HNTs electrode. The V3 S4 @C NS-HNTs are endowed with strong structural rigidness owing to the layered VS2 subunits and interlayer occupied V atoms, and efficient alkali-ion adsorption/diffusion thanks to the electroactive V3 S4 -C interfaces. The resulting V3 S4 @C NS-HNTs anode exhibit distinct alkali-ion-dependent charge storage mechanisms and exceptional long-durability cyclic performance in storage of K+ , benefiting from synergistic contributions of pseudocapacitive and reversible intercalation/de-intercalation behaviors superior to those of the conversion-reaction-based Li+ -/Na+ -storage counterparts.

构建基于有机材料的高性能柔性储钠电极面临诸多挑战. 本 工作通过可控组装及还原的方式, 实现了铁基配位聚合物纳米线/还原氧化石墨烯柔性薄膜的构筑. 多维复合薄膜可直接用作钠离 子电池自支撑负极, 且具有较高的储钠容量(200 mA g−1电流密度 下可逆容量为319 mA h g−1)和优异的倍率性能(3000 mA g−1大电 流密度下比容量可保持在∼120 mA h g−1). 研究表明有机配体(氨 三乙酸)中的羧基及氨基为储钠活性位点, 而配位金属离子(Fe2+)不 参与电化学反应.

Flexible non-lithium (Na+, K+, Zn2+, and Al3+) based rechargeable batteries are promising power sources in the emerging field of flexible and wearable electronic devices due to their low cost and wide availability. In this review, we mainly summarized the latest contributions and progress in non-lithium based secondary batteries. Initially, a brief introduction to the structural superiority and working mechanisms of non-lithium based rechargeable batteries is provided. Then, smart design and rational construction of flexible electrodes involved in non-lithium ion batteries, including various flexible substrates and inorganic/organic/metal–organic framework based cathode and anode materials, were elaborately discussed and evaluated. Besides, electrolytes, especially the solid-state electrolytes and separators for flexible batteries were further comprehensively discussed, along with the design and packaging of flexible full batteries. Finally, the future trends, challenges and prospects are proposed for advanced flexible non-lithium based rechargeable batteries.

Metallic bismuth has drawn attention as a promising alloying anode for advanced potassium ion batteries (PIBs). However, serious volume expansion/electrode pulverization and sluggish kinetics always lead to its inferior cycling and rate properties for practical applications. Therefore, advanced Bi-based anodes via structural/compositional optimization and sur-/interface design are needed. Herein, we develop a bottom-up avenue to fabricate nanoscale Bi encapsulated in a 3D N-doped carbon nanocages (Bi@N-CNCs) framework with a void space by using a novel Bi-based metal-organic framework as the precursor. With elaborate regulation in annealing temperatures, the optimized Bi@N-CNCs electrode exhibits large reversible capacities and long-duration cyclic stability at high rates when evaluated as competitive anodes for PIBs. Insights into the intrinsic K+ -storage processes of the Bi@N-CNCs anode are put forward from comprehensive in situ characterizations.

Background: Mesenchymal stem cells (MSCs) selectively differentiate into adipocytes or osteoblasts, and several molecules control the fate determination of MSCs. Understanding these key checkpoints greatly contributes to the ability to induce specific MSC differentiation for clinical applications. In this study, we aimed to explore whether TNF receptor-associated factor 4 (TRAF4) affects MSC adipogenic differentiation, which we previously reported that could positively regulated the osteogenic differentiation. Methods: Western blotting and Real-time Polymerase Chain Reaction were used to detected the expression pattern of TRAF4 during adipogenic differentiation. Lentivirus was constructed to regulate TRAF4 expression, and oil red O staining and Western blotting were used to assess its role in adipogenesis, which was confirmed in vivo by implanting an MSC-matrigel mixture into nude mice. Western blotting was used to detect the activated signaling pathways, and a specific inhibitor and agonist were used to clear the roles of the key signaling pathways. Additionaly, Co-Immunoprecipitation was conducted to find that Pyruvate kinase isozyme type M2 (PKM2) interacts with TRAF4, and to further explore their binding and functional domains. Finally, an RNA-binding protein immunoprecipitation assay and Western blotting were used to detect whether N6-methyladenosine mediates the decreased TRAF4 expression during adipogenic differentiation. Findings: The results demonstrated that TRAF4 negatively regulates MSC adipogenesis in vitro and in vivo. Mechanistically, we revealed that TRAF4 binds to PKM2 to activate the kinase activity of PKM2, which subsequently activates β-catenin signaling and then inhibits adipogenesis. Furthermore, TRAF4 downregulation during adipogenesis is regulated by ALKBH5-mediated N6-methyladenosine RNA demethylation. Interpretation: TRAF4 negatively regulates the adipogenesis of MSCs by activating PKM2 kinase activity, which may act as a checkpoint to fine-tune the balance of adipo-osteogenic differentiation, and suggests that TRAF4 may be a novel target of MSCs in clinical use and may also illuminate the underlying mechanisms of bone metabolic diseases. Funding: This study was supported by the National Natural Science Foundation of China (81871750 and 81971518) and the Science and Technology Project of Guangdong Province (2019B02023600 and 2017A020215070). Keywords: TRAF4, Mesenchymal stem cells, Adipogenic differentiation, PKM2

Recently, there has been an urgent need for flexible and low cost rechargeable batteries for the emerging flexible and wearable electronic devices. Herein, MnS nanoparticles embedded in carbon nanowires/reduced graphene oxide (MnS@CNWs/rGO) composite paper were synthesized via a simple yet scalable strategy with Mn based coordination nanowires and graphene oxide as precursors. The combination of multi-dimensional subunits offers not only a robust structure but also abundant pathways for fast electron/ion diffusion. When directly used as a free-standing electrode for sodium ion batteries (SIBs), the ultra-flexible paper anode exhibits excellent mechanical and electrochemical performance, benefitting from the synergistic effects between nano-dimensional MnS encapsulated in CNWs and conductive rGO nanosheets. Remarkably, a high reversible gravimetric/volumetric capacity of ∼560 mA h g-1/∼362.3 mA h cm-3 is obtained using the self-supported flexible electrode at a current density of 0.1 A g-1, which is almost 92.4% of the theoretical capacity of MnS. More competitively, the flexible MnS@CNWs/rGO anode exhibits an unprecedented long cycle life with a high reversible capacity of ∼150 mA h g-1 at 1 A g-1 after 10, 000 cycles. This highly favours the promising application of MnS@CNWs/rGO paper in advanced flexible SIBs as an appealing anode.

Although metal ore mining activities are well known as an important source of heavy metals, soil pollution caused by small-scale mining activities has long been overlooked. This study investigated the pollution of surface soils in an area surrounding a recently abandoned small-scale polymetallic mining district in Guangdong province of south China. A total of 13 tailing samples, 145 surface soil samples, and 29 water samples were collected, and the concentrations of major heavy metals, including Cr, Mn, Co, Ni, Cu, Zn, As, Cd, Pb, and Se, were determined. The results show that the tailings contained high levels of heavy metals, with Cu, Zn, As, Cd, and Pb occurring in the ranges of 739–4.15 × 103, 1.81 × 103–5.00 × 103, 118–1.26 × 103, 8.14–57.7, and 1.23 × 103–6.99 × 103 mg/kg, respectively. Heavy metals also occurred at high concentrations in the mine drainages (15.4–17.9 mg/L for Cu, 21.1–29.3 mg/L for Zn, 0.553–0.770 mg/L for Cd, and 1.17–2.57 mg/L for Pb), particularly those with pH below 3. The mean contents of Cu, Zn, As, Cd, and Pb in the surface soils of local farmlands were up to 7 times higher than the corresponding background values, and results of multivariate statistical analysis clearly indicate that Cu, Zn, Cd, and Pb were largely contributed by the mining activities. The surface soils from farmlands surrounding the mining district were moderately to seriously polluted, while the potential ecological risk of heavy metal pollution was extremely high. It was estimated that the input fluxes from the mining district to the surrounding farmlands were approximately 17.1, 59.2, 0.311, and 93.8 kg/ha/yr for Cu, Zn, Cd, and Pb, respectively, which probably occurred through transport of fine tailings by wind and runoff, and mine drainage as well. These findings indicate the significant need for proper containment of the mine tailings at small-scale metal ore mines.

Small-scale mining activities in many developing countries have caused severe environmental issues to the surrounding areas, which ultimately threatened the health of local populations. Based on detailed characterization of the local drinking water and surface soil, as well as foodstuffs, this study comprehensively assessed the public health risk of toxic metal(loid)s to the population living in three villages surrounding an abandoned small-scale polymetallic mine in southern China. The agricultural soils contained elevated levels of Cu, Zn, As, Cd, and Pb, which originated from the mining district, and as expected, the locally cultivated rice and vegetables were contaminated by As, Cd, and Pb to varying extents. Arsenic occurred in both inorganic and organic forms in the rice and vegetables, with inorganic As (i-As) accounting for 82.2% (45.4–100%) and 94.7% (65.2–100%) of the total As contents in rice and vegetables, respectively. Results of health risk assessment indicate that the residents in the impacted villages had serious non-carcinogenic and carcinogenic risk. Dietary exposure to i-As and Cd through rice and vegetable consumption was the primary cause of non-carcinogenic risk, while i-As intake was the dominant contributor of carcinogenic risk. These findings suggest that significant environmental pollution by toxic metal(loid)s could result from small-scale metal mines, even after being abandoned, and the accumulation of the toxic metal(loid)s in food crops could pose significant health risk to the local residents. Immediate actions should be taken to discourage them from consuming the locally produced food crops, while long-term control measures for containment of toxic metal(loid) pollution are being developed, and high priority should be given to the remediation of Cd and As in the contaminated soils.

Heavy metal(loid)s are natural constituents of the Earth's crust, and apportionment of their sources in surface soils is a challenging task. This study evaluated the application of positive matrix factorization (PMF) model, assisted with regression modeling and geospatial mapping, in the quantitative source apportionment of heavy metal(loid)s in the agricultural soils of Handan, a region covering >12,000 km2. Obvious enrichment of As, Cd, Cu, Pb, and Zn was found in the surface soils, with Cd alone accounted for 73 % of the overall potential ecological risk. PMF model revealed that Cd (56.9 %) and Pb (47.8 %) in the region's agricultural soils were predominantly contributed by industrial sources, Fe (71.8 %), Cr (60.0 %), V (52.9 %), Cu (50.7 %), Ni (42.2 %), and Mn (41.4 %) were primarily of lithogenic origin, while Co (54.1 %), As (42.9 %), and Zn (40.0 %) mainly came from the mixed sources of natural background, agricultural sources, and vehicle emissions. Uncertainty analysis showed that the contributions of pollution sources to the soil heavy metal(loid)s estimated by PMF model had considerable variations. While quantitative source apportionment of heavy metal(loid)s in soils could be achieved with PMF based on their spatial distributions, combination with emission inventory and reactive transport are probably necessary to obtain more accurate results.

Ultrathin core-shell V

The Zimudang gold deposit is a large Carlin-type gold deposit in the Southwest Guizhou Province, China, with an average Au content of 6.2 g/t. Gold is mainly hosted in the fault zone and surrounding strata of the F1 fault and Permian Longtan Formation, and the ore bodies are strictly controlled by both the faults and strata. Detailed mineralogy and geochemistry studies are conducted to help judge the nature of ore-forming fluids. The results indicate that the Au is generally rich in the sulfides of both ores and wall rocks in the deposit, and the arsenian pyrite and arsenopyrite are the main gold-bearing sulfides. Four subtypes of arsenian pyrite are found in the deposit, including the euhedral and subhedral pyrite, framboidal pyrite, pyrite aggregates and pyrite veins. The euhedral and subhedral pyrite, which can take up about 80% of total pyrite grains, is the dominant type. Au distributed unevenly in the euhedral and subhedral pyrite, and the content of the Au in the rim is relatively higher than in the core. Au in the pyrite veins and pyrite aggregates is lower than the euhedral and subhedral pyrite. No Au has been detected in the points of framboidal pyrites in this study. An obvious highly enriched As rim exists in the X-ray images of euhedral pyrites, implying the ore-forming fluids may be rich in As. The relationship between Au and As reveals that the Au may host as a solid solution (Au+) and nanoparticles of native gold (Au0) in the sulfides. The high Co/Ni ratio (>1) of sulfides and the enrichment of W in the ores all reflect that the gold-bearing minerals and ore-forming process were mainly related to the hydrothermal fluids, but the magmatic and volcanic activities cannot be neglected. The general existence of Au and As in the sulfides of both ores and wall rocks and the REE results suggest that the ore-forming fluids may mainly be derived from the basin itself. The enrichment of Tl suggests that the ore-forming fluids may be enriched in Cl. The Ce and Eu show slightly or apparently negative anomalies, which means the ore fluids were probably formed under reducing environment. The Y/Ho ratios of ore samples fluctuate around 28, implying the bicarbonate complexation and fluorine were both involved in the ore-forming process. Combined with the previous studies and our results, we infer that the ore-forming fluids enriched Au, As, HS− and halogen (F, Cl) were derived from the mixture of reducing basinal fluids and magmatic or volcanic hydrothermal fluids.

Abandoned tailings are one of the most important sources of heavy metal pollution in the areas surrounding mining districts, and significant leaching of heavy metals could be brought by precipitation. This study investigated the leaching of heavy metals from the tailings of a small-scale abandoned polymetallic mine in south China by rainwater with batch and column tests and evaluated the associated environmental impact. The mean contents of Cr, Ni, Cu, Zn, As, Cd, and Pb in the un-weathered mine tailings were 1.46 × 102, 3.11 × 102, 4.10 × 103, 2.18 × 104, 2.82 × 102, 5.65 × 102, and 8.74 × 103 mg/kg, respectively, and appreciable fractions of Cd, Zn, Cu, and Cr in the tailings were present in the acid soluble form. Batch and column leaching tests consistently showed that significant quantities of heavy metals could be released from the mine tailings. Based on the results of column leaching tests, it was estimated that the average fluxes of Cr, Ni, Cu, Zn, As, Cd, and Pb from the mine tailings at the studied mining district leached by precipitation were 3.20, 38.3, 12.5, 1.52 × 104, 104, 1.08, and 9.26 g/ha/yr, respectively. The metal-rich tailing leachate would impact the quality of surface water and soils downhill of the mining district, and pose significant potential ecological risk to the farmland soils, which are irrigated by local surface water. These findings indicate the importance of tailings as a source of heavy metals in the mining districts of south China with heavy precipitation, as well as the need for mitigating the releases of heavy metals and the associated environmental impact from abandoned mine tailings.

The traditional industrial operations are well recognized as an important source of heavy metal pollution, while that caused by the e-waste recycling activities, which have sprouted in some developing countries, is often overlooked. This study was carried out to compare the status of soil heavy metal pollution caused by the traditional industrial operations and the e-waste recycling activities in the Pearl River Delta, and assess whether greater attention should be paid to control the pollution arising from e-waste recycling activities. Both the total contents and the chemical fractionation of major heavy metals (As, Cr, Cd, Ni, Pb, Cu, and Zn) in 50 surface soil samples collected from the e-waste recycling areas and 20 soil samples from the traditional industrial zones were determined. The results show that the soils in the e-waste recycling areas were mainly polluted by Cu, Zn, As, and Cd, while Cu, Zn, As, Cd, and Pb were the major heavy metals in the soils from the traditional industrial zones. Statistical analyses consistently show that Cu, Cd, Pb, and Zn in the surface soils from both types of sites were contributed mostly by human activities, while As, Cr, and Ni in the soils were dominated by natural background. No clear distinction was found on the pollution characteristic of heavy metals in the surface soils between the e-waste recycling areas and traditional industrial zones. The potential ecological risk posed by heavy metals in the surface soils from both types of sites, which was dominated by that from Cd, ranged from low to moderate. Given the much shorter development history of e-waste recycling and its largely unregulated nature, significant efforts should be made to crack down on illegal e-waste recycling and strengthen pollution control for related activities.

Past work on object detection has emphasized the issues of feature extraction and classification, however, relatively less attention has been given to the critical issue of feature selection. The main trend in feature extraction has been representing the data in a lower dimensional space, for example, using principal component analysis (PCA). Without using an effective scheme to select an appropriate set of features in this space, however, these methods rely mostly on powerful classification algorithms to deal with redundant and irrelevant features. In this paper, we argue that feature selection is an important problem in object detection and demonstrate that genetic algorithms (GAs) provide a simple, general, and powerful framework for selecting good subsets of features, leading to improved detection rates. As a case study, we have considered PCA for feature extraction and support vector machines (SVMs) for classification. The goal is searching the PCA space using GAs to select a subset of eigenvectors encoding important information about the target concept of interest. This is in contrast to traditional methods selecting some percentage of the top eigenvectors to represent the target concept, independently of the classification task. We have tested the proposed framework on two challenging applications: vehicle detection and face detection. Our experimental results illustrate significant performance improvements in both cases.

Developing on-board automotive driver assistance systems aiming to alert drivers about driving environments, and possible collision with other vehicles has attracted a lot of attention lately. In these systems, robust and reliable vehicle detection is a critical step. This paper presents a review of recent vision-based on-road vehicle detection systems. Our focus is on systems where the camera is mounted on the vehicle rather than being fixed such as in traffic/driveway monitoring systems. First, we discuss the problem of on-road vehicle detection using optical sensors followed by a brief review of intelligent vehicle research worldwide. Then, we discuss active and passive sensors to set the stage for vision-based vehicle detection. Methods aiming to quickly hypothesize the location of vehicles in an image as well as to verify the hypothesized locations are reviewed next. Integrating detection with tracking is also reviewed to illustrate the benefits of exploiting temporal continuity for vehicle detection. Finally, we present a critical overview of the methods discussed, we assess their potential for future deployment, and we present directions for future research.

As one of the most promising applications of computer vision, vision-based vehicle detection for driver assistance has received considerable attention over the last 15 years. There are at least three reasons for the blooming research in this field: first, the startling losses both in human lives and finance caused by vehicle accidents; second, the availability of feasible technologies accumulated within the last 30 years of computer vision research; and third, the exponential growth of processor speed has paved the way for running computation-intensive video-processing algorithms even on a low-end PC in realtime. This paper provides a critical survey of recent vision-based on-road vehicle detection systems appeared in the literature (i.e., the cameras are mounted on the vehicle rather than being static such as in traffic/driveway monitoring systems).

Despite the considerable amount of research work on the application of Gabor filters in pattern classification, their design and selection have been mostly done on a trial and error basis. Existing techniques are either only suitable for a small number of filters or less problem-oriented. A systematic and general evolutionary Gabor filter optimization (EGFO) approach that yields a more optimal, problem-specific, set of filters is proposed in this study. The EGFO approach unifies filter design with filter selection by integrating genetic algorithms (GAs) with an incremental clustering approach. Specifically, filter design is performed using GAs, a global optimization approach that encodes the parameters of the Gabor filters in a chromosome and uses genetic operators to optimize them. Filter selection is performed by grouping together filters having similar characteristics (i.e., similar parameters) using incremental clustering in the parameter space. Each group of filters is represented by a single filter whose parameters correspond to the average parameters of the filters in the group. This step eliminates redundant filters, leading to a compact, optimized set of filters. The average filters are evaluated using an application-oriented fitness criterion based on support vector machines (SVMs). To demonstrate the effectiveness of the proposed framework, we have considered the challenging problem of vehicle detection from gray-scale images. Our experimental results illustrate that the set of Gabor filters, specifically optimized for the problem of vehicle detection, yield better performance than using traditional filter banks.

Feature subset selection has received considerable attention in the machine learning literature, however, it has not been fully explored or exploited in the computer vision area. In this paper, we consider the problem of object detection using genetic algorithms (GA) for feature subset selection. We argue that feature selection is an important problem in object detection, and demonstrate that GA provide a simple, general, and powerful framework for selecting good sets of features, leading to lower detection error rates. As a case study, we have chosen to perform feature extraction using the popular method of principal component analysis (PCA) and classification using support vector machines (SVM). We have tested this framework on. two difficult and practical object detection problems: vehicle detection and face detection. Experimental results demonstrate significant performance improvements in both cases.

The focus of this work is on the problem of feature selection and classification for on-road vehicle detection. In particular, we propose using quantized Haar wavelet features and Support Vector Machines (SVMs) for rear-view vehicle detection. Wavelet features are particularly attractive for vehicle detection because they form a compact representation, encode edge information, capture information from multiple scales, and can be computed efficiently. Traditionally, methods using wavelet features for classification truncate the coefficients by keeping only the ones with largest magnitude. We believe that the actual values of the wavelet coefficients are not very important for vehicle detection. In fact, the coefficient magnitudes indicate local oriented intensity differences, information that cold be very different even for the same vehicle under different lighting conditions. Therefore, we argue and demonstrate experimentally that the actual coefficient values are less important compared to the simple presence or absence of those coefficients. Specifically, we propose quantizing large negative coefficients to -1, large positive coefficients to 1, and setting the rest coefficients to 0. The quantized coefficients seem to encode important information about the general shape and structure of vehicles, while ignoring fine details and allowing for sufficient inter-class variability. Experimental results and comparisons using real data demonstrate the superiority of the proposed approach which has achieved an average accuracy of 93.94% on completely novel test images.

We present a distributed vision-based surveillance system. The system acquires and processes grey level images through one or multiple camera units monitoring certain area(s) via a local area network (LAN) and is capable of combining information from multiple camera units to obtain a consensus decision. It can be trained to detect certain type of intrusions, for example pedestrians, a group of pedestrians, vehicles, pets, etc., and minimizes false alerts due to other non-interested intrusions. As a case study, we aim to detect pedestrian/vehicle in an observation area. Our vision-based intrusion detection approach consists of two main steps: background subtraction based hypothesis generation (HG) and appearance-based hypothesis verification (HV). HG hypothesizes possible threats (intrusions), and HV verifies those hypotheses using a Gabor filter for feature extraction and support vector machines (SVMs) for classification. The system has been tested in an unconstrained outdoor environment, illustrating good performance.

This paper presents an in-vehicle real-time monocular precrash vehicle detection system. The system acquires grey level images through a forward facing low light camera and achieves an average detection rate of 10Hz. The vehicle detection algorithm consists of two main steps: multi-scale driven hypothesis generation and appearance-based hypothesis verification. In the multi-scale hypothesis generation step, possible image locations where vehicles might be present are hypothesized. This step uses multi-scale techniques to speed up detection but also to improve system robustness by making system performance less sensitive to the choice of certain parameters. Appearance-base hypothesis verification verifies those hypothesis using Haar Wavelet decomposition for feature extraction and Support Vector Machines (SVMs) for classification. The monocular system was tested under different traffic scenarios (e.g., simply structured highway, complex urban street, varying weather conditions), illustrating good performance.

We consider the problem of gender classification from frontal facial images using feature selection and neural networks. We argue that feature selection is an important issue in gender classification and we demonstrate that by removing features that do not encode important gender information from the image representation of faces, the error rate can be reduced significantly. Automatic feature subset selection is used. First, principal component analysis (PCA) is used to represent each image as a feature vector (i.e., eigen-features) in a low-dimensional space, spanned by the eigenvectors of the covariance matrix of the training images (i.e., coefficients of the linear expansion). A genetic algorithm (GA) is then used to select a subset of features from the low-dimensional representation by removing certain eigenvectors. Finally, a neural network is trained to perform gender classification using the selected eigen-feature subset. Experimental results demonstrate a significant improvement in error rate reduction. Using a subset of eigen-features containing only 18% of the features in the complete set, the average NN classification error goes down to 11.3% from an average error rate of 17.7%.

On-road vehicle detection is an important problem with application to driver assistance systems and autonomous, self-guided vehicles. The focus of this paper is on the problem of feature extraction and classification for rear-view vehicle detection. Specifically, we propose using Gabor filters for vehicle feature extraction and support vector machines (SVM) for vehicle detection. Gabor filters provide a mechanism for obtaining some degree of invariance to intensity due to global illumination, selectivity in scale, and selectivity in orientation. Basically, they are orientation and scale tunable edge and line detectors. Vehicles do contain strong edges and lines at different orientation and scales, thus, the statistics of these features (e.g., mean, standard deviation, and skewness) could be very powerful for vehicle detection. To provide robustness, these statistics are not extracted from the whole image but rather are collected from several subimages obtained by subdividing the original image into subwindows. These features are then used to train a SVM classifier. Extensive experimentation and comparisons using real data, different features (e.g., based on principal components analysis (PCA)), and different classifiers (e.g., neural networks (NN)) demonstrate the superiority of the proposed approach which has achieved an average accuracy of 94.81% on completely novel test images.

Appearance-based methods represent a promising research direction to the problem of vehicle detection. These methods learn the characteristics of the vehicle class from a set of training images which capture the variability in vehicle appearance. First, training images are represented by a set of features. Then, the decision boundary between the vehicle and nonvehicle classes is computed by modelling the probability distribution of the features In each class or through learning. The purpose of this study is to investigate the effectiveness of two important types of features for vehicle detection based on Haar wavelets and Gabor filters. In both cases, the decision boundary is computed using support vector machines (SVMs). Wavelet features encode edge information. Gabor filters provide a mechanism for obtaining orientation and scale tunable edge and line detectors. Our experimental results and comparisons using real data illustrate the effectiveness of both types of features for vehicle detection, with Gabor features performing the better. The two feature sets yield different misclassification errors which led us to the idea of combining them for improving performance. The combined set of features outperformed each feature set alone on completely novel test images.