Vehicle Vision Robust Detection and Recognition Method

Research on Vehicle Detection and Recognition Technology Based on Artificial Intelligence

Robust and reliable vehicle detection is a challenging task under the conditions of variable size and distance, various weather and illumination, cluttered background, the relative motion between the host vehicle and background. In this paper we investigate real-time vehicle detection using machine vision for active safety in vehicle applications. The conventional search method of vehicle detection is a full search one using image pyramid,which processes the image patches in same way and costs same computing time, even for no vehicle region according to prior knowledge.Our vehicle detection approach includes two basic phases. In the hypothesis generation phase, we determine the Regions of Interest (ROI) in an image according to lane vanishing points; furthermore, near, middle, and far ROIs, each with a different resolution, are extracted from the image. From the analysis of horizontal and vertical edges in the image, vehicle hypothesis lists are generated for each ROI. Finally, a hypothesis list for the whole image is obtained by combining these three lists. In the hypothesis validation phase, we propose a vehicle validation approach using Support Vector Machine (SVM) and Gabor feature. The experimental results show that the average right detection rate reach 90% and the average execution time is 30ms using a Pentium(R)4 CPU 2.4GHz.

Transportation plays an important role in urban development However, the vehicle growth in Indonesia is not supported by the number of road. Due to this fact, traffic congestion is easily occurred, especially in big cities. Intelligent Transportation System (ITS) has huge contribution to decrease the traffic congestion. In ITS, vehicle detection is one of challenging issue for traffic surveillance. In this paper, adaptive selective background learning and hole filling algorithm are applied to improve the vehicle detection. The validity of the proposed method is tested by using three scenarios and two parameters. The scenarios are bad weather close range (BW-CR), normal weather close range (NW-CR) and normal weather wide range (NW-WR). While, the parameters are the time duration of stopped vehicle detection and the pixel accuracy. Then, the proposed method (Adaptive Selective Background Learning-Hole Filling algorithm) is compared by another previous vehicle detection method. Generally, the result shows that the proposed method yields a significant improvement in vehicle detection. ASBL-HF can detect the stopped and moved vehicle with free noises. Moreover, ASBL-HF has the best accuracy. The accuracy value is about 98.2%.

With the rapid development of intelligent transportation systems, vehicle detection and traffic flow statistics have become increasingly critical in traffic management and the construction of smart cities. However, complex weather conditions lead to challenges in vehicle detection algorithms, increasing false negatives and false positives. Additionally, the rapid growth of vehicle ownership makes traditional algorithms inadequate for meeting practical application needs. Therefore, this paper proposes a collaborative optimization-based method combining YOLOv11 for vehicle detection and traffic flow analysis under complex weather conditions. Vehicle information is accurately extracted through object detection and tracking methods, and further analyzed using Logistic Regression to assess traffic congestion. The main contributions of this study are: (1) Integration of the CBAM attention mechanism into YOLOv11, enhancing the vehicle feature extraction ability through collaborative channel and spatial attention, thereby improving interference resistance under complex weather conditions. (2) Adoption of the Shape-IoU loss function, with dynamic weight allocation to optimize the accuracy of detecting smaller vehicles and the precision of bounding box regression in images where vehicles are located at a greater distance. (3) Construction of a traffic congestion prediction model based on the detection results, integrated with Logistic Regression. The proposed method's effectiveness is validated on the public dataset UA-DETRAC. Through the synergistic effect of the CBAM attention mechanism and Shape-IoU loss function, the improved model's mAP50-95 shows a 0.3 percentage point increase over the YOLOv11 baseline, providing efficient and reliable technical support for intelligent transportation systems. This study enhances the prediction performance of traffic states under complex weather conditions by effectively calculating traffic flow parameters such as vehicle count and density, thus offering foundational data support for traffic management decision-making. The research conclusions are of practical significance for urban traffic improvement.

With the potential to increase road safety and provide economic benefits, intelligent vehicles have elicited a significant amount of interest from both academics and industry. A robust and reliable vehicle detection and tracking system is one of the key modules for intelligent vehicles to perceive the surrounding environment. The millimeter-wave radar and the monocular camera are two vehicular sensors commonly used for vehicle detection and tracking. Despite their advantages, the drawbacks of these two sensors make them insufficient when used separately. Thus, the fusion of these two sensors is considered as an efficient way to address the challenge. This paper presents a collaborative fusion approach to achieve the optimal balance between vehicle detection accuracy and computational efficiency. The proposed vehicle detection and tracking design is extensively evaluated with a real-world data set collected by the developed intelligent vehicle. Experimental results show that the proposed system can detect on-road vehicles with 92.36% detection rate and 0% false alarm rate, and it only takes ten frames (0.16 s) for the detection and tracking of each vehicle. This system is installed on Kuafu-II intelligent vehicle for the fourth and fifth autonomous vehicle competitions, which is called “Intelligent Vehicle Future Challenge” in China.

Characterization of the spatial distribution and abundance of protein and RNAs within a tissue enables deep understanding of biologic systems. The ability to interrogate multiple targets simultaneously from clinical samples enhances the potential for discovery, but has proven to be challenging for formalin fixed paraffin embedded (FFPE) tissue. NanoString Technologies® has developed the Digital Spatial Profiling (DSP) platform to enable highly multiplexed profiling of protein or RNA from FFPE tissue using a non-destructive protocol. This technology can be used to quantify the abundance of targets within user-defined regions of interest (ROI) using a variety of masking strategies to select and define those regions, including geometric, phenotypic, or automatically generated masks based on expression of certain markers within the tissue. The current technology is capable of profiling dozens of targets, and future iterations could enable hundreds or thousands of targets to be profiled simultaneously. Controlling ROI size, shape, and features defines the heterogeneity and granularity of the information generated. To date, studies on the DSP have largely focused on profiling regions of interest in a nonsystematic way. However, profiling a tissue in a gridded fashion enables a uniform sampling of protein expression either across a whole tissue section or within a smaller area of the tissue to enable nonbiased evaluation, which may aid in biomarker discovery. This study utilizes DSP and a gridded ROI selection strategy to profile protein distribution within normal and tumor tissue. FFPE tissue from tonsil or colorectal tumors are stained with a 44-plex cocktail of antibodies conjugated to unique DNA oligos via a photocleavable linker and up to 3 fluorescent antibodies. Visual images of the tissue are collected with the fluorescent antibodies, and selected regions of interest are subsequently illuminated with UV light to release the oligos for collection. ROIs are profiled serially and captured oligos are then quantitated in the standard NanoString assay. In this study, tonsil and colorectal cancer (CRC) tissue is profiled on the DSP platform using a gridded ROI selection strategy to collect protein profiles from an entire tissue section either via low-resolution sampling or from specific regions within the tissue via high-resolution sampling. Low-resolution sampling of the tissue provides an estimate of the protein distribution across a section of tonsil or CRC. It further enables identification of regions of interest within the tumor suitable for deeper profiling with the high-resolution approach. The higher resolution profiling enables reconstruction of the tissue morphology at the molecular level by looking solely at the distribution of proteins. A distinct protein profiles for the germinal centers and T-cell zones within lymph nodes. Furthermore, we see variations emerges of protein expression between different lymph nodes that may contribute to different biologic functionality. Similarly, in CRC tissue, we see protein profiles that distribute uniquely between the tumor and the tumor microenvironment. For example, there is highly localized expression of immune cell type markers in particular, discrete regions of the tissue rather than distributed throughout the tumor microenvironment, and strong expression of pathways know to be involved in tumor growth/progression distributed to tumor-rich region of the tissue. Furthermore, there appears to be expression of select immune checkpoint molecules (e.g., PD-L1 and TIM3) in regions of the tissue that are predominantly tumor. Whether this represents tumor infiltration by immune cells or aberrant expression of the checkpoint molecules remains to be determined. These results demonstrate the utility of a gridded ROI selection strategy for deep profiling of tissue in an unbiased way to interrogate the immune biology within FFPE tissue using the NanoString DSP platform. Citation Format: Sarah Church, Chris Merritt, Giang Ong, Andrew White, Kristi Zevin, Sarah Warren, Joseph M. Beechem. Gridded tissue profiling strategy with digital spatial profiling for unbiased tissue sampling [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr B006.

The JPEG2000 image coding standard defines two kinds of region of interest (ROI) coding methods-the general scaling based method and the maximum shift (maxshift) method. The former requires shape coding of the ROIs, which leads to increased complexity of codec implementations and limits the choice of ROI shapes (currently, only rectangle and ellipse shapes are defined). The latter allows for arbitrarily shaped ROI coding without explicitly transmitting any shape information to the decoder, but does not have the flexibility to select an arbitrary scaling value to define the relative importance of the ROI and the background wavelet coefficients. We propose a bitplane-by-bitplane shift (BbBShift) method, which supports both arbitrary ROI shape and arbitrary scaling without shape coding.

Traffic congestion is a widespread problem. Dynamic traffic routing systems and congestion pricing are getting importance in recent research. Lane prediction and vehicle density estimation is an important component of such systems. We introduce a novel problem of vehicle self positioning which involves predicting the number of lanes on the road and vehicle's position in those lanes using videos captured by a dashboard camera. We propose an integrated closed-loop approach where we use the presence of vehicles to aid the task of self-positioning and vice versa. To incorporate multiple factors and high-level semantic knowledge into the solution, we formulate this problem as a Bayesian framework. In the framework, the number of lanes, the vehicle's position in those lanes and the presence of other vehicles are considered as parameters. We also propose a bounding box selection scheme to reduce the number of false detections and increase the computational efficiency. We show that the number of box proposals decreases by a factor of 6 using the selection approach. It also results in large reduction in the number of false detections. The entire approach is tested on real-world videos and is found to give acceptable results.

Vehicle detection plays a vital role in the design of Automatic Driving System (ADS), which has achieved remarkable improvements in recent years. However, vehicle detection in night scenes still has considerable challenges for the reason that the vehicle features are not obvious and are easily affected by complex road lighting or lights from vehicles. In this paper, a high-accuracy vehicle detection algorithm is proposed to detect vehicles in night scenes. Firstly, an improved Generative Adversarial Network (GAN), named Attentive GAN, is used to enhance the vehicle features of nighttime images. Then, with the purpose of achieving a higher detection accuracy, a multiple local regression is employed in the regression branch, which predicts multiple bounding box offsets. An improved Region of Interest (RoI) pooling method is used to get distinguishing features in a classification branch based on Faster Region-based Convolutional Neural Network (R-CNN). Cross entropy loss is introduced to improve the accuracy of classification branch. The proposed method is examined with the proposed dataset, which is composed of the selected nighttime images from BDD-100k dataset (Berkeley Diverse Driving Database, including 100,000 images). Compared with a series of state-of-the-art detectors, the experiments demonstrate that the proposed algorithm can effectively contribute to vehicle detection accuracy in nighttime.

Due to the dramatical grow of road transport, Advanced Driver Assistance Systems (ADAS) are being one of the most popular system. The main challenge of these systems is to improve driving safety and reduce accidents. Robust and effective vehicle detection is a critical step. However, vehicle detection meets many difficulties such as complex background, different size, model and orientations of vehicle. To solve this problem, this paper introduces an approach for traffic vehicle detection based on multi-scale covariance descriptor (MSCOV) for the image description and support vector machine classifier (SVM) for the data classification. This method is evaluated and compared to existing detection approach. The result of this approach outperforms existing vehicle detection system using the same dataset.

Nighttime vehicle detection is part of an intelligent transportation system for road safety, driving navigation, and surveillance at night. However, previous nighttime vehicle detection methods only deal with a single class or two classes of vehicles. This paper presents an effective system based on cascade feature selection and a coarse-to-fine mechanism for detecting preceding multiclass vehicles at night. First, we train a coarse-level classifier using contrast features. Second, we combine a cascade selection framework with feature mutual correlation, similarity and regions’ overlap to select a set of image regions, where we can extract effective features. Three features, including local binary pattern, histogram of oriented gradients, and four direction features, are extracted from the selected regions for training a fine-level multiclass vehicle classifier. During the detection stage, we utilize a coarse-to-fine mechanism. In the coarse level, a multiscale sliding window is classified by the coarse-level classifier to find regions of interest (ROIs) that are likely to be vehicles. In the fine level, these ROIs are identified by the trained multiclass vehicle classifier. Evaluations on a Hong Kong nighttime multiclass vehicle dataset show that our proposed system successfully detects a car, taxi, bus, and minibus within nighttime images under different scenes. Quantitatively, our proposed system obtains 95.48% detection rate at 0.055 false positives per image, outperforming some state-of-the-art detection approaches including two current nighttime vehicle detection methods, and two deep learning-based methods.

Vision based intelligent traffic management system is a robust framework that manages the on road traffic flow in real time by estimating traffic density near traffic signals. We have proposed a simple yet efficient algorithm to calculate the number of vehicles at various signals on a road to efficiently manage the traffic by controlling traffic signals to avoid congestion and traffic jam. The proposed system works by detection of vehicles in video frames acquired by cameras installed on roads and then perform accurate counting of vehicles at the same time. Dynamic background subtraction technique and morphological operations for vehicle detection have been used to achieve better detection efficiency. In order to attain accurate vehicle count in least possible time, we have used Region of Interest based method for vehicle calculation. The proposed framework is designed and implemented in several simulation test cases. It is expected that this work will provide an insight into the design and development of traffic signaling based system and also serves as a basis for practical implementation of a computer vision technology in real-time environment. Furthermore, this work also contributes to new design schemes to increase traffic signaling system's intelligence.

This paper presents a robustness benchmark evaluation and optimization for vehicle detection. Real-time vehicle detection has become an essential means of data perception in the transportation field, covering various aspects such as intelligent transportation systems, video surveillance, and autonomous driving. However, evaluating and optimizing the robustness of vehicle detection in real traffic scenarios remains challenging. When data distributions change, such as the impact of adverse weather or sensor damages, model reliability cannot be guaranteed. We first conducted a large-scale robustness benchmark evaluation for vehicle detection. Analysis revealed that adverse weather, motion, and occlusion are the most detrimental factors to vehicle detection performance. The impact of color changes and noise, while present, is relatively less pronounced. Moreover, the robustness of vehicle detection is closely linked to its baseline performance and model size. And as the severity of corruption intensifies, the performance of models experiences a sharp drop. When the data distribution of images changes, the features of the vehicles that the model focuses on are weakened, making the activation level of the targets significantly reduced. By evaluation, we provided guidance and direction for optimizing detection robustness. Based on these findings, we propose TDIRM, a traffic-degraded image restoration model based on stable diffusion, designed to efficiently restore degraded images in real traffic scenarios and thereby enhance the robustness of vehicle detection. The model introduces an image semantics encoder (ISE) module to extract features that align with the latent description of the real background while excluding degradation-related information. Additionally, a triple control embedding attention (TCE) module is proposed to fully integrate all condition controls. Through a triple condition control mechanism, TDIRM achieves restoration results with high fidelity and consistency. Experimental results demonstrate that TDIRM improves vehicle detection mAP by 6.92% on real dense fog data, especially for small distant vehicles that were severely obscured by fog. By enabling semantic-structural-content collaborative optimization within the diffusion framework, TDIRM establishes a novel paradigm for traffic scene image restoration.

Intelligent Transportation System (ITS) is the emerging transportation which comprises of the advanced technologies such as wireless communications, video vehicle detection, distributed system architectures, human machine interface, sensing and actuating, to improve safety of the passengers, traffic congestion, fuel consumption and optimize other services of the transportation system such as real time traffic situation, electronic toll collection, automatic road enforcement and hot lanes. The current ITS systems lacks self-adaptive capability to take spontaneous decisions. The environment of the ITS system is very complex so optimal decisions cannot be taken but a good enough decision can be taken which fulfills the requirement of the system. To build an ITS system having the capability to adapt itself, this paper presents the self-adaptive requirements for the ITS. To elicit the requirements for ITS, we use scenarios based model and goal based model using the requirements Engineering processes (Elicitation, Elaboration, Specification, Validation) with MAPE-K. We use Knowledge Acquisition on Automated Specification (KAOS) approach to build goal model using Objectiver tool, and use RELAX syntax to elicit adaptive requirements for ITS.

The dismal performance of the traditional traffic congestion measures and motorist attitudes on Nigeria roads have not only made traffic flow unpredictable and worsening but also worrisome and defiling every measures put-in-place by the Government. To solve these problems, the study assessed traffic congestion and Intelligent Transport System (ITS) in Ibadan, Nigeria. In achieving this, the study examined among others, factors aiding traffic congestion; effectiveness of the existing traffic congestion measures; and the performance and disposition to the use of ITS in the study area. Systematic random sampling was used to administer 386 copies of a questionnaire on commercial motorists, while data collected were analyzed using both descriptive and inferential (Pearson Chi-Square and Fisher's Exact Tests) statistics. Findings revealed that faulty intersections (3.586) and excesses of traffic officers (3.558) were worsening factors aiding traffic congestion in Ibadan city. Meanwhile, electronic traffic signalization (3.558) and law enforcement and regulation (3.499) were rated the most effective existing traffic congestion measures. In other words, real-time traffic signal devices (3.609); automated travel advisory system (3.595) and road congestion pricing devices (3.581) were highly rated and favourably disposed to use ITS. However, the Chi-Square Tests shows that the use Intelligent Transport System is statistically influenced by users' level of education (Sig. value < 0.05). The study concludes that there is a total necessity and willingness to use ITS in the study area. Hence, the study recommends exigent and holistic approach in the provision and enforcement of ITS towards mitigating congestion and other suggestions for improved traffic flow in Nigerian cities including Ibadan.