Vehicle Type Recognition Research Articles

基于改进的HMAX算法的车型识别应用

本文实现了将改进的HMAX算法应用于车型识别领域。算法的主要创新点是采用ITTI显著性算法选取车辆图片显著点构造图库，通过采用对颜色、亮度和朝向三方面敏感度的综合，构造更具代表性的模板库，提高了算法的识别率。此外，通过计算同一模板与不同图片响应度值的方差，消除了冗余模板，降低了算法识别时间。分析和实验表明，改进的HMAX算法可以有效地实现车型识别，新提出的算法相对原先HMAX模型在识别率可以提高1%~2%，相对当前存在的其他车型识别算法在识别率方面可提升约5%~10%，在保证特征数量的前提下，识别率接近95%；此外，加入模版筛选方法的改进算法相对于原先HMAX模型在识别率基本保持不变的条件下，识别时间可以缩减到原先的1/4，最终根据效益值评估的最优组合中，识别率约为92%，识别时间为0.6 s/幅图片，相对于原先方法均达到了一定的提升。 The paper implements the improved HMAX model for vehicle type (make and model) recognition. The main improvement of this algorithm is using ITTI model with the characteristic of quick response to color, intensity or orien- tation discrimination, to select the saliency area. By this way the template can be more representative and more beneficial to the detection rate. In addition, the calculating of the variance of response degree between different images and the same template eliminates redundant templates, which make contribution to reducing the time of classification. Analyz- ing and experiments claims that the improved HMAX model can be effective and reliable by the detection rate at 95%, with enough features extracted, 5% - 10% higher than current vehicle type recognition methods and 1% - 2% higher than original HMAX model. In addition, with the template screening method added, the improved HMAX model can keep the detection rate and curtail the classification time by quarter time. According to the optimal setting by the final value estimation, the detection rate is about 92%, and the classification time is 0.6 s/image, with promotion to the past.

Fast and simple vehicle logo location method

Simple appearance of the vehicle cannot be used for vehicle type recognition because of the different view angles,but vehicle logo can be used.The authors proposed a fast and simple method for vehicle logo location.Firstly,the vehicle license in front of the vehicle body could be roughly located by the texture of it,and then the position of the vehicle logo could be roughly found by taking account of the symmetrical prior knowledge.After that,the vehicle logo could be exactly detected by edge feature and morphology operation.Taking into account all kinds of noise,as well as the impact of morphology,template matching was used for precision positioning and rough identification.Experimental results prove the effectiveness of the proposed method.

Automated Identification of Vehicles using Acoustic Signal Processing

Automated identification of unauthorised intruding vehicles approaching protected infrastructure is becoming increasingly important for security purposes. This three‐year project, which is now in its second year, aims to develop a real‐time acoustic vehicle type recognition system that will be predominantly composed of three main parts; acoustic signal pre‐processing, feature extraction, and decision making (or classification). The main study area covers various signal processing techniques in time, time‐frequency and potentially frequency domains with signal classification implemented using a range of artificial intelligence techniques such as artificial neural networks. So far the focus has been on time domain signal processing and neural network classification. Whilst the work is at an early stage, the time domain methods such as Time Domain Signal Coding (TDSC) and Co‐Occurrence Matrix combined with neural networks have already shown some promise. The presentation will introduce the project, describe methodologies involved as well as the results to date for the novel acoustic identification of categories of vehicles.

Recognition of Proceeding Vehicles with Specific Information

Although the recognition of a license plate number or vehicle type has been researched, the recognition of vehicles using all features has not been studied due to its complexity. In this paper a novel method is proposed to identify vehicles with specific information, i.e., color, license plate, and vehicle's model. Low level image processing and texture descriptors are computed from the front image of vehicles. Then, two three- layer neural networks were built and trained for license plate and vehicle's model.

多重構造ニュラルネットワークを用いた車載センサ情報統合に基づく車種認識

An algorithm of vehicle type recognition integrating on-board sensor information is proposed. This method, compensates for apparent size differences due to the objective distances of on-board sensors, the distance and intensity detected by a scanning laser radar, and the images input from an on-board CCD camera. A frame recognition which consists of a multiplex structured neural network using the three kinds of adjusted data is performed, then the results are integrated to improve the accuracy of the recognition of each object. Experiments using road images captured while driving and the sensing data show that this method is effective : average recognition rates were above 96.1% for various road environments.

Acoustic and seismic signals of heavy military vehicles for co-operative verification

To analyze the potential of ground sensors for co-operative verification, the Bochum Verification Project has measured sound and soil vibration produced by military vehicles. Both signals contain contributions from the engine which usually dominate in the acoustic channel. With tracked vehicles, the track produces additional components which dominate the seismic signal at close to medium range and render it more than ten-fold stronger than that of wheeled vehicles. The acoustic amplitude decreases roughly with the inverse distance, while keeping the signal shape. The seismic amplitude decreases faster, but shows local variations; the seismic signal shape varies strongly with position. Acoustic and seismic spectra consist mostly of harmonic line series, caused by the engine and, if present, the track. Acoustic and seismic sensors can detect heavy vehicles passively, independent of weather and daylight, at more than 100 m range. More complex tasks, such as vehicle-type recognition or trajectory determination, require further research.

Acoustic–seismic detection and classification of military vehicles—developing tools for disarmament and peace-keeping

Acoustic and seismic signals of military vehicles passing between two newly-developed sensor stations separated by about 100 m distance have been measured. Five tracked and five wheeled vehicles drove with seven speeds in both directions on four lanes. Rms values rose above background at distances of several 100 m for tanks, 60–80 m for a medium truck, and 40–50 m for a light van. By processing the acoustic and seismic maxima together, tracked and wheeled vehicles were separated correctly in 94% of the passes. An algorithm for detecting lines and harmonic series in acoustic spectra worked for eight vehicles. For vehicle-type recognition, learning-vector quantisation was used, working on the relative powers of the first 15 harmonics. This was successful in 95–98% of the tracked-vehicle passes and 69–95% of the wheeled-vehicle passes. These experiences and results suggest that sensor monitoring of control lines is feasible.