Moving Vehicle Classification Using Cloud Model

2011 ◽  
Vol 467-469 ◽  
pp. 2123-2128
Author(s):  
Zhi Yuan Zeng ◽  
Bo Li ◽  
Xiao Jun Tan ◽  
Jian Zhong Zhou
Keyword(s):  
Cloud Model ◽  
Classification Model ◽  
Vehicle Classification ◽  
Model Parameters ◽  
Decision Tree Classifier ◽  
Shape Information ◽  
Vehicle Class ◽  
Moving Vehicle ◽  
Tree Classifier ◽  
Stable Performance

In this paper, we proposed a vehicle classification algorithm based on cloud model. Cloud model is a new theory which can express the relationship between randomness and fuzziness. Vehicle features, such as vehicle size, shape information, contour information and edge information are extracted for cloud model. Each vehicle class is expressed through cloud model parameters, such as Ex (expectation), En (entropy), with multi-dimensional feature. And cloud classification model is employed to judge the optimal class for each vehicle. Furthermore, attribute similarity is introduced to judge the weight of each feature in classification. Decision tree classifier is utilized for classification. The algorithm’s evaluations on video image series, the results show that cloud model ensures a promising and stable performance in recognizing these vehicle classes, and the algorithm can achieve accuracy and real-time.

Vehicle Classification and Tracking Based on Particle Swarm Optimization and Meanshift

2010 ◽  
Vol 121-122 ◽  
pp. 417-422
Author(s):  
Bo Li ◽  
Zhi Yuan Zeng ◽  
Ji Xiong Chen
Keyword(s):  
Particle Swarm Optimization ◽  
Fitness Function ◽  
Particle Swarm ◽  
Vehicle Classification ◽  
Shape Information ◽  
Vehicle Class ◽  
Swarm Optimization ◽  
Optimal Position ◽  
Moving Vehicle ◽  
Moving Vehicles

Vehicle classification and tracking is considered as one of the most challenging problems in the field of pattern recognition. In this paper, Particle Swarm Optimization (PSO) based method is exploited to recognize vehicle classes. Vehicle features, such as vehicle size, shape information, contour information are extracted. Each vehicle class is encoded as a centroid with multidimensional feature and PSO is employed to search the optimal position for each class centroid based on fitness function. After vehicle classification, an improved meanshift algorithm is presented for vehicle tracking. The algorithm’s evaluations on video image series, moving vehicle detection, vehicle classification and tracking are respectively conducted. The results show that PSO ensures a promising and stable performances in recognizing these vehicle classes, and the improved meanshift algorithm can achieve accuracy and real-time for tracking moving vehicles.

scholarly journals Classification of Level of Rice Pest Incidence Based on Weather Information

2021 ◽  
pp. 32-43
Author(s):  
V. Jinubala ◽  
P. Jeyakumar
Keyword(s):  
Gall Midge ◽  
Weather Conditions ◽  
Stem Borer ◽  
Vital Role ◽  
Classification Model ◽  
Decision Tree Classifier ◽  
Weather Information ◽  
Tree Classifier ◽  
Rice Pest

Aims: To classify the rice pest data based on the weather attributes using a machine learning approach, a decision tree classifier, and to validate the performance results with other existing techniques through comparison. Design: Rice pest classification using C5.0 algorithm Methodology: We collected rice pest data from the crop fields of various regions in the state of Maharashtra of India. The dataset contains the name of the region (Taluk), period (week), pest data, temperature, rainfall, and relative humidity. The data is collected from 39 taluks within four districts in different weeks of the year of 2013-2014. The weather information plays a vital role in this rice pest data analysis, because based on the weather, pest infestation varies in all the regions. The pests considered in this research are Yellow Stem borer, Gall midge, Leaf folder, and Planthopper. The collected dataset is given as input to the classifier, where 75% of data from the dataset is used for training, and 25% of data are used for testing the classifier. Results: The proposed C5.0 algorithm performed better in the classification of rice pest dataset based on weather attributes. The C5.0 algorithm achieved 88.99% accuracy, 78.81% sensitivity, and 89.11% specificity, which are higher in performance when compared with other techniques. Compared with the other different methods, the C5.0 algorithm achieved 1.3 to 8.5% improved accuracy, 2.4 to 9% improved sensitivity, and 0.8 to 7.8% improved specificity. Conclusion: Early detection of pest and pest based diseases is an essential process to avoid major crop losses. The proposed classification model is designed to classify the level of pest infestations based on weather attributes, as level of infestations caused by the rice pest varies based on weather conditions. The C5.0 algorithm classified the rice pest data based on the weather attributes in the dataset.

Vehicle Classification Model for Loop Detectors Using Neural Networks

2005 ◽  
Vol 1917 (1) ◽  
pp. 164-172 ◽  
Author(s):  
Yong-Kul Ki ◽  
Doo-Kwon Baik
Keyword(s):  
Neural Networks ◽  
Road Traffic ◽  
Recognition Rate ◽  
Classification Model ◽  
Vehicle Classification ◽  
Vehicle Class ◽  
Traffic Measurement ◽  
Loop Detectors ◽  
Variation Rate ◽  
The Neural Networks

Vehicle class is an important parameter in the process of road traffic measurement. Inductive loop detectors (ILD) and image sensors are rarely used for vehicle classification because of their low accuracy. To improve their accuracy, a new algorithm is suggested for ILD using backpropagation neural networks. In the developed algorithm, inputs to the neural networks are the variation rate of frequency and occupancy time. The output is five classified vehicles. The developed algorithm was assessed at test sites, and the recognition rate was 91.7%. Results verified that, compared with the conventional method based on ILD, the proposed algorithm improves the vehicle classification accuracy.

scholarly journals Development of Computer Vision System for Fruits

2021 ◽  
pp. 1-11
Author(s):  
Y. Dileep Sean ◽  
D.D. Smith ◽  
V.S.P. Bitra ◽  
Vimala Bera ◽  
Sk. Nafeez Umar
Keyword(s):  
Machine Learning ◽  
Computer Vision ◽  
Fruit Quality ◽  
Vision System ◽  
Machine Learning Algorithms ◽  
Classification Model ◽  
Support Vector ◽  
Small Scale ◽  
Decision Tree Classifier ◽  
Tree Classifier

Automated defect detection of fruits using computer vision and machine learning concepts has ‎become a significant area of research. In ‎this work, working prototype hardware model of conveyor with PC is designed, constructed and implemented to analyze the fruit quality. The prototype consists of low-cost microcontrollers, USB camera and MATLAB user interface. The automated classification model rejects or accepts the fruit based on the quality i.e., good (ripe, unripe) and bad. For the classification of fruit quality, machine learning algorithms such as Support Vector Machine, KNN, Random Forest classifier, Decision Tree classifier and ANN are used. The dataset used in this work consists of the following fruit varieties i.e., apple, orange, tomato, guava, lemon, and pomegranate. We trained, tested and ‎compared the performance of these five machine learning approaches and found out that the ANN based fruit detection performs better. The overall accuracy obtained by the ANN model for the dataset is 95.6%. In addition, the response time of the system is 50 seconds per fruit which is very low. Therefore, it will be very suitable and useful for small-scale industries and farmers to grow up their business.

scholarly journals An Efficient Cancer Classification Model Using Microarray and High-Dimensional Data

2021 ◽  
Vol 2021 ◽  
pp. 1-14
Author(s):  
Hanaa Fathi ◽  
Hussain AlSalman ◽  
Abdu Gumaei ◽  
Ibrahim I. M. Manhrawy ◽  
Abdelazim G. Hussien ◽  
...  
Keyword(s):  
Classification Accuracy ◽  
Large Scale ◽  
Gene Selection ◽  
Expression Profiles ◽  
Cancer Classification ◽  
Classification Model ◽  
Machine Learning Techniques ◽  
High Dimensional ◽  
Decision Tree Classifier ◽  
Tree Classifier

Cancer can be considered as one of the leading causes of death widely. One of the most effective tools to be able to handle cancer diagnosis, prognosis, and treatment is by using expression profiling technique which is based on microarray gene. For each data point (sample), gene data expression usually receives tens of thousands of genes. As a result, this data is large-scale, high-dimensional, and highly redundant. The classification of gene expression profiles is considered to be a (NP)-Hard problem. Feature (gene) selection is one of the most effective methods to handle this problem. A hybrid cancer classification approach is presented in this paper, and several machine learning techniques were used in the hybrid model: Pearson’s correlation coefficient as a correlation-based feature selector and reducer, a Decision Tree classifier that is easy to interpret and does not require a parameter, and Grid Search CV (cross-validation) to optimize the maximum depth hyperparameter. Seven standard microarray cancer datasets are used to evaluate our model. To identify which features are the most informative and relative using the proposed model, various performance measurements are employed, including classification accuracy, specificity, sensitivity, F1-score, and AUC. The suggested strategy greatly decreases the number of genes required for classification, selects the most informative features, and increases classification accuracy, according to the results.

scholarly journals A Framework for AI-Assisted Detection of Patent Ductus Arteriosus from Neonatal Phonocardiogram

Healthcare ◽  
2021 ◽  
Vol 9 (2) ◽  
pp. 169
Author(s):  
Sergi Gómez-Quintana ◽  
Christoph E. Schwarz ◽  
Ihor Shelevytsky ◽  
Victoriya Shelevytska ◽  
Oksana Semenova ◽  
...  
Keyword(s):  
Patent Ductus Arteriosus ◽  
Heart Defects ◽  
Ductus Arteriosus ◽  
Area Under The Curve ◽  
Clinical Decision ◽  
Heart Sounds ◽  
Decision Tree Classifier ◽  
Support Tool ◽  
Tree Classifier ◽  
Patent Ductus

The current diagnosis of Congenital Heart Disease (CHD) in neonates relies on echocardiography. Its limited availability requires alternative screening procedures to prioritise newborns awaiting ultrasound. The routine screening for CHD is performed using a multidimensional clinical examination including (but not limited to) auscultation and pulse oximetry. While auscultation might be subjective with some heart abnormalities not always audible it increases the ability to detect heart defects. This work aims at developing an objective clinical decision support tool based on machine learning (ML) to facilitate differentiation of sounds with signatures of Patent Ductus Arteriosus (PDA)/CHDs, in clinical settings. The heart sounds are pre-processed and segmented, followed by feature extraction. The features are fed into a boosted decision tree classifier to estimate the probability of PDA or CHDs. Several mechanisms to combine information from different auscultation points, as well as consecutive sound cycles, are presented. The system is evaluated on a large clinical dataset of heart sounds from 265 term and late-preterm newborns recorded within the first six days of life. The developed system reaches an area under the curve (AUC) of 78% at detecting CHD and 77% at detecting PDA. The obtained results for PDA detection compare favourably with the level of accuracy achieved by an experienced neonatologist when assessed on the same cohort.

scholarly journals Optimization of Physical Activity Recognition for Real-Time Wearable Systems: Effect of Window Length, Sampling Frequency and Number of Features

2019 ◽  
Vol 9 (22) ◽  
pp. 4833 ◽  
Author(s):  
Ardo Allik ◽  
Kristjan Pilt ◽  
Deniss Karai ◽  
Ivo Fridolin ◽  
Mairo Leier ◽  
...  
Keyword(s):  
Physical Activity ◽  
Feature Selection ◽  
Real Time ◽  
Sampling Frequency ◽  
Classification Performance ◽  
Window Length ◽  
Decision Tree Classifier ◽  
Acceleration Signal ◽  
Wearable Systems ◽  
Tree Classifier

The aim of this study was to develop an optimized physical activity classifier for real-time wearable systems with the focus on reducing the requirements on device power consumption and memory buffer. Classification parameters evaluated in this study were the sampling frequency of the acceleration signal, window length of the classification fragment, and the number of classification features, found with different feature selection methods. For parameter evaluation, a decision tree classifier was created based on the acceleration signals recorded during tests, where 25 healthy test subjects performed various physical activities. Overall average F1-score achieved in this study was about 0.90. Similar F1-scores were achieved with the evaluated window lengths of 5 s (0.92 ± 0.02) and 3 s (0.91 ± 0.02), while classification performance with 1 s were lower (0.87 ± 0.02). Tested sampling frequencies of 50 Hz, 25 Hz, and 13 Hz had similar results with most classified activity types, with an exception of outdoor cycling, where differences were significant. Using forward sequential feature selection enabled the decreasing of the number of features from initial 110 features to about 12 features without lowering the classification performance. The results of this study have been used for developing more efficient real-time physical activity classifiers.

Sign in / Sign up

Export Citation Format

Share Document