Joshan Athanesious | Anna University (original) (raw)
Papers by Joshan Athanesious
2022 1st International Conference on Computational Science and Technology (ICCST)
International Journal of Digital Multimedia Broadcasting
After cataract, glaucoma is one of the second leading retinal diseases in the world. This paper p... more After cataract, glaucoma is one of the second leading retinal diseases in the world. This paper presents the methodology to detect the glaucoma using principal component analysis. The images are involved in dilation as a preprocessing, enhancement using the contrast limited adaptive histogram equalization method, and followed by the extraction of features using principal component analysis. The extracted features are classified using support vector machine, Naive Bayes, and K-nearest neighbor classifiers. Comparing with other classifiers, the Naive Bayes provides high accuracy of 95% which demonstrates the effectiveness of the feature extraction and the classifier.
PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON RECENT ADVANCES IN MANUFACTURING ENGINEERING RESEARCH 2021: ICRAMER 2021
The aim is to develop an alternative means to Industrial Metrology, by integrating 3D Scanning an... more The aim is to develop an alternative means to Industrial Metrology, by integrating 3D Scanning and 3D Reconstruction. Traditional mean of modern metrology includes Manual Measurement, Measurement by CMM and Laser Bases Scanning. These means of "measurement" are often too expensive as is in the case of CMMs or too prone to human error. Thus, this resulted in the need for development of a new method for metrology for mass manufactured components. This need lead to the integration of 3D Scanning by means of a simple webcam and 3D Reconstruction of the object using Context capture. The camera is rotated, by means of a DC Motor, around the object that is fixed in its position. At regular intervals of rotation, takes pictures and stores them. A curved guideway is provided for the webcam mount so that pictures can be taken from different angles of orientation. The pictures are then fed into Context Capture, where they are stitched and result in a reconstructed model. For profile validation we use Compare Vidia, where we compare the results of the two scans, one of master or reference piece and one of the test or samples. After setting angular thresholds and tolerances, the two geometry profiles are compared and depending on the tolerances, the result is a "match" or "not-a-match".
2021 4th International Conference on Computing and Communications Technologies (ICCCT), 2021
Plant disease is a significant factor that affects plant growth and yield leading to a loss in mo... more Plant disease is a significant factor that affects plant growth and yield leading to a loss in money and time. A professional farmer can identify the disease type with his experience. Still, a new one has to rely on an agricultural expert, which takes more time and risks disease spreading to remaining crops. We have chosen the paddy crop among various plant types as it is the widely cultivated plant type in the South-Asian subcontinent. We have primarily determined three common disease types, namely Leaf blight, Brown spot, and Blast. Our solution is to identify the disease type from paddy leaf images using a Convolutional neural network(CNN). Our dataset has 1100 images, of which 700 are taken for training and 400 for testing. We use the TensorFlow Keras model for image pre-processing and classification, in which the image will be resized and converted to RGB. We operate transfer learning methods for classification as it takes less time compared to other algorithms. Finally, the disease type is displayed to the user.
2021 4th International Conference on Computing and Communications Technologies (ICCCT), 2021
Farming involves various activities such as cultivation, irrigation, harvesting, and more. In the... more Farming involves various activities such as cultivation, irrigation, harvesting, and more. In these activities, searching for weeds that affect the crops all over the field is a tedious process. Weeds are unwanted plants that grow along with crops. Many weeds look very similar to crops which makes it hard for farmers to categorize weeds among crops. We have various crops all over the globe in which we are going to take sesame as the main crop, and other unwanted plants that affect sesame are considered weeds. Our solution is to detect the weeds and crops using Region-Based Convolutional Neural Networks(RCNN). In our dataset, we have 1300 images of sesame crops and other crops (weeds). We will use, Tensorflow Keras model for image classification, in which we will have background, crop, and weeds as classes. We Train our model using RCNN to get fine-tuned images and SVM to improve the model's overall prediction.
Procedia Computer Science, 2019
A significant portion of the time allocated to a faculty for teaching purposes is consumed on the... more A significant portion of the time allocated to a faculty for teaching purposes is consumed on the task of taking attendance of the students. This is an issue because it takes the valuable time of teachers which could be spent on more productive tasks such as teaching and interacting with students. In excess to the increase in chaos and loss of decorum in the classroom environment, the presence of proxy attendance also plagues the existing method of manual attendance keeping. To counter these issues, this paper proposed the Deep Learning Assisted Attendance System (DPAAS); which keeps track of students attending a particular class with the help of a continuous stream of pictures captured from a video streaming device located inside a classroom connected to the remote server. The proposed DPAAS method reduces the amount of time spent by the faculty on taking attendance, and leads to a reduction in chaos inside a classroom. DPASS is proposed handles the issues in existing systems such as multi-class identification for multiple individuals in a classroom, occlusion and differing light scenarios. The DPAAS methodology compares the results of the state of art algorithms, and uses the best fit architecture which provides the lowest false rate on evaluation. There is no need of user interaction in the proposed DPAAS. Experimental results show that the proposed DPAAS method gives 94.66% accuracy which is better than the other existing methods.
IET Image Processing, 2020
Detection of abnormal events in the traffic scene is very challenging and is a significant proble... more Detection of abnormal events in the traffic scene is very challenging and is a significant problem in video surveillance. The authors proposed a novel scheme called super orientation optical flow (SOOF)-based clustering for identifying the abnormal activities. The key idea behind the proposed SOOF features is to efficiently reproduce the motion information of a moving vehicle with respect to superorientation motion descriptor within the sequence of the frame. Here, the authors adopt the mean absolute temporal difference to identify the anomalies by motion block (MB) selection and localisation. SOOF features obtained from MB are used as motion descriptor for both normal and abnormal events. Simple and efficient K-means clustering is used to study the normal motion flow during the training. The abnormal events are identified using the nearest-neighbour searching technique in the testing phase. The experimental outcome shows that the proposed work is effectively detecting anomalies and found to give results better than the state-of-the-art techniques.
Multimedia Tools and Applications, 2019
Detection of abnormal trajectories in a traffic scene is an important problem in Video Traffic Su... more Detection of abnormal trajectories in a traffic scene is an important problem in Video Traffic Surveillance (VTS). Recently, General Potential Data Field (GPDf)-based trajectory clustering scheme has been adopted for detecting abnormal events such as illegal U-turn, wrong side and unusual driving behaviors and it uses spatial and temporal attributes explicitly. The concept of data field is used to discover the relation between the spatial points in data-space and grouping them into clusters based on their mutual interaction. Existing methodologies related to potential data field-based clustering have certain limitations such as pre-defined cluster size, non-effective cluster center identification, and limitation in range estimation using isotropic impact factor (h) which leads to inaccurate results. In order to address the above-mentioned issues, this paper proposes an efficient anomaly detection scheme based on General Potential Data field with Spectral Clustering (GPDfSC). The proposed GPDfSC scheme utilizes potential data field technique along with spectral clustering for effective identification of abnormalities. The Limitation in impact factor(h) is overcome by using anisotropic impact parameter Bmat. Further, Bayesian Decision theory is used to classify the events as normal or abnormal. The proposed scheme is implemented in real time using GPU and from the results it is found that it gives 12% better accuracy in detecting abnormalities than the state of art technique. Keywords General potential data field. Impact factor matrix. Spatial trajectory data. Dynamic time warping. Spectral clustering. Abnormal detection Multimedia Tools and Applications
2015 Seventh International Conference on Advanced Computing (ICoAC), 2015
Detecting anomalies such as rule violations, accidents, unusual driving and other suspicious acti... more Detecting anomalies such as rule violations, accidents, unusual driving and other suspicious action increase the need for automatic analysis in Traffic Video Surveillance (TVS). Most of the works in Traffic rule violation systems are based on probabilistic methods of classification for detecting the events as normal and abnormal. This paper proposes an un-supervised clustering technique namely Novel Anomaly Detection-Density Based Spatial Clustering of Applications with Noise (NAD-DBSCAN) which clusters the trajectories of moving objects of varying sizes and shapes. A trajectory is said to be abnormal if the event that never fit with the trained model. Epsilon (Eps) and Minimum Points (MinPts) are essential parameters for dynamically calculating the sum of clusters for a data point. The proposed system is validated using benchmark traffic dataset and found to perform accurately in detecting anomalies.
Procedia Computer Science, 2015
In this paper, a novel approach for automatic segmentation and classification of skin lesions is ... more In this paper, a novel approach for automatic segmentation and classification of skin lesions is proposed. Initially, skin images are filtered to remove unwanted hairs and noise and then the segmentation process is carried out to extract lesion areas. For segmentation, a region growing method is applied by automatic initialization of seed points. The segmentation performance is measured with different well known measures and the results are appreciable. Subsequently, the extracted lesion areas are represented by color and texture features. SVM and k-NN classifiers are used along with their fusion for the classification using the extracted features. The performance of the system is tested on our own dataset of 726 samples from 141 images consisting of 5 different classes of diseases. The results are very promising with 46.71% and 34% of F-measure using SVM and k-NN classifier respectively and with 61% of F-measure for fusion of SVM and k-NN.
2015 Seventh International Conference on Advanced Computing (ICoAC), 2015
Detection of vehicles in a pedestrian pathway is important for the safety of pedestrians. Existin... more Detection of vehicles in a pedestrian pathway is important for the safety of pedestrians. Existing methods examine the entire frame for feature extraction and detect vehicles. However in real-time surveillance, analyzing the entire frame is computationally complex. To reduce the computational complexity, this paper proposes an efficient vehicle detection scheme. Using defined range approach, the background is subtracted and foreground blobs obtained from the background subtraction, the probability of the blobs in which the presence of the vehicles is expected is identified by a statistically computed predefined area criterian. These identified blobs within the defined range are taken as the Region of Interest (ROI). The Haar features are extracted from these ROI and then provided to Cascaded Classifier, which is trained apriori to detect vehicles. The proposed scheme is implemented using OpenCV, tested in real-time and found to give faster and better detection rate compared to existing method.
2019 International Conference on Wireless Communications Signal Processing and Networking (WiSPNET), 2019
A new approach is proposed to analysis and localize abnormal events in crowded scenes, such us de... more A new approach is proposed to analysis and localize abnormal events in crowded scenes, such us detecting suspicious behavior, sudden assemble and dispersion of pedestrian in the crowd scenes. The proposed methodology uses streak flow technique along with LDA (Latent Dirichlet Allocation)to detect anomalies in crowd scenes. This method divides the frames into blocks for accurate representation of spatial and temporal changes in scene. Optical flow algorithm is used to estimate the motion and direction of the pedestrian in crowd scene. Crowd flow motion is extracted using streak flow with the combination of streak line and potential functions. Experimental result shows that the detection of abnormalities in crowd scenes using streak flow approach gives better accuracy of 93.37% than the state of Art of techniques.
2019 International Conference on Wireless Communications Signal Processing and Networking (WiSPNET), 2019
In today’s digital age, detecting abnormal events in the traffic scene is a more significant prob... more In today’s digital age, detecting abnormal events in the traffic scene is a more significant problem in video surveillance. The proposed scheme is adopted for detecting abnormalities such as wrong side driving, people crossing the road illegally, vehicle moving in pedestrain pathway and which uses spatio-temporal attributes. Although numerous machine learning approaches are described in existing literature focusing on pixel-wise differnce with less performance in identifying the abnormalities in video traffic scenes. In this paper, a novel frame work for identifying abnormal detection using unsupervised deep learning algorithm. The scheme uses joint based methodology (ConvLSTM with kmeans) with a combination of reconstruction and clustering loss respectively. The convolutional neural network with LSTM is used to detect and recognize anomalies in traffic scenes with refernce of previous frame information. To identify the normal or abnormal events, the proposed work is composed of tra...
Advances in Computerized Analysis in Clinical and Medical Imaging, 2019
1298 www.ijarcet.org Abstract Object tracking in video sequences is one of the important ongoing ... more 1298 www.ijarcet.org Abstract Object tracking in video sequences is one of the important ongoing exploration areas in the field of computer vision. Computer vision is an arena that comprises methods for acquiring, processing, analyzing images and also covers the essential technology of automatic image analysis which is used in various fields. The aim of object tracking is to find the trajectory of the target objects through a number of frames from an image sequence. Object Tracking is identification of interesting object, especially on tracking of walkers or moving vehicles. Tracking is an interesting problem owing to, object occlusion, varying of illumination, unexpected object motion and camera motion. Normally many algorithms were developed for successful tracking. Object Tracking is mainly classified of three stages: object extraction, object recognition and tracking, and decisions about activities. In this paper we have implemented some algorithms and comparison table are analy...
Object tracking in video structures is one of the important ongoing exploration areas in the fiel... more Object tracking in video structures is one of the important ongoing exploration areas in the field of computer vision. The aim of object tracking is to find an object of a pre-defined class in a video frame.Video structures consists of multiple frames and huge facts, hence video tracking is a time overriding procedures. Tracking is nothing but identification of interest, especially on tracking of moving vehicles or walkers. In event of any strange actions, an attentive should be provided. Normally a video tracking system combines three stages of data treating: object extraction, object recognition and tracking, and decisions about activities. In this paper, we analysis various object tracking techniques.
Int. J. Commun. Syst., 2021
2018 Tenth International Conference on Advanced Computing (ICoAC)
Journal of Intelligent & Fuzzy Systems
Detection of abnormal events in a traffic scene is a highly challenging task due to vast field of... more Detection of abnormal events in a traffic scene is a highly challenging task due to vast field of view, continuous stream of video data, various object interactions and complex events in Video Surveillance. Hence, this research proposes novel schemes using machine learning approach to detect abnormal events such as illegal U-turn, presence of pedestrian in driving region, wrong side driving and frequent lane change. Recently, Density Based Spatial Clustering of Applications with Noise (DBSCAN) is a popular method that has been used for clustering the trajectory datasets. The existing Density Based Clustering approach used for Abnormal detection in traffic scene uses random selection of cluster radius (Eps) and minimum points (minpts) needed to form a cluster. This random selection is time consuming and inefficient clustering results in accuracy reduction in abnormal detection. So, Adaptive Density based Spatial Clustering of Applications with Noise (ADBSCAN) is proposed for the dete...
2022 1st International Conference on Computational Science and Technology (ICCST)
International Journal of Digital Multimedia Broadcasting
After cataract, glaucoma is one of the second leading retinal diseases in the world. This paper p... more After cataract, glaucoma is one of the second leading retinal diseases in the world. This paper presents the methodology to detect the glaucoma using principal component analysis. The images are involved in dilation as a preprocessing, enhancement using the contrast limited adaptive histogram equalization method, and followed by the extraction of features using principal component analysis. The extracted features are classified using support vector machine, Naive Bayes, and K-nearest neighbor classifiers. Comparing with other classifiers, the Naive Bayes provides high accuracy of 95% which demonstrates the effectiveness of the feature extraction and the classifier.
PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON RECENT ADVANCES IN MANUFACTURING ENGINEERING RESEARCH 2021: ICRAMER 2021
The aim is to develop an alternative means to Industrial Metrology, by integrating 3D Scanning an... more The aim is to develop an alternative means to Industrial Metrology, by integrating 3D Scanning and 3D Reconstruction. Traditional mean of modern metrology includes Manual Measurement, Measurement by CMM and Laser Bases Scanning. These means of "measurement" are often too expensive as is in the case of CMMs or too prone to human error. Thus, this resulted in the need for development of a new method for metrology for mass manufactured components. This need lead to the integration of 3D Scanning by means of a simple webcam and 3D Reconstruction of the object using Context capture. The camera is rotated, by means of a DC Motor, around the object that is fixed in its position. At regular intervals of rotation, takes pictures and stores them. A curved guideway is provided for the webcam mount so that pictures can be taken from different angles of orientation. The pictures are then fed into Context Capture, where they are stitched and result in a reconstructed model. For profile validation we use Compare Vidia, where we compare the results of the two scans, one of master or reference piece and one of the test or samples. After setting angular thresholds and tolerances, the two geometry profiles are compared and depending on the tolerances, the result is a "match" or "not-a-match".
2021 4th International Conference on Computing and Communications Technologies (ICCCT), 2021
Plant disease is a significant factor that affects plant growth and yield leading to a loss in mo... more Plant disease is a significant factor that affects plant growth and yield leading to a loss in money and time. A professional farmer can identify the disease type with his experience. Still, a new one has to rely on an agricultural expert, which takes more time and risks disease spreading to remaining crops. We have chosen the paddy crop among various plant types as it is the widely cultivated plant type in the South-Asian subcontinent. We have primarily determined three common disease types, namely Leaf blight, Brown spot, and Blast. Our solution is to identify the disease type from paddy leaf images using a Convolutional neural network(CNN). Our dataset has 1100 images, of which 700 are taken for training and 400 for testing. We use the TensorFlow Keras model for image pre-processing and classification, in which the image will be resized and converted to RGB. We operate transfer learning methods for classification as it takes less time compared to other algorithms. Finally, the disease type is displayed to the user.
2021 4th International Conference on Computing and Communications Technologies (ICCCT), 2021
Farming involves various activities such as cultivation, irrigation, harvesting, and more. In the... more Farming involves various activities such as cultivation, irrigation, harvesting, and more. In these activities, searching for weeds that affect the crops all over the field is a tedious process. Weeds are unwanted plants that grow along with crops. Many weeds look very similar to crops which makes it hard for farmers to categorize weeds among crops. We have various crops all over the globe in which we are going to take sesame as the main crop, and other unwanted plants that affect sesame are considered weeds. Our solution is to detect the weeds and crops using Region-Based Convolutional Neural Networks(RCNN). In our dataset, we have 1300 images of sesame crops and other crops (weeds). We will use, Tensorflow Keras model for image classification, in which we will have background, crop, and weeds as classes. We Train our model using RCNN to get fine-tuned images and SVM to improve the model's overall prediction.
Procedia Computer Science, 2019
A significant portion of the time allocated to a faculty for teaching purposes is consumed on the... more A significant portion of the time allocated to a faculty for teaching purposes is consumed on the task of taking attendance of the students. This is an issue because it takes the valuable time of teachers which could be spent on more productive tasks such as teaching and interacting with students. In excess to the increase in chaos and loss of decorum in the classroom environment, the presence of proxy attendance also plagues the existing method of manual attendance keeping. To counter these issues, this paper proposed the Deep Learning Assisted Attendance System (DPAAS); which keeps track of students attending a particular class with the help of a continuous stream of pictures captured from a video streaming device located inside a classroom connected to the remote server. The proposed DPAAS method reduces the amount of time spent by the faculty on taking attendance, and leads to a reduction in chaos inside a classroom. DPASS is proposed handles the issues in existing systems such as multi-class identification for multiple individuals in a classroom, occlusion and differing light scenarios. The DPAAS methodology compares the results of the state of art algorithms, and uses the best fit architecture which provides the lowest false rate on evaluation. There is no need of user interaction in the proposed DPAAS. Experimental results show that the proposed DPAAS method gives 94.66% accuracy which is better than the other existing methods.
IET Image Processing, 2020
Detection of abnormal events in the traffic scene is very challenging and is a significant proble... more Detection of abnormal events in the traffic scene is very challenging and is a significant problem in video surveillance. The authors proposed a novel scheme called super orientation optical flow (SOOF)-based clustering for identifying the abnormal activities. The key idea behind the proposed SOOF features is to efficiently reproduce the motion information of a moving vehicle with respect to superorientation motion descriptor within the sequence of the frame. Here, the authors adopt the mean absolute temporal difference to identify the anomalies by motion block (MB) selection and localisation. SOOF features obtained from MB are used as motion descriptor for both normal and abnormal events. Simple and efficient K-means clustering is used to study the normal motion flow during the training. The abnormal events are identified using the nearest-neighbour searching technique in the testing phase. The experimental outcome shows that the proposed work is effectively detecting anomalies and found to give results better than the state-of-the-art techniques.
Multimedia Tools and Applications, 2019
Detection of abnormal trajectories in a traffic scene is an important problem in Video Traffic Su... more Detection of abnormal trajectories in a traffic scene is an important problem in Video Traffic Surveillance (VTS). Recently, General Potential Data Field (GPDf)-based trajectory clustering scheme has been adopted for detecting abnormal events such as illegal U-turn, wrong side and unusual driving behaviors and it uses spatial and temporal attributes explicitly. The concept of data field is used to discover the relation between the spatial points in data-space and grouping them into clusters based on their mutual interaction. Existing methodologies related to potential data field-based clustering have certain limitations such as pre-defined cluster size, non-effective cluster center identification, and limitation in range estimation using isotropic impact factor (h) which leads to inaccurate results. In order to address the above-mentioned issues, this paper proposes an efficient anomaly detection scheme based on General Potential Data field with Spectral Clustering (GPDfSC). The proposed GPDfSC scheme utilizes potential data field technique along with spectral clustering for effective identification of abnormalities. The Limitation in impact factor(h) is overcome by using anisotropic impact parameter Bmat. Further, Bayesian Decision theory is used to classify the events as normal or abnormal. The proposed scheme is implemented in real time using GPU and from the results it is found that it gives 12% better accuracy in detecting abnormalities than the state of art technique. Keywords General potential data field. Impact factor matrix. Spatial trajectory data. Dynamic time warping. Spectral clustering. Abnormal detection Multimedia Tools and Applications
2015 Seventh International Conference on Advanced Computing (ICoAC), 2015
Detecting anomalies such as rule violations, accidents, unusual driving and other suspicious acti... more Detecting anomalies such as rule violations, accidents, unusual driving and other suspicious action increase the need for automatic analysis in Traffic Video Surveillance (TVS). Most of the works in Traffic rule violation systems are based on probabilistic methods of classification for detecting the events as normal and abnormal. This paper proposes an un-supervised clustering technique namely Novel Anomaly Detection-Density Based Spatial Clustering of Applications with Noise (NAD-DBSCAN) which clusters the trajectories of moving objects of varying sizes and shapes. A trajectory is said to be abnormal if the event that never fit with the trained model. Epsilon (Eps) and Minimum Points (MinPts) are essential parameters for dynamically calculating the sum of clusters for a data point. The proposed system is validated using benchmark traffic dataset and found to perform accurately in detecting anomalies.
Procedia Computer Science, 2015
In this paper, a novel approach for automatic segmentation and classification of skin lesions is ... more In this paper, a novel approach for automatic segmentation and classification of skin lesions is proposed. Initially, skin images are filtered to remove unwanted hairs and noise and then the segmentation process is carried out to extract lesion areas. For segmentation, a region growing method is applied by automatic initialization of seed points. The segmentation performance is measured with different well known measures and the results are appreciable. Subsequently, the extracted lesion areas are represented by color and texture features. SVM and k-NN classifiers are used along with their fusion for the classification using the extracted features. The performance of the system is tested on our own dataset of 726 samples from 141 images consisting of 5 different classes of diseases. The results are very promising with 46.71% and 34% of F-measure using SVM and k-NN classifier respectively and with 61% of F-measure for fusion of SVM and k-NN.
2015 Seventh International Conference on Advanced Computing (ICoAC), 2015
Detection of vehicles in a pedestrian pathway is important for the safety of pedestrians. Existin... more Detection of vehicles in a pedestrian pathway is important for the safety of pedestrians. Existing methods examine the entire frame for feature extraction and detect vehicles. However in real-time surveillance, analyzing the entire frame is computationally complex. To reduce the computational complexity, this paper proposes an efficient vehicle detection scheme. Using defined range approach, the background is subtracted and foreground blobs obtained from the background subtraction, the probability of the blobs in which the presence of the vehicles is expected is identified by a statistically computed predefined area criterian. These identified blobs within the defined range are taken as the Region of Interest (ROI). The Haar features are extracted from these ROI and then provided to Cascaded Classifier, which is trained apriori to detect vehicles. The proposed scheme is implemented using OpenCV, tested in real-time and found to give faster and better detection rate compared to existing method.
2019 International Conference on Wireless Communications Signal Processing and Networking (WiSPNET), 2019
A new approach is proposed to analysis and localize abnormal events in crowded scenes, such us de... more A new approach is proposed to analysis and localize abnormal events in crowded scenes, such us detecting suspicious behavior, sudden assemble and dispersion of pedestrian in the crowd scenes. The proposed methodology uses streak flow technique along with LDA (Latent Dirichlet Allocation)to detect anomalies in crowd scenes. This method divides the frames into blocks for accurate representation of spatial and temporal changes in scene. Optical flow algorithm is used to estimate the motion and direction of the pedestrian in crowd scene. Crowd flow motion is extracted using streak flow with the combination of streak line and potential functions. Experimental result shows that the detection of abnormalities in crowd scenes using streak flow approach gives better accuracy of 93.37% than the state of Art of techniques.
2019 International Conference on Wireless Communications Signal Processing and Networking (WiSPNET), 2019
In today’s digital age, detecting abnormal events in the traffic scene is a more significant prob... more In today’s digital age, detecting abnormal events in the traffic scene is a more significant problem in video surveillance. The proposed scheme is adopted for detecting abnormalities such as wrong side driving, people crossing the road illegally, vehicle moving in pedestrain pathway and which uses spatio-temporal attributes. Although numerous machine learning approaches are described in existing literature focusing on pixel-wise differnce with less performance in identifying the abnormalities in video traffic scenes. In this paper, a novel frame work for identifying abnormal detection using unsupervised deep learning algorithm. The scheme uses joint based methodology (ConvLSTM with kmeans) with a combination of reconstruction and clustering loss respectively. The convolutional neural network with LSTM is used to detect and recognize anomalies in traffic scenes with refernce of previous frame information. To identify the normal or abnormal events, the proposed work is composed of tra...
Advances in Computerized Analysis in Clinical and Medical Imaging, 2019
1298 www.ijarcet.org Abstract Object tracking in video sequences is one of the important ongoing ... more 1298 www.ijarcet.org Abstract Object tracking in video sequences is one of the important ongoing exploration areas in the field of computer vision. Computer vision is an arena that comprises methods for acquiring, processing, analyzing images and also covers the essential technology of automatic image analysis which is used in various fields. The aim of object tracking is to find the trajectory of the target objects through a number of frames from an image sequence. Object Tracking is identification of interesting object, especially on tracking of walkers or moving vehicles. Tracking is an interesting problem owing to, object occlusion, varying of illumination, unexpected object motion and camera motion. Normally many algorithms were developed for successful tracking. Object Tracking is mainly classified of three stages: object extraction, object recognition and tracking, and decisions about activities. In this paper we have implemented some algorithms and comparison table are analy...
Object tracking in video structures is one of the important ongoing exploration areas in the fiel... more Object tracking in video structures is one of the important ongoing exploration areas in the field of computer vision. The aim of object tracking is to find an object of a pre-defined class in a video frame.Video structures consists of multiple frames and huge facts, hence video tracking is a time overriding procedures. Tracking is nothing but identification of interest, especially on tracking of moving vehicles or walkers. In event of any strange actions, an attentive should be provided. Normally a video tracking system combines three stages of data treating: object extraction, object recognition and tracking, and decisions about activities. In this paper, we analysis various object tracking techniques.
Int. J. Commun. Syst., 2021
2018 Tenth International Conference on Advanced Computing (ICoAC)
Journal of Intelligent & Fuzzy Systems
Detection of abnormal events in a traffic scene is a highly challenging task due to vast field of... more Detection of abnormal events in a traffic scene is a highly challenging task due to vast field of view, continuous stream of video data, various object interactions and complex events in Video Surveillance. Hence, this research proposes novel schemes using machine learning approach to detect abnormal events such as illegal U-turn, presence of pedestrian in driving region, wrong side driving and frequent lane change. Recently, Density Based Spatial Clustering of Applications with Noise (DBSCAN) is a popular method that has been used for clustering the trajectory datasets. The existing Density Based Clustering approach used for Abnormal detection in traffic scene uses random selection of cluster radius (Eps) and minimum points (minpts) needed to form a cluster. This random selection is time consuming and inefficient clustering results in accuracy reduction in abnormal detection. So, Adaptive Density based Spatial Clustering of Applications with Noise (ADBSCAN) is proposed for the dete...