A F M Saifuddin Saif | University of Helsinki (original) (raw)
Papers by A F M Saifuddin Saif
Journal of Engineering and Science Research, 2021
Crowd flow estimation from Drones or normally referred as Unmanned Aerial Vehicle (UAV ) for crow... more Crowd flow estimation from Drones or normally referred as Unmanned Aerial Vehicle (UAV ) for crowd management and monitoring is an essential research problem for adaptive monitoring and controlling dynamic crowd gatherings. Various challenges exist in this context, i.e. variation in density, scale, brightness, height from UAV platform, occlusion and inefficient pose estimation. Currently, gathering of crowd is mostly monitored by Close Circuit Television (CCTV) cameras where various problems exist, i.e. coverage in little area and constant involvement of human to monitor crowd which encourage researchers to move towards deep learning and computer vision techniques to minimize the need of human operator and thus develop intelligent crowd counting techniques. Deep learning frameworks are promising for intelligent crowd analysis from frames of video despite the fact of various challenges for detecting humans from unstable UAV camera platforms. This research presents rigorous investigat...
Fast and computationally less complex feature extraction for moving object detection using aerial... more Fast and computationally less complex feature extraction for moving object detection using aerial images from unmanned aerial vehicles (UAVs) remains as an elusive goal in the field of computer vision research. The types of features used in current studies concerningmov-ing object detection are typically chosen based on improving detection rate rather than on providing fast and computationally less complex feature extraction methods. Because mov-ing object detection using aerial images from UAVs involves motion as seen from a certain altitude, effective and fast feature extraction is a vital issue for optimum detection perfor-mance. This research proposes a two-layer bucket approach based on a new feature ex-traction algorithm referred to as the moment-based feature extraction algorithm (MFEA). Because a moment represents thecoherent intensity of pixels and motion estimation is a motion pixel intensity measurement, this research used this relation to develop the pro-posed algorithm....
Proceedings of the International Conference on Computing Advancements, 2020
A toddler learns to hold various types of objects using hands in their everyday life without any ... more A toddler learns to hold various types of objects using hands in their everyday life without any specific instruction from an adult. The main media of interaction happens using the hands. A toddler's sensory-motor is developed by manipulating various types of objects. Exploring object holding-activity of a toddler can provide new insight into how sensory-motor develops in the formative years. Goal of this research is to learn how a toddler develops the sensory-motor system from directly hand interaction of an object. A video dataset was developed to study the patterns and hand-pose of a toddler while using their hands to hold an object. The birds-eye view of the scene was captured using a top-mounted camera, where the toddler was interacting with various types of objects with hands. Parents were guided to engage their child to interact with the object. The different hand-pose of the toddler's hands were observed, while the toddler was manipulating different types of objects. A detailed analysis result has presented on the hand-pose distribution of a toddler's hands for object holding-activities.
Journal of Information Technology Research, 2021
Recent years have seen a rise in the use of various machine learning techniques in computer visio... more Recent years have seen a rise in the use of various machine learning techniques in computer vision, particularly in posing feature-based human action recognition which includes convolutional neural networks (CNN) and recurrent neural network (RNN). CNN-based methods are useful in recognizing human actions for combined motions (i.e., standing up, hand shaking, walking). However, in case of uncertainty of camera motion, occlusion, and multiple people, CNN suppresses important feature information and is not efficient enough to recognize variations for human action. Besides, RNN with long short-term memory (LSTM) requires more computational power to retain memories to classify human actions. This research proposes an extended framework based on capsule network using silhouette pose features to recognize human actions. Proposed extended framework achieved high accuracy of 95.64% which is higher than previous research methodology. Extensive experimental validation of the proposed extended...
Moving object detection from UAV aerial images has stated to get attention in recent years where ... more Moving object detection from UAV aerial images has stated to get attention in recent years where fast abrupt motion, low resolution, noisy imagery, cluttered background, low contrast and small target size need to be compensated with an updated moving object detection framework to increase detection performance. This paper presents 6 UCF (Uncertainty Constraint Factors) issues which are the recent categorized challenges to handle performance metric for moving object detection from UAV aerial images. A critical review for adaptive motion analysis to handle these 6 UCF factors is presented in this paper. Besides these 6 UCF factors, feature extraction for moving object detection is another unsolved issue. Large feature vector is suitable for optimum detection performance but increase computation time for overall detection process. This paper presents critical review of existing feature extraction techniques which can be compensated with adaptive motion analysis framework to influence o...
pertanika journal of science and technology, 2017
Face detection and analysis is an important area in computer vision. Furthermore, face detection ... more Face detection and analysis is an important area in computer vision. Furthermore, face detection has been an active research field in the recent years following the advancement in digital image processing. The visualisation of visual entities or sub-pattern composition may become complex to visualise due to the high frequency of noise and light effect during examination. This study focuses on evaluating the ability of Haar classifier in detecting faces from three paired Min-Max values used on histogram stretching. Min-Max histogram stretching was the selected method for implementation given that it appears to be the appropriate technique from the observation carried out. Experimental results show that, 60-240 MinMax values, Haar classifier can accurately detect faces compared to the two values.
International Journal of Image, Graphics and Signal Processing, 2021
Advances in Visual Informatics, 2019
Efficient feature extraction for moving object using aerial images is still an unsolved issue in ... more Efficient feature extraction for moving object using aerial images is still an unsolved issue in computer vision, image processing and pattern recognition research domains. Aerial types of images contain various environmental constraints due to capture frames from various altitudes level, i.e. illumination, shadows, occlusion. For this reason, appropriate feature selection for those types of images needs more attention by the researchers to improve detection rate with fast and computationally less complex features extraction method. This research performed comprehensive review with critical analysis for using various features with various methods for moving object detection using aerial images. In this context, three aspects for critical analysis of using various features are identified followed by challenges of using various features. After that, existing methods with advantages and barriers are comprehensively described with various constraints claimed by the previous research. Next, justification for the need of new feature selection is elaborated for optimum detection performance. Later, adequate validation matrics are illustrated to evaluate various features based moving object detection using aerial images performed in the previous research. The overall review performed in this paper have been comprehensively studied and expected to contribute significantly in computer vision, image processing pattern recognition research field.
Proceedings of the International Conference on Computing Advancements, 2020
In Bangladesh, a central exam is given by lots of students each year to get themselves admitted i... more In Bangladesh, a central exam is given by lots of students each year to get themselves admitted into their preferred medical college and only few get chosen by their merits. Assigning each student to a college and checking the validity of the assignment is a major concern here. This particular problem dates back to 1962 when college admission problem was introduced by Gale and Shapley [5] where stable marriage approach was used as a solution. Stability criteria was also introduced in their research which can be used to check validation of fair matching when priority is involved. But, Complicacy arises as in Bangladesh some seats are reserved for special students known as quota which can not be solved with plain stable marriage algorithm. Stability criteria also has to be defined properly for this specific problem and an algorithm is needed to handle this. A max-flow, min-cut solution combined with stable marriage algorithm is presented in this research to solve medical admission pro...
Proceedings of the International Conference on Computing Advancements, 2020
Every year thousands of people lose their lives due to traffic accidents. Road accidents, especia... more Every year thousands of people lose their lives due to traffic accidents. Road accidents, especially the ones on the highways are the most fatal ones. Accidents not only cut peoples' lives short but also cause intense financial loss to the country. Many people who survive disastrous accidents are often left with critical injury or paralysis. Particularly in Bangladesh majority of the drivers are minimally educated. They do not have sufficient knowledge and tend to ignore the traffic rules often. As a result the roads are filled with careless drivers. Consequently most of the accidents on the highways and city roads occur due to the lack of awareness of the drivers. Additionally, many of the roads are poorly lit which makes it difficult to drive in unfavorable weather. An autonomous lane detection system can play an important role as a solution to the problem by assisting the driver in seeing the lane clearly. It can also generate warning to the driver in case of an unintentional...
Education gives the insight to take us to the future which has not been possible to grasp due to ... more Education gives the insight to take us to the future which has not been possible to grasp due to lack of research methodologies towards learning process. Traditional formal education methodologies feel like living in the old age as existing learning processes are not advanced like technology which starts from primary education. As a result, children are losing interest in education due to the way of learning. In this context, learning is a continuous process where the quality cannot be justified by just including reading and writing, the overall process should be creative and interactive as well. However, existing primary education system seems to abandon the main idea about education and learning process to lean towards memorization and rote learning. The way of acquisition of education should be through discussion, research, storytelling, training which should be fun and interesting but our traditional primary education is based on two-dimensional materials and does not provide an...
Edge features or characteristics are the significant local intensity variation of pictorial struc... more Edge features or characteristics are the significant local intensity variation of pictorial structure in an image or frame or video sequence. Edge detection process finds the existence and position of pixels causes by significant differences in intensity of the images or frames or video sequences. However, selection of appropriate edge features specially to detect moving objects using aerial types of video sequences is still an elusive research concern because of various constraints for aerial frames, i.e. various altitudes, lack of features available for detection, motion variation etc. This research proposes comparative evaluation of different edge features based moving object detection methods to make it easy to decide which edge features based detection method is appropriate for extraction of moving object from aerial frames. Proposed research selected three edge features for detection, i.e. Sobel, Prewitt and Canny for moving object extraction to process high and low pictorial ...
AIUB Journal of Science and Engineering (AJSE), 2020
Detection of vacant parking space is becoming a challenging task gradually. Space utilization and... more Detection of vacant parking space is becoming a challenging task gradually. Space utilization and management of vehicle space is now a demandable field of research. Searching for an empty parking space in congested traffic is a time-consuming process. The existing vacant parking space detection methods are not robust or generalized for images captured from different camera viewpoints. Finding a proper parking space in a busy city is really a challenging issue and people are facing this problem on a daily basis. The main purpose of this research is to comprehensively discuss the previous researches of vacant parking space detection and compare them from different aspects. Methods used in previous researches are descriptively discussed along with their advantages and disadvantages. The frameworks of previous researches were compared on six generalized phases and the experimental results are compared in terms of dataset, accuracy, processing time and other performance measures. This r...
Soft Computing Approach for Mathematical Modeling of Engineering Problems, 2021
The Fourth industrial revolution (IR 4.0) saw the emergence of computer vision and artificial int... more The Fourth industrial revolution (IR 4.0) saw the emergence of computer vision and artificial intelligence in creating smart imaging systems that can replace human vision and decision making especially to predict models for autonomous vehicles. In this context, advanced prediction of probable collision in real time scenario is an unsolved problem especially in the use of artificial intelligence and computer vision for autonomous vehicles. This research proposed an efficient collision avoidance model to avoid collision in real time scenario. Proposed model differs from other methods in a way that it does not require any other equipment like sensors for measuring distance between the vehicles. Proposed collision avoidance model estimates the relation between distance and size of the vehicle in real time scenario to generate an approximate notion of distance between the vehicles. Then, the ratio of distance between vehicles and size of the vehicle was used to depict vehicles that are in potentially dangerous positions for probable collision. Proposed collision avoidance model was experimented in the real-time traffic and experimental results showed that the model could detect vehicles in order to avoid the probable collisions efficiently. Proposed model is expected to be a possible tool in dealing with future demand of autonomous vehicles with the increase of 4IR technologies.
International Journal of Advanced Computer Science and Applications, 2020
International Journal of Information Systems and Engineering, 2013
Object detection is a fundamental step for automated video analysis in many highlevel UAV surveil... more Object detection is a fundamental step for automated video analysis in many highlevel UAV surveillance tasks, including cooperative UAV path planning, navigation control, and automated information analysis. Moving object detection in aerial video is still a challenging problem for the reason that when capturing the video the camera (or the platform) is moving all the time. As a result, the problem is detecting moving object from moving background which is much more difficult than the case that the background is constant. Moving object detection in stationary scene usually modelling the pixel value changes over time, but in aerial video the change does not have regular patterns. Therefore, we model the motion of the background rather than modelling the background directly.In this paper we present a low complexity long term motion analysis based moving object detection approach by using the ideas of robust analysis and spatiotemporal clustering. Here we present an efficient algorithm to estimate the global vehicle-camera motion. Our experimental results show the efficiency of the proposed algorithm.
International Journal of Engineering Trends and Technology, 2017
Determining a Pupil size, diameter and center are fundamental for pupil orientation. It is import... more Determining a Pupil size, diameter and center are fundamental for pupil orientation. It is important features to detect eye and It is considered as a significant verification method for human computer interaction. In This paper, We investigated existing methods and presented the framework to detect pupil to calculate its distance. The existing methods of pupil orientation have classified in 4 section which are estimation and measurement, localization, detection, and tracking. We have shown the tabular study of an algorithm, detected feature and accuracy for each classification sector. There are several investigations that are running to classified all the sectors accurately. We have also proposed a framework to calculate pupil distance from images. We have described the algorithms to detect and straighten face, detect eyes. We have also proposed that a modified Circle Equation can be better to detect and exact pupils based on circle size, object polarity, and sensitivity. Although, we have discussed distance calculation.
International Journal of Image, Graphics and Signal Processing, 2019
High returning rate of garments products have become a notable problem for online fashion shoppin... more High returning rate of garments products have become a notable problem for online fashion shopping. This problem is partially caused by using different standards for measuring cloth sizes on different websites. In this research, we have designed a set of equipment to capture images of t-shirts of any color and propose an automatic cloth measurement approach using image processing techniques. A method has been introduced to recognize feature points, which has been used to calculate the cloth sizes. The method has provided a useful and efficient tool for cloth measurement. The photographs have been taken in a controlled environment, and then clothes have been categorized with the proportions of the neck, shoulder, chest width, upper waist, lower waist, and length. In this method, we have measured the t-shirt size for men by calculating the chest width and length of men. For this, a dataset has been created in a specific environment. This method has integrated with a webbased application. We have validated our work by calculating RMSE values.
International Journal of Education and Management Engineering, 2019
International Journal of Education and Management Engineering, 2019
Journal of Engineering and Science Research, 2021
Crowd flow estimation from Drones or normally referred as Unmanned Aerial Vehicle (UAV ) for crow... more Crowd flow estimation from Drones or normally referred as Unmanned Aerial Vehicle (UAV ) for crowd management and monitoring is an essential research problem for adaptive monitoring and controlling dynamic crowd gatherings. Various challenges exist in this context, i.e. variation in density, scale, brightness, height from UAV platform, occlusion and inefficient pose estimation. Currently, gathering of crowd is mostly monitored by Close Circuit Television (CCTV) cameras where various problems exist, i.e. coverage in little area and constant involvement of human to monitor crowd which encourage researchers to move towards deep learning and computer vision techniques to minimize the need of human operator and thus develop intelligent crowd counting techniques. Deep learning frameworks are promising for intelligent crowd analysis from frames of video despite the fact of various challenges for detecting humans from unstable UAV camera platforms. This research presents rigorous investigat...
Fast and computationally less complex feature extraction for moving object detection using aerial... more Fast and computationally less complex feature extraction for moving object detection using aerial images from unmanned aerial vehicles (UAVs) remains as an elusive goal in the field of computer vision research. The types of features used in current studies concerningmov-ing object detection are typically chosen based on improving detection rate rather than on providing fast and computationally less complex feature extraction methods. Because mov-ing object detection using aerial images from UAVs involves motion as seen from a certain altitude, effective and fast feature extraction is a vital issue for optimum detection perfor-mance. This research proposes a two-layer bucket approach based on a new feature ex-traction algorithm referred to as the moment-based feature extraction algorithm (MFEA). Because a moment represents thecoherent intensity of pixels and motion estimation is a motion pixel intensity measurement, this research used this relation to develop the pro-posed algorithm....
Proceedings of the International Conference on Computing Advancements, 2020
A toddler learns to hold various types of objects using hands in their everyday life without any ... more A toddler learns to hold various types of objects using hands in their everyday life without any specific instruction from an adult. The main media of interaction happens using the hands. A toddler's sensory-motor is developed by manipulating various types of objects. Exploring object holding-activity of a toddler can provide new insight into how sensory-motor develops in the formative years. Goal of this research is to learn how a toddler develops the sensory-motor system from directly hand interaction of an object. A video dataset was developed to study the patterns and hand-pose of a toddler while using their hands to hold an object. The birds-eye view of the scene was captured using a top-mounted camera, where the toddler was interacting with various types of objects with hands. Parents were guided to engage their child to interact with the object. The different hand-pose of the toddler's hands were observed, while the toddler was manipulating different types of objects. A detailed analysis result has presented on the hand-pose distribution of a toddler's hands for object holding-activities.
Journal of Information Technology Research, 2021
Recent years have seen a rise in the use of various machine learning techniques in computer visio... more Recent years have seen a rise in the use of various machine learning techniques in computer vision, particularly in posing feature-based human action recognition which includes convolutional neural networks (CNN) and recurrent neural network (RNN). CNN-based methods are useful in recognizing human actions for combined motions (i.e., standing up, hand shaking, walking). However, in case of uncertainty of camera motion, occlusion, and multiple people, CNN suppresses important feature information and is not efficient enough to recognize variations for human action. Besides, RNN with long short-term memory (LSTM) requires more computational power to retain memories to classify human actions. This research proposes an extended framework based on capsule network using silhouette pose features to recognize human actions. Proposed extended framework achieved high accuracy of 95.64% which is higher than previous research methodology. Extensive experimental validation of the proposed extended...
Moving object detection from UAV aerial images has stated to get attention in recent years where ... more Moving object detection from UAV aerial images has stated to get attention in recent years where fast abrupt motion, low resolution, noisy imagery, cluttered background, low contrast and small target size need to be compensated with an updated moving object detection framework to increase detection performance. This paper presents 6 UCF (Uncertainty Constraint Factors) issues which are the recent categorized challenges to handle performance metric for moving object detection from UAV aerial images. A critical review for adaptive motion analysis to handle these 6 UCF factors is presented in this paper. Besides these 6 UCF factors, feature extraction for moving object detection is another unsolved issue. Large feature vector is suitable for optimum detection performance but increase computation time for overall detection process. This paper presents critical review of existing feature extraction techniques which can be compensated with adaptive motion analysis framework to influence o...
pertanika journal of science and technology, 2017
Face detection and analysis is an important area in computer vision. Furthermore, face detection ... more Face detection and analysis is an important area in computer vision. Furthermore, face detection has been an active research field in the recent years following the advancement in digital image processing. The visualisation of visual entities or sub-pattern composition may become complex to visualise due to the high frequency of noise and light effect during examination. This study focuses on evaluating the ability of Haar classifier in detecting faces from three paired Min-Max values used on histogram stretching. Min-Max histogram stretching was the selected method for implementation given that it appears to be the appropriate technique from the observation carried out. Experimental results show that, 60-240 MinMax values, Haar classifier can accurately detect faces compared to the two values.
International Journal of Image, Graphics and Signal Processing, 2021
Advances in Visual Informatics, 2019
Efficient feature extraction for moving object using aerial images is still an unsolved issue in ... more Efficient feature extraction for moving object using aerial images is still an unsolved issue in computer vision, image processing and pattern recognition research domains. Aerial types of images contain various environmental constraints due to capture frames from various altitudes level, i.e. illumination, shadows, occlusion. For this reason, appropriate feature selection for those types of images needs more attention by the researchers to improve detection rate with fast and computationally less complex features extraction method. This research performed comprehensive review with critical analysis for using various features with various methods for moving object detection using aerial images. In this context, three aspects for critical analysis of using various features are identified followed by challenges of using various features. After that, existing methods with advantages and barriers are comprehensively described with various constraints claimed by the previous research. Next, justification for the need of new feature selection is elaborated for optimum detection performance. Later, adequate validation matrics are illustrated to evaluate various features based moving object detection using aerial images performed in the previous research. The overall review performed in this paper have been comprehensively studied and expected to contribute significantly in computer vision, image processing pattern recognition research field.
Proceedings of the International Conference on Computing Advancements, 2020
In Bangladesh, a central exam is given by lots of students each year to get themselves admitted i... more In Bangladesh, a central exam is given by lots of students each year to get themselves admitted into their preferred medical college and only few get chosen by their merits. Assigning each student to a college and checking the validity of the assignment is a major concern here. This particular problem dates back to 1962 when college admission problem was introduced by Gale and Shapley [5] where stable marriage approach was used as a solution. Stability criteria was also introduced in their research which can be used to check validation of fair matching when priority is involved. But, Complicacy arises as in Bangladesh some seats are reserved for special students known as quota which can not be solved with plain stable marriage algorithm. Stability criteria also has to be defined properly for this specific problem and an algorithm is needed to handle this. A max-flow, min-cut solution combined with stable marriage algorithm is presented in this research to solve medical admission pro...
Proceedings of the International Conference on Computing Advancements, 2020
Every year thousands of people lose their lives due to traffic accidents. Road accidents, especia... more Every year thousands of people lose their lives due to traffic accidents. Road accidents, especially the ones on the highways are the most fatal ones. Accidents not only cut peoples' lives short but also cause intense financial loss to the country. Many people who survive disastrous accidents are often left with critical injury or paralysis. Particularly in Bangladesh majority of the drivers are minimally educated. They do not have sufficient knowledge and tend to ignore the traffic rules often. As a result the roads are filled with careless drivers. Consequently most of the accidents on the highways and city roads occur due to the lack of awareness of the drivers. Additionally, many of the roads are poorly lit which makes it difficult to drive in unfavorable weather. An autonomous lane detection system can play an important role as a solution to the problem by assisting the driver in seeing the lane clearly. It can also generate warning to the driver in case of an unintentional...
Education gives the insight to take us to the future which has not been possible to grasp due to ... more Education gives the insight to take us to the future which has not been possible to grasp due to lack of research methodologies towards learning process. Traditional formal education methodologies feel like living in the old age as existing learning processes are not advanced like technology which starts from primary education. As a result, children are losing interest in education due to the way of learning. In this context, learning is a continuous process where the quality cannot be justified by just including reading and writing, the overall process should be creative and interactive as well. However, existing primary education system seems to abandon the main idea about education and learning process to lean towards memorization and rote learning. The way of acquisition of education should be through discussion, research, storytelling, training which should be fun and interesting but our traditional primary education is based on two-dimensional materials and does not provide an...
Edge features or characteristics are the significant local intensity variation of pictorial struc... more Edge features or characteristics are the significant local intensity variation of pictorial structure in an image or frame or video sequence. Edge detection process finds the existence and position of pixels causes by significant differences in intensity of the images or frames or video sequences. However, selection of appropriate edge features specially to detect moving objects using aerial types of video sequences is still an elusive research concern because of various constraints for aerial frames, i.e. various altitudes, lack of features available for detection, motion variation etc. This research proposes comparative evaluation of different edge features based moving object detection methods to make it easy to decide which edge features based detection method is appropriate for extraction of moving object from aerial frames. Proposed research selected three edge features for detection, i.e. Sobel, Prewitt and Canny for moving object extraction to process high and low pictorial ...
AIUB Journal of Science and Engineering (AJSE), 2020
Detection of vacant parking space is becoming a challenging task gradually. Space utilization and... more Detection of vacant parking space is becoming a challenging task gradually. Space utilization and management of vehicle space is now a demandable field of research. Searching for an empty parking space in congested traffic is a time-consuming process. The existing vacant parking space detection methods are not robust or generalized for images captured from different camera viewpoints. Finding a proper parking space in a busy city is really a challenging issue and people are facing this problem on a daily basis. The main purpose of this research is to comprehensively discuss the previous researches of vacant parking space detection and compare them from different aspects. Methods used in previous researches are descriptively discussed along with their advantages and disadvantages. The frameworks of previous researches were compared on six generalized phases and the experimental results are compared in terms of dataset, accuracy, processing time and other performance measures. This r...
Soft Computing Approach for Mathematical Modeling of Engineering Problems, 2021
The Fourth industrial revolution (IR 4.0) saw the emergence of computer vision and artificial int... more The Fourth industrial revolution (IR 4.0) saw the emergence of computer vision and artificial intelligence in creating smart imaging systems that can replace human vision and decision making especially to predict models for autonomous vehicles. In this context, advanced prediction of probable collision in real time scenario is an unsolved problem especially in the use of artificial intelligence and computer vision for autonomous vehicles. This research proposed an efficient collision avoidance model to avoid collision in real time scenario. Proposed model differs from other methods in a way that it does not require any other equipment like sensors for measuring distance between the vehicles. Proposed collision avoidance model estimates the relation between distance and size of the vehicle in real time scenario to generate an approximate notion of distance between the vehicles. Then, the ratio of distance between vehicles and size of the vehicle was used to depict vehicles that are in potentially dangerous positions for probable collision. Proposed collision avoidance model was experimented in the real-time traffic and experimental results showed that the model could detect vehicles in order to avoid the probable collisions efficiently. Proposed model is expected to be a possible tool in dealing with future demand of autonomous vehicles with the increase of 4IR technologies.
International Journal of Advanced Computer Science and Applications, 2020
International Journal of Information Systems and Engineering, 2013
Object detection is a fundamental step for automated video analysis in many highlevel UAV surveil... more Object detection is a fundamental step for automated video analysis in many highlevel UAV surveillance tasks, including cooperative UAV path planning, navigation control, and automated information analysis. Moving object detection in aerial video is still a challenging problem for the reason that when capturing the video the camera (or the platform) is moving all the time. As a result, the problem is detecting moving object from moving background which is much more difficult than the case that the background is constant. Moving object detection in stationary scene usually modelling the pixel value changes over time, but in aerial video the change does not have regular patterns. Therefore, we model the motion of the background rather than modelling the background directly.In this paper we present a low complexity long term motion analysis based moving object detection approach by using the ideas of robust analysis and spatiotemporal clustering. Here we present an efficient algorithm to estimate the global vehicle-camera motion. Our experimental results show the efficiency of the proposed algorithm.
International Journal of Engineering Trends and Technology, 2017
Determining a Pupil size, diameter and center are fundamental for pupil orientation. It is import... more Determining a Pupil size, diameter and center are fundamental for pupil orientation. It is important features to detect eye and It is considered as a significant verification method for human computer interaction. In This paper, We investigated existing methods and presented the framework to detect pupil to calculate its distance. The existing methods of pupil orientation have classified in 4 section which are estimation and measurement, localization, detection, and tracking. We have shown the tabular study of an algorithm, detected feature and accuracy for each classification sector. There are several investigations that are running to classified all the sectors accurately. We have also proposed a framework to calculate pupil distance from images. We have described the algorithms to detect and straighten face, detect eyes. We have also proposed that a modified Circle Equation can be better to detect and exact pupils based on circle size, object polarity, and sensitivity. Although, we have discussed distance calculation.
International Journal of Image, Graphics and Signal Processing, 2019
High returning rate of garments products have become a notable problem for online fashion shoppin... more High returning rate of garments products have become a notable problem for online fashion shopping. This problem is partially caused by using different standards for measuring cloth sizes on different websites. In this research, we have designed a set of equipment to capture images of t-shirts of any color and propose an automatic cloth measurement approach using image processing techniques. A method has been introduced to recognize feature points, which has been used to calculate the cloth sizes. The method has provided a useful and efficient tool for cloth measurement. The photographs have been taken in a controlled environment, and then clothes have been categorized with the proportions of the neck, shoulder, chest width, upper waist, lower waist, and length. In this method, we have measured the t-shirt size for men by calculating the chest width and length of men. For this, a dataset has been created in a specific environment. This method has integrated with a webbased application. We have validated our work by calculating RMSE values.
International Journal of Education and Management Engineering, 2019
International Journal of Education and Management Engineering, 2019