An accurate and fast human detection is a crucial task for a wide variety of applications such as automotive and person identification. The histogram of oriented gradients (hog) algorithm is one of the most reliable a...
详细信息
An accurate and fast human detection is a crucial task for a wide variety of applications such as automotive and person identification. The histogram of oriented gradients (hog) algorithm is one of the most reliable and applied algorithms for this task However the hog algorithm is also a compute intensive task. This paper presents three different implementations using the Zynq SoC that consists of an ARM processor and an FPGA. The first uses OpenCV functions and runs on the ARM processor. A speedup of 249 x is achieved due to several optimizations that are implemented in this OpenCV-based hog approach. The second is a HW/SW Co-Design implemented on the ARM processor and the FPGA. The third is completely implemented on the FPGA and optimized for an FPGA implementation to achieve the highest performance for high resolution images (1920 x 1080). This implementation achieves 39.6 fps which is a speedup of 503.9x compared to the OpenCV-based approach and 2x compared to this implementation with optimizations. The HW/SW Co-Design achieves a speedup of approximately 9x compared to an original hog implementation running on the ARM processor. 2017 Published by Elsevier Inc.
In this work, we dealt with topics related to computer vision, with particular focus on solutions aimed at ensuring computational efficiency. One of the algorithms often used in computer vision is Histogram of Oriente...
详细信息
ISBN:
(纸本)9798350375701;9788363578268
In this work, we dealt with topics related to computer vision, with particular focus on solutions aimed at ensuring computational efficiency. One of the algorithms often used in computer vision is Histogram of Oriented Gradients (hog), often used in object detection in video sequences. In many applications, this algorithm should be able to run in real time, simultaneously on devices with low computing power. In this work, we attempted to implement one of its key stages at the level of digital specialized circuits (also suitable for integrated circuit realization). At this stage, it is determined which of the bins the vector calculated on the basis of previously calculated gradients for the x and y coordinates belongs to. We presented solutions for different numbers of bins, but also solutions that can be considered more general, i.e. allowing to change the number of bins.
Nowadays, computer technology is developing rapidly, and image recognition technology makes a very important part of it used in many places, such as surveillance, face recognition, image matching, etc. Image recogniti...
详细信息
Accident prevention encompasses a wide range of strategies, practices, and safety measures aimed at reducing the likelihood of accidents, injuries, and fatalities in various settings, including on the road, at home, i...
详细信息
The current computer vision-based elderly falling behavior recognition algorithm mainly uses target detection followed by behavior recognition for a single human body, but there are mostly numerous human targets and n...
详细信息
The current computer vision-based elderly falling behavior recognition algorithm mainly uses target detection followed by behavior recognition for a single human body, but there are mostly numerous human targets and noise interference in the actual scene, which makes it difficult to accurately detect and recognize the anomalous behavior of the moving individual, for which detection and tracking of the moving human body is needed. In this paper, the fusion applications of histogram of oriented gradients human body detection algorithm, Camshift human body tracking algorithm, and spatial temporal graph convolutional networks gesture recognition algorithm are investigated for the recognition of anomalous behaviors in elderly people during falls. After testing with 100 motion target groups, the time lag of this paper's fusion algorithm for motion human falling behavior recognition is < 0.5 s on average, and the average accuracy of anomalous falling behavior recognition is 97.0%, which is better than that of falling behavior recognition based on acquisition devices such as environmental sensors and deep learning models such as YOLOv5 and thus provides more powerful protection for the personal safety of elderly people.
This paper focuses on detecting a pedestrian in an image. This real time application aims for high detection accuracy as well as faster computation. For higher accuracy and detection rate Histogram of Oriented Gradien...
详细信息
ISBN:
(纸本)9781509064717
This paper focuses on detecting a pedestrian in an image. This real time application aims for high detection accuracy as well as faster computation. For higher accuracy and detection rate Histogram of Oriented Gradients (hog) algorithm is used. Further, Linear Support Vector Machine (LSVM) classification is used for faster and reliable classification. Since the hog algorithm is compute expensive several modifications have been made in order to get the best results for real time application. We have used bilinear interpolation and L2-normalisation for more reliable output. Further since the data is linearly separable a LSVM is designed in Matlab. The proposed algorithm provides an accuracy of 93.27% with a high true positive rate of 92.27% and a minor false positive rate of 4%.
Sudden pedestrian crossing is the major cause of accident on the road, especially in cities. Most of the results show that drivers tend to lose their attention and always feel drowsy while driving. Some of them just t...
详细信息
ISBN:
(纸本)9781728183091
Sudden pedestrian crossing is the major cause of accident on the road, especially in cities. Most of the results show that drivers tend to lose their attention and always feel drowsy while driving. Some of them just take it easy, not focusing and giving their full attention while driving on the road. Therefore, the main objective of this project is to build a driving assistance device that can detect sudden pedestrians crossing the road using Raspberry Pi microcontroller. This can be accomplished with the following methodological steps;Open CV is used to develop a detection algorithm where the Pi camera is used to capture the image and image processing algorithm as well as the fact that a warning system is programmed via Phyton language to give an early warning to driver. A buzzer sound is used to get the driver's attention to slow down the speed of vehicle or just to stop the car. At this vehicle's speed rate, the crashing can be reduced if the driver gets the notification earlier. In conclusion, such a warning system should be available in any car to warn drivers. By creating such device, it can contribute towards reducing the percentage of pedestrian death.
The most prevalent neurodegenerative condition that substantially impairs elderly people's motor capabilities is Parkinson's disease (PD). Even today, in less developed regions of the world, the diagnosis and ...
详细信息
ISBN:
(数字)9789819990405
ISBN:
(纸本)9789819990399;9789819990405
The most prevalent neurodegenerative condition that substantially impairs elderly people's motor capabilities is Parkinson's disease (PD). Even today, in less developed regions of the world, the diagnosis and monitoring of PD remain an expensive and difficult process. This is comparative study of methods for predicting Parkinson's disease using hand-drawn spiral pictures using computer vision and machine learning approaches. In this paper, five machine learning algorithms are used, which are Decision Tree, K-Nearest Neighbors, Support Vector Classifier, Logistic Regression, and Random Forest. Along with this, hog feature descriptor is used for the extraction of the features from the spiral images. In this work, the KNN machine learning algorithm performed with the best accuracy of 90%. Similarly, the Random Forest is the second best algorithm with accuracy 83.3%.
暂无评论