scholarly journals FPGA Implementation of Multi-scale Face Detection Using HOG Features and SVM Classifier

2016 ◽  
Vol 21 (3) ◽  
pp. 27-44 ◽  
Author(s):  
Michał Drożdż ◽  
Tomasz Kryjak
Keyword(s):  
Face Detection ◽  
Vision System ◽  
Video Stream ◽  
Svm Classifier ◽  
Vast Number ◽  
Real Time Processing ◽  
Multi Scale ◽  
Svm Algorithm ◽  
Detection Window ◽  
Embedded Vision System

Abstract In this paper an FPGA based embedded vision system for face detection is presented. The sliding detection window, HOG+SVM algorithm and multi-scale image processing were used and extensively described. The applied computation parallelizations allowed to obtain real-time processing of a 1280 × 720 @ 50Hz video stream. The presented module has been verified on the Zybo development board with Zynq SoC device from Xilinx. It can be used in a vast number of vision systems, including diver fatigue monitoring.

scholarly journals FPGA Implementation of Multi-scale Pedestrian Detection in Thermal Images

2016 ◽  
Vol 21 (3) ◽  
pp. 55-67
Author(s):  
Tomasz Kańka ◽  
Tomasz Kryjak ◽  
Marek Gorgon
Keyword(s):  
Vision System ◽  
Pedestrian Detection ◽  
Detection Algorithm ◽  
Thermal Images ◽  
Real Time Processing ◽  
Computing Platform ◽  
Multi Scale ◽  
Field Programmable ◽  
Window Approach ◽  
Silhouette Detection

Abstract In this paper an embedded vision system for human silhouette detection in thermal images is presented. As the computing platform a reprogrammable device (FPGA – Field Programmable Gate Array) is used. The detection algorithm is based on a sliding window approach, which content is compared with a probabilistic template. Moreover, detection is four scales in supported. On the used test database, the proposed method obtained 97% accuracy, with average one false detection per frame. Due to the used parallelization and pipelining real-time processing for 720 × 480 @ 50 fps and 1280 × 720 @ 50 fps video streams was achieved. The system has been practically verified in a test setup with a thermal camera.

scholarly journals Development of Non Expensive Technologies for Precise Maneuvering of Completely Autonomous Unmanned Aerial Vehicles

Sensors ◽  
2021 ◽  
Vol 21 (2) ◽  
pp. 391
Author(s):  
Luca Bigazzi ◽  
Stefano Gherardini ◽  
Giacomo Innocenti ◽  
Michele Basso
Keyword(s):  
Unmanned Aerial Vehicles ◽  
Degrees Of Freedom ◽  
Vision System ◽  
Video Stream ◽  
Measurement Unit ◽  
Absolute Position ◽  
Light Load ◽  
Aerial Vehicles ◽  
Angular Velocities ◽  
Custom Hardware

In this paper, solutions for precise maneuvering of an autonomous small (e.g., 350-class) Unmanned Aerial Vehicles (UAVs) are designed and implemented from smart modifications of non expensive mass market technologies. The considered class of vehicles suffers from light load, and, therefore, only a limited amount of sensors and computing devices can be installed on-board. Then, to make the prototype capable of moving autonomously along a fixed trajectory, a “cyber-pilot”, able on demand to replace the human operator, has been implemented on an embedded control board. This cyber-pilot overrides the commands thanks to a custom hardware signal mixer. The drone is able to localize itself in the environment without ground assistance by using a camera possibly mounted on a 3 Degrees Of Freedom (DOF) gimbal suspension. A computer vision system elaborates the video stream pointing out land markers with known absolute position and orientation. This information is fused with accelerations from a 6-DOF Inertial Measurement Unit (IMU) to generate a “virtual sensor” which provides refined estimates of the pose, the absolute position, the speed and the angular velocities of the drone. Due to the importance of this sensor, several fusion strategies have been investigated. The resulting data are, finally, fed to a control algorithm featuring a number of uncoupled digital PID controllers which work to bring to zero the displacement from the desired trajectory.

scholarly journals A Fast and Lightweight Method with Feature Fusion and Multi-Context for Face Detection

2018 ◽  
Vol 10 (8) ◽  
pp. 80
Author(s):  
Lei Zhang ◽  
Xiaoli Zhi
Keyword(s):  
Face Detection ◽  
Graphics Processing Units ◽  
High Performance ◽  
Feature Fusion ◽  
Local Context ◽  
Data Set ◽  
Global Context ◽  
Detection Algorithms ◽  
Multi Scale ◽  
Benchmark Datasets

Convolutional neural networks (CNN for short) have made great progress in face detection. They mostly take computation intensive networks as the backbone in order to obtain high precision, and they cannot get a good detection speed without the support of high-performance GPUs (Graphics Processing Units). This limits CNN-based face detection algorithms in real applications, especially in some speed dependent ones. To alleviate this problem, we propose a lightweight face detector in this paper, which takes a fast residual network as backbone. Our method can run fast even on cheap and ordinary GPUs. To guarantee its detection precision, multi-scale features and multi-context are fully exploited in efficient ways. Specifically, feature fusion is used to obtain semantic strongly multi-scale features firstly. Then multi-context including both local and global context is added to these multi-scale features without extra computational burden. The local context is added through a depthwise separable convolution based approach, and the global context by a simple global average pooling way. Experimental results show that our method can run at about 110 fps on VGA (Video Graphics Array)-resolution images, while still maintaining competitive precision on WIDER FACE and FDDB (Face Detection Data Set and Benchmark) datasets as compared with its state-of-the-art counterparts.

A Novel Approach of Face Detection Based on Skin Color Segmentation and SVM

2011 ◽  
Vol 225-226 ◽  
pp. 437-441
Author(s):  
Jing Zhang ◽  
You Li
Keyword(s):  
Face Detection ◽  
Skin Color ◽  
Information Access ◽  
Color Segmentation ◽  
Support Vector ◽  
Svm Classifier ◽  
Novel Approach ◽  
Great Performance ◽  
Face Detection And Recognition ◽  
Detection And Recognition

Nowadays, face detection and recognition have gained importance in security and information access. In this paper, an efficient method of face detection based on skin color segmentation and Support Vector Machine(SVM) is proposed. Firstly, segmenting image using color model to filter candidate faces roughly; And then Eye-analogue segments at a given scale are discovered by finding regions which are darker than their neighborhoods to filter candidate faces farther; at the end, SVM classifier is used to detect face feature in the test image, SVM has great performance in classification task. Our tests in this paper are based on MIT face database. The experimental results demonstrate that the proposed method is encouraging with a successful detection rate.

scholarly journals Technical vision system to ensure the movement of a group of multicopters

2020 ◽  
pp. 13-24
Author(s):  
Yurii Bobkov ◽  
Pavlo Pishchela
Keyword(s):  
Vision System ◽  
Video Stream ◽  
Reference Object ◽  
Maximum Relative Error ◽  
Software Environment ◽  
Technical Vision ◽  
Additional Equipment ◽  
C Programming ◽  
In The Beginning ◽  
Color Depth

The actual task of controlling a group of multicopters performing coordinated actions and are locating at short distances from each other, cannot be performed with the help of a standard on-board autopilot on GPS or GLONASS signals, which give large errors. The solution to this problem is possible due to additional equipment that allows you to set the distance between the multicopters and their relative position. To do this, it is proposed to mark each multicopter with an image label in the form of a standard geometric figure or a geometric body of a given color and size, and to use technical vision system and image recognition algorithms. The structure of the technical vision system for the multicopter was developed and algorithms for image processing and calculation of the change of coordinates of the neighboring multicopter, which are transmitted to the control system to introduce the necessary motion correction, were proposed. The method to identify the reference object in the image of the scene by its color was used in this work. This method is very effective compared to other methods, because it requires only one pass per pixel, which gives a significant advantage in speed during video stream frame processing. RGB color model with a color depth of 24-bit was chosen based on the analysis. Since the lighting during the flight can change, the color is set by the limits of change of the components R, G, B. To determine the distance between multicopters, a very simple but effective method of determination the area of the recognition object (labels on the neighboring multicopter) with next comparation it with the actual value is used. Since the reference object is artificial, its area can be specified with high accuracy. The offset of the center of the object from the center of the frame is used to calculate the other two coordinates. In the beginning, the specific camera instance is calibrated both for a known value of the area of the object and for its displacement along the axes relative to the center of the frame. The technical vision system model in the Simulink software environment of the Matlab system was created to test the proposed algorithms. Based on the simulation results in Simulink, you can generate code in the C programming language for further implementation of the system in real time. A series of studies of the model was conducted using a Logitech C210 webcam with a 0.3 megapixel photo matrix (640x480 resolution). According to the results of the experiment, it was found that the maximum relative error in determining the coordinates of the multicopter did not exceed 6.8 %.

scholarly journals AN OPTIONAL THRESHOLD WITH SVM CLOUD DETECTION ALGORITHM AND DSP IMPLEMENTATION

2016 ◽  
Vol XLI-B8 ◽  
pp. 771-777 ◽  
Author(s):  
Guoqing Zhou ◽  
Xiang Zhou ◽  
Tao Yue ◽  
Yilong Liu
Keyword(s):  
Real Time ◽  
Detection Algorithm ◽  
High Accuracy ◽  
Svm Classifier ◽  
Detection Accuracy ◽  
Landsat 8 ◽  
Cloud Detection ◽  
Threshold Method ◽  
Dsp Implementation ◽  
Svm Algorithm

This paper presents a method which combines the traditional threshold method and SVM method, to detect the cloud of Landsat-8 images. The proposed method is implemented using DSP for real-time cloud detection. The DSP platform connects with emulator and personal computer. The threshold method is firstly utilized to obtain a coarse cloud detection result, and then the SVM classifier is used to obtain high accuracy of cloud detection. More than 200 cloudy images from Lansat-8 were experimented to test the proposed method. Comparing the proposed method with SVM method, it is demonstrated that the cloud detection accuracy of each image using the proposed algorithm is higher than those of SVM algorithm. The results of the experiment demonstrate that the implementation of the proposed method on DSP can effectively realize the real-time cloud detection accurately.

Sign in / Sign up

Export Citation Format

Share Document