RGB-D Image Multi-Target Detection Method Based on 3D DSF R-CNN

2019 ◽  
Vol 33 (08) ◽  
pp. 1954026 ◽  
Author(s):  
Qi Hu ◽  
Lang Zhai
Keyword(s):  
Target Detection ◽  
Color Image ◽  
Detection Algorithm ◽  
Detection Methods ◽  
Experimental Result ◽  
Depth Information ◽  
Two Dimensional ◽  
Information Need ◽  
Image Detection ◽  
Depth Cameras

At present, the application of deep learning algorithms in two-dimensional color image detection is being continuously innovated and broken. With the popularity of depth cameras, color image detection methods with depth information need to be upgraded. To solve this problem, a multi-target detection algorithm based on 3D DSF R-CNN (Double Stream Faster R-CNN, Convolution Neural Network based on Candidate Region) is proposed in this paper. The RGB information and the depth information of the image are given to two input elements of the convolution network with the same structure and weight sharing, and an optimal fusion weight algorithm is used to determine the weight of the fusion target in accordance with the recognition accuracy of the recognition targets under the single modal information, so as to ensure the most efficient fusion result. After several convolution operations, the independent features are extracted and the two networks are fused according to the optimal weights in the convolution layer. With the conducting of convolution and extract the fused features, and finally get the output through the full link layer. Compared with the previous two-dimensional convolution network algorithm, this algorithm improves the detection rate and success rate while ensuring the detection time. The experimental result shows that this method has strong robustness for complex illumination and partial occlusion, and has excellent detection results under non-restrictive conditions.

scholarly journals Real-Time Vehicle Detection Algorithm Based on Vision and Lidar Point Cloud Fusion

2019 ◽  
Vol 2019 ◽  
pp. 1-9 ◽  
Author(s):  
Hai Wang ◽  
Xinyu Lou ◽  
Yingfeng Cai ◽  
Yicheng Li ◽  
Long Chen
Keyword(s):  
Deep Learning ◽  
Real Time ◽  
Autonomous Vehicles ◽  
Point Cloud ◽  
Vehicle Detection ◽  
Detection Algorithm ◽  
Detection Methods ◽  
Depth Information ◽  
Detection Accuracy ◽  
Classification Rate

Vehicle detection is one of the most important environment perception tasks for autonomous vehicles. The traditional vision-based vehicle detection methods are not accurate enough especially for small and occluded targets, while the light detection and ranging- (lidar-) based methods are good in detecting obstacles but they are time-consuming and have a low classification rate for different target types. Focusing on these shortcomings to make the full use of the advantages of the depth information of lidar and the obstacle classification ability of vision, this work proposes a real-time vehicle detection algorithm which fuses vision and lidar point cloud information. Firstly, the obstacles are detected by the grid projection method using the lidar point cloud information. Then, the obstacles are mapped to the image to get several separated regions of interest (ROIs). After that, the ROIs are expanded based on the dynamic threshold and merged to generate the final ROI. Finally, a deep learning method named You Only Look Once (YOLO) is applied on the ROI to detect vehicles. The experimental results on the KITTI dataset demonstrate that the proposed algorithm has high detection accuracy and good real-time performance. Compared with the detection method based only on the YOLO deep learning, the mean average precision (mAP) is increased by 17%.

scholarly journals Image target detection algorithm compression and pruning based on neural network

2021 ◽  
Vol 18 (2) ◽  
pp. 499-516
Author(s):  
Yan Sun ◽  
Zheping Yan
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Target Detection ◽  
Network Model ◽  
Detection Algorithm ◽  
Image Feature ◽  
Propagation Mode ◽  
Learning Technology ◽  
Image Detection ◽  
Detection Algorithms

The main purpose of target detection is to identify and locate targets from still images or video sequences. It is one of the key tasks in the field of computer vision. With the continuous breakthrough of deep machine learning technology, especially the convolutional neural network model shows strong Ability to extract image feature in the field of digital image processing. Although the model research of target detection based on convolutional neural network is developing rapidly, but there are still some problems in practical applications. For example, a large number of parameters requires high storage and computational costs in detected model. Therefore, this paper optimizes and compresses some algorithms by using early image detection algorithms and image detection algorithms based on convolutional neural networks. After training and learning, there will appear forward propagation mode in the application of CNN network model, providing the model for image feature extraction, integration processing and feature mapping. The use of back propagation makes the CNN network model have the ability to optimize learning and compressed algorithm. Then research discuss the Faster-RCNN algorithm and the YOLO algorithm. Aiming at the problem of the candidate frame is not significant which extracted in the Faster- RCNN algorithm, a target detection model based on the Significant area recommendation network is proposed. The weight of the feature map is calculated by the model, which enhances the saliency of the feature and reduces the background interference. Experiments show that the image detection algorithm based on compressed neural network image has certain feasibility.

scholarly journals A Self-Adaptive Wildfire Detection Algorithm with Two-Dimensional Otsu Optimization

2020 ◽  
Vol 2020 ◽  
pp. 1-12
Author(s):  
Guoyong Zhang ◽  
Bo Li ◽  
Jing Luo ◽  
Lifu He
Keyword(s):  
Transmission Lines ◽  
Detection Algorithm ◽  
Power Grids ◽  
Detection Methods ◽  
Two Dimensional ◽  
Electrical Transmission ◽  
Brightness Temperatures ◽  
Otsu Method ◽  
Test Algorithm ◽  
Self Adaptive

The gradual increase in wildfires has caused frequent trips and outages along electrical transmission lines, which is a serious threat to the operational stability of power grids. A self-adaptive wildfire detection algorithm has been developed and tested in this paper. Most of existing wildfire detection methods employed fixed thresholds to identify potential wildfire pixels while the background pixels were ignored. By calculating two-dimensional histogram of the brightness temperatures of mid-infrared channel, the threshold selection is self-adaptive and potential pixels containing scenes of fire can be distinguished automatically. Based on the two-dimensional Otsu method and contextual test algorithm, an improved wildfire detection algorithm that uses multitemporal Visible and Infrared Radiometer (VIRR) data is described. The wildfire detection results within three kilometers of electrical transmission lines demonstrate the effectiveness of the proposed method, which has accurate low-temperature wildfire detection ability.

Real-Time Text Detection in Color Images Based on Color Projection

2011 ◽  
Vol 480-481 ◽  
pp. 84-88
Author(s):  
Jian Jie Wu ◽  
Yu Hui Zhang
Keyword(s):  
Real Time ◽  
Production Line ◽  
Color Image ◽  
Detection Algorithm ◽  
Text Detection ◽  
Color Images ◽  
Two Dimensional ◽  
Defect Inspection ◽  
One Dimensional ◽  
Different Types

In SMT production line, different types of components may have the same shape but providing different functions. The only difference between these components is the text on surface of a component indicating its type. Therefore, not only geometry defect inspection but also text detection is needed in component inspection to avoid wrong use of components. Traditional algorithms based on pixel comparison of text image are time consuming and sensitive to tiny change of the text as well. A concise text detection algorithm based on color projection is proposed. The algorithm transfers two-dimensional color image to one-dimensional curve for comparison by projection of the text image, which greatly reduces the computing amount, increases speed and makes the algorithm less sensitive to displacement or rotation of the text. Experiments show that the algorithm can ensure effective real-time text detection.

scholarly journals A Self-Improving Framework for Joint Depth Estimation and Underwater Target Detection from Hyperspectral Imagery

2021 ◽  
Vol 13 (9) ◽  
pp. 1721
Author(s):  
Jiahao Qi ◽  
Pengcheng Wan ◽  
Zhiqiang Gong ◽  
Wei Xue ◽  
Aihuan Yao ◽  
...  
Keyword(s):  
Target Detection ◽  
Structural Information ◽  
Hyperspectral Imagery ◽  
Contextual Information ◽  
Water Environment ◽  
Depth Estimation ◽  
Detection Methods ◽  
Depth Information ◽  
Underwater Target Detection ◽  
Underwater Target

Underwater target detection (UTD) is one of the most attractive research topics in hyperspectral imagery (HSI) processing. Most of the existing methods are presented to predict the signatures of desired targets in an underwater context but ignore the depth information which is position-sensitive and contributes significantly to distinguishing the background and target pixels. So as to take full advantage of the depth information, in this paper a self-improving framework is proposed to perform joint depth estimation and underwater target detection, which exploits the depth information and detection results to alternately boost the final detection performance. However, it is difficult to calculate depth information under the interference of a water environment. To address this dilemma, the proposed framework, named self-improving underwater target detection framework (SUTDF), employs the spectral and spatial contextual information to pick out target-associated pixels as the guidance dataset for depth estimation work. Considering the incompleteness of the guidance dataset, an expectation-maximum liked updating scheme has also been developed to iteratively excavate the statistical and structural information from input HSI for further improving the diversity of the guidance dataset. During each updating epoch, the calculated depth information is used to yield a more diversified dataset for the target detection network, leading to a more accurate detection result. Meanwhile, the detection result will in turn contribute in detecting more target-associated pixels as the supplement for the guidance dataset, eventually promoting the capacity of the depth estimation network. With this specific self-improving framework, we can provide a more precise detection result for a hyperspectral UTD task. Qualitative and quantitative illustrations verify the effectiveness and efficiency of SUTDF in comparison with state-of-the-art underwater target detection methods.

scholarly journals Vehicle Detection in Remote Sensing Image Based on Machine Vision

2021 ◽  
Vol 2021 ◽  
pp. 1-12
Author(s):  
Liming Zhou ◽  
Chang Zheng ◽  
Haoxin Yan ◽  
Xianyu Zuo ◽  
Baojun Qiao ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Target Detection ◽  
Detection Algorithm ◽  
Detection Methods ◽  
Small Target ◽  
Remote Sensing Images ◽  
Target Feature ◽  
Original Algorithm ◽  
Wide Range ◽  
Small Targets

Target detection in remote sensing images is very challenging research. Followed by the recent development of deep learning, the target detection algorithm has obtained large and fast growth. However, in the application of remote sensing images, due to the small target, wide range, small texture, and complex background, the existing target detection methods cannot achieve people’s hope. In this paper, a target detection algorithm named IR-PANet for remote sensing images of an automobile is proposed. In the backbone network CSPDarknet53, SPP is used to strengthen the learning content. Then, IR-PANet is used as the neck network. After the upper sampling, depthwise separable convolution is used to greatly avoid the lack of small target feature information in the convolution of the shallow network and increase the semantic information in the high-level network. Finally, Gamma correction is used to preprocess the image before image training, which effectively reduces the interference of shadow and other factors on training. The experiment proves that the method has a better effect on small targets obscured by shadows and under the color similar to the background of the picture, and the accuracy is significantly improved based on the original algorithm.

Two-dimensional non-reconstruction compressive sensing adaptive target detection algorithm

2019 ◽  
Vol 48 (1) ◽  
pp. 126001
Author(s):  
曹文焕 Cao Wenhuan ◽  
黄树彩 Huang Shucai ◽  
赵 炜 Zhao Wei ◽  
黄 达 Huang Da
Keyword(s):  
Compressive Sensing ◽  
Target Detection ◽  
Detection Algorithm ◽  
Two Dimensional

Target Detection of Sonar Image Based on Bis-Parameter with Adaptive Windows

2013 ◽  
Vol 347-350 ◽  
pp. 3407-3410
Author(s):  
Ke Li ◽  
Zhong Liu ◽  
Sheng Liang Hu ◽  
Yang Liu
Keyword(s):  
Target Detection ◽  
Detection Algorithm ◽  
High Accuracy ◽  
New Method ◽  
Experimental Result ◽  
Image Process ◽  
Sar Image ◽  
Sonar Image

The detection algorithm CFAR is very mature in SAR image process field and the efficiency is very good. In the paper, CFAR is tried to be used in sonar image process. In order to solve the problem that the target part leaking to the background, a new method target detection of sonar image based on bis-parameter with adaptive windows is proposed. The size of adaptive windows can be adjusted to totally cover different targets. The experimental result showed that the complex multi target can be detected by the proposed method in a high accuracy.

Anti-Noise Car License Plate Location Algorithm Based on Mathematical Morphology Edge Detection

2014 ◽  
Vol 513-517 ◽  
pp. 1052-1054 ◽  
Author(s):  
Wen Bo Liu ◽  
Tao Wang
Keyword(s):  
Edge Detection ◽  
Mathematical Morphology ◽  
Color Image ◽  
Detection Algorithm ◽  
Experimental Result ◽  
License Plate ◽  
Vertical Projection ◽  
Edge Detection Algorithm ◽  
Open Operation ◽  
Location Algorithm

We proposed an improved mathematical morphology edge detection algorithm, aimed at the significance of car location in car license plate recognition system. The first step is the true color image pre-processing, an improved mathematical morphology edge detection algorithm is used to detect the edge of the car image and after the image binarization, the morphology method is used to fill the image, and then get the candidate area after corrosion expansion after open operation. Then, according to the area of the candidate region than the circumference and vertical projection used in comprehensive analysis, the license plate area can be located accurately. The experimental result shows that the method is an effective anti-noise car plate lisence location algorithm.

scholarly journals A time-saving automobile assembly state monitoring system based on channel-pruned YOLOv4 algorithm

2021 ◽  
Author(s):  
Jiang Daqi ◽  
Hong Wang
Keyword(s):  
Neural Network ◽  
Monitoring System ◽  
Target Detection ◽  
Detection Algorithm ◽  
Experimental Result ◽  
State Monitoring ◽  
Time Saving ◽  
Industrial Environment ◽  
Automobile Assembly ◽  
Speed And Accuracy

Abstract A time-saving automobile assembly state monitoring system in industrial environment is presented in this paper. The system only needs to input a video which contains the whole detected parts and manually label in the first frame. By finding the best point for tracking and tracking the point, the dataset can be automatically generated which saves time spent on manufacturing the dataset and makes the assembly state monitoring system easy to deploy into a practical industrial environment. The target detection algorithm uses the channel-pruned YOLOv4 neural network. The experimental result shows the algorithm balances speed and accuracy. Compared to original YOLOv4, our proposed method is two times faster and the mAP is nearly equal to it. It shows that the channel pruning process dynamically improves the speed of the forward propagation without sacrifice accuracy.

Sign in / Sign up

Export Citation Format

Share Document