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

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.

Download Full-text

Parameter Measurement of Biaxial Braided Composite Preform Based on Phase Congruency

Autex Research Journal ◽

10.1515/aut-2018-0012 ◽

2019 ◽

Vol 19 (1) ◽

pp. 8-16

Author(s):

Zhitao Xiao ◽

Lei Pei ◽

Fang Zhang ◽

Ying Sun ◽

Lei Geng ◽

...

Keyword(s):

Automatic Measurement ◽

Detection Algorithm ◽

High Accuracy ◽

Block Matching ◽

Experimental Results ◽

Corner Detection ◽

New Method ◽

Two Dimensional ◽

Phase Congruency ◽

Braided Composite

Abstract In this paper, a new method based on phase congruency is proposed to measure pitch lengths and surface braiding angles of two-dimensional biaxial braided composite preforms. Lab space transform and BM3D (block-matching and 3D filter) are used first to preprocess the original acquired images. A corner detection algorithm based on phase congruency is then proposed to detect the corners of the preprocessed images. Pitch lengths and surface braiding angles are finally measured based on the detected corner maps. Experimental results show that our method achieves the automatic measurement of pitch lengths and the surface braiding angles of biaxial braided composite preforms with high accuracy.

Download Full-text

А Nоvel Data-Driven Орtimаl Methоdоlоgy fоr Deteсting Shiр from Sаr Images Bаsed on Аrtifiсiаl Intelligenсe

Indian Journal of Artificial Intelligence and Neural Networking ◽

10.35940/ijainn.c1035.061321 ◽

2021 ◽

Vol 1 (3) ◽

pp. 17-22

Author(s):

M.S.Antony Vigil ◽

Rishabh Jain ◽

Tanmay Agarwal ◽

Abhinav Chandra

Keyword(s):

Target Detection ◽

Detection Rate ◽

Target Identification ◽

Detection Algorithm ◽

High Accuracy ◽

Data Driven ◽

Identification Rate ◽

Ship Detection ◽

Hyper Spectral ◽

Multiple Issues

There are a variety of deep learning algorithms available in the supervision of ships, but they are dealing with multiple issues of inaccurate identification rate and inadequate target detection speed. At this stage, an algorithm is given оn Соnvоlutiоnаl Neural Network for target identification and detection using the ship image. The study involves the investigation of the reactions of hyper spectral atmospheric rectification on the accurate and precise results of ship detection. The ship features which were detected from two atmospheric rectified algorithms on airborne hyper spectral data were corrected by the application of these algorithms with the help of an unsupervised target detection procedure. High accuracy and fast ship identification was a result of this algorithm and using unique modules, improving the loss function and enlargement of data for the smaller targets. The results of the experiments show that our algorithm has given much better detection rate as compared to target detection algorithm using traditional machine learning.

Download Full-text

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

10.21203/rs.3.rs-1026299/v1 ◽

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.

Download Full-text

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

International Journal of Pattern Recognition and Artificial Intelligence ◽

10.1142/s0218001419540260 ◽

2019 ◽

Vol 33 (08) ◽

pp. 1954026 ◽

Cited By ~ 4

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.

Download Full-text