Smart Boat Detection Based on Feature Pyramid Network and Deformable Convolution

Abstract The submarine cable guarantees the electricity and communication of the island residents. The operation of fishing boats poses a major threat to the submarine cable. Due to the complex environment, manual monitoring has defects such as strong subjective factors and easy fatigue. The paper adopts the intelligent monitoring method, using the object detection algorithm based on deep learning and the camera to monitor the boats on the sea. Use feature pyramid network to enhance the detection of smaller and farther boats. Use deformable convolution to solve the problem of few samples. Experimental results show that model can detect boats. The detection ability of the feature pyramid network is stronger, especially for the distant and smaller boat targets. Using deformable convolution can improve the accuracy of models trained on small dataset.

Download Full-text

Investigation of Sensor Placement in Lamb Wave-Based SHM Method

Key Engineering Materials ◽

10.4028/www.scientific.net/kem.518.174 ◽

2012 ◽

Vol 518 ◽

pp. 174-183 ◽

Cited By ~ 3

Author(s):

Pawel Malinowski ◽

Tomasz Wandowski ◽

Wiesław M. Ostachowicz

Keyword(s):

Wave Propagation ◽

Damage Detection ◽

Guided Waves ◽

Sensor Placement ◽

Detection Algorithm ◽

Piezoelectric Sensors ◽

Optimal Placement ◽

Laser Vibrometer ◽

Contact Method ◽

Monitoring Method

In this paper the investigation of a structural health monitoring method for thin-walled parts of structures is presented. The concept is based on the guided elastic wave propagation phenomena. This type of waves can be used in order to obtain information about structure condition and possibly damaged areas. Guided elastic waves can travel in the medium with relatively low attenuation, therefore they enable monitoring of extensive parts of structures. In this way it is possible to detect small defects in their early stage of growth. It is essential because undetected damage can endanger integrity of a structure. In reported investigation piezoelectric transducer was used to excite guided waves in chosen specimens. Dispersion of guided waves results in changes of velocity with the wave frequency, therefore a narrowband signal was used. Measurement of the wave field was realized using laser scanning vibrometer that registered the velocity responses at points belonging to a defined mesh. An artificial discontinuity was introduced to the specimen. The goals of the investigation was to detect it and find optimal sensor placement for this task. Determination of the optimal placement of sensors is a very challenging mission. In conducted investigation laser vibrometer was used to facilitate the task. The chosen mesh of measuring points was the basis for the investigation. The purpose was to consider various configuration of piezoelectric sensors. Instead of using vast amount of piezoelectric sensors the earlier mentioned laser vibrometer was used to gather the necessary data from wave propagation. The signals gather by this non-contact method for the considered network were input to the damage detection algorithm. Damage detection algorithm was based on a procedure that seeks in the signals the damage-reflected waves. Knowing the wave velocity in considered material the damage position can be estimated.

Download Full-text

A Temperature Monitoring Method for Power Electronic Converter Based on Infrared Image and Object Detection Algorithm

10.1109/cieec50170.2021.9510854 ◽

2021 ◽

Author(s):

Hongcheng Yang ◽

Yu Chen ◽

Yong Kang

Keyword(s):

Object Detection ◽

Infrared Image ◽

Detection Algorithm ◽

Temperature Monitoring ◽

Power Electronic ◽

Monitoring Method ◽

Power Electronic Converter

Download Full-text

GC-YOLOv3: You Only Look Once with Global Context Block

Electronics ◽

10.3390/electronics9081235 ◽

2020 ◽

Vol 9 (8) ◽

pp. 1235

Author(s):

Yang Yang ◽

Hongmin Deng

Keyword(s):

Object Detection ◽

Irrelevant Information ◽

Detection Algorithm ◽

Visual Object ◽

Detection Accuracy ◽

Feature Maps ◽

Average Precision ◽

Global Context ◽

Pascal Voc ◽

Feature Pyramid

In order to make the classification and regression of single-stage detectors more accurate, an object detection algorithm named Global Context You-Only-Look-Once v3 (GC-YOLOv3) is proposed based on the You-Only-Look-Once (YOLO) in this paper. Firstly, a better cascading model with learnable semantic fusion between a feature extraction network and a feature pyramid network is designed to improve detection accuracy using a global context block. Secondly, the information to be retained is screened by combining three different scaling feature maps together. Finally, a global self-attention mechanism is used to highlight the useful information of feature maps while suppressing irrelevant information. Experiments show that our GC-YOLOv3 reaches a maximum of 55.5 object detection mean Average Precision (mAP)@0.5 on Common Objects in Context (COCO) 2017 test-dev and that the mAP is 5.1% higher than that of the YOLOv3 algorithm on Pascal Visual Object Classes (PASCAL VOC) 2007 test set. Therefore, experiments indicate that the proposed GC-YOLOv3 model exhibits optimal performance on the PASCAL VOC and COCO datasets.

Download Full-text

Smart Machine for Crack Detection of Railway Track using Photodiode Assembly

International Journal of Innovative Technology and Exploring Engineering - Special Issue ◽

10.35940/ijitee.l2640.1081219 ◽

2019 ◽

Vol 8 (12) ◽

pp. 3679-3681

Keyword(s):

Visual Inspection ◽

Crack Detection ◽

Ultrasonic Inspection ◽

Cost Effective ◽

Detection Algorithm ◽

Railway Track ◽

Monitoring Method ◽

Robust Solution ◽

Periodic Maintenance ◽

Smart Machine

Most of the commercial goods are transported using railway trains and therefore, any problem in above network has the capacity to incur damage to the economy of that country. This model illustrates a cost effective yet robust solution to the issues related to railway crack detection. The project discusses the technical and design aspects in details alongside a better crack detection algorithm. The model also presents data related to all the components used in this system. The currently prevailing solution in the detection of cracks in rails involve periodic maintenance connected to occasional monitoring method like visual inspection, ultrasonic inspection, eddy current and laser methods

Download Full-text

Water surface object detection using panoramic vision based on improved single-shot multibox detector

EURASIP Journal on Advances in Signal Processing ◽

10.1186/s13634-021-00831-6 ◽

2021 ◽

Vol 2021 (1) ◽

Author(s):

Aofeng Li ◽

Xufang Zhu ◽

Shuo He ◽

Jiawei Xia

Keyword(s):

Object Detection ◽

Water Surface ◽

Detection Algorithm ◽

Single Shot ◽

Panoramic Vision ◽

Feature Pyramid ◽

The Mean ◽

Detection Effect ◽

Surface Object ◽

Improved Algorithm

AbstractIn view of the deficiencies in traditional visual water surface object detection, such as the existence of non-detection zones, failure to acquire global information, and deficiencies in a single-shot multibox detector (SSD) object detection algorithm such as remote detection and low detection precision of small objects, this study proposes a water surface object detection algorithm from panoramic vision based on an improved SSD. We reconstruct the backbone network for the SSD algorithm, replace VVG16 with a ResNet-50 network, and add five layers of feature extraction. More abundant semantic information of the shallow feature graph is obtained through a feature pyramid network structure with deconvolution. An experiment is conducted by building a water surface object dataset. Results showed the mean Average Precision (mAP) of the improved algorithm are increased by 4.03%, compared with the existing SSD detecting Algorithm. Improved algorithm can effectively improve the overall detection precision of water surface objects and enhance the detection effect of remote objects.

Download Full-text