Vision-based force measurement using geometric moment invariants

2012 ◽  
Vol 226 (10) ◽  
pp. 2589-2601
Author(s):  
Wen Hu ◽  
Shigang Wang ◽  
Chun Hu ◽  
Hongtao Liu ◽  
Jinqiu Mo
Keyword(s):  
Measurement Method ◽  
Feature Vector ◽  
Force Measurement ◽  
Shape Descriptor ◽  
Support Vector ◽  
Moment Invariants ◽  
Geometric Moment ◽  
New Vision ◽  
Multi Class Classification ◽  
Standard Library

This article presents a new vision-based force measurement method to measure microassembly forces without directly computing the deformation. The shape descriptor of geometric moment invariants is used as a feature vector to describe the implicit relationship between an applied force and the deformation. Then, a standard library is established to map the corresponding relationship between the deformed cantilever under known forces and a set of feature vectors. Finally, a support vector machine compares the feature vector of deformed cantilever under an unknown force with those in the standard library, implements multi-class classification and predicts the unknown force. The vision-based force measurement method is validated for eight simulated microcantilevers of different sizes. Both regional and boundary moment invariants are used to constitute the feature vector. Simulated results show that the force measurement precision varies with length, width and height of cantilevers. If length increases and width and height decrease, the precision is higher. This trend can provide a reference for mechanism design of microcantilevers and microgrippers.

A visual terrain classification method for mobile robots’ navigation based on convolutional neural network and support vector machine

2021 ◽  
pp. 014233122098791
Author(s):  
Wanli Wang ◽  
Botao Zhang ◽  
Kaiqi Wu ◽  
Sergey A Chepinskiy ◽  
Anton A Zhilenkov ◽  
...  
Keyword(s):  
Neural Network ◽  
Support Vector Machine ◽  
Mobile Robots ◽  
Convolutional Neural Network ◽  
Hybrid Method ◽  
Classification Accuracy ◽  
Support Vector ◽  
High Classification Accuracy ◽  
Enhance Efficiency ◽  
Multi Class Classification

In this paper, a hybrid method based on deep learning is proposed to visually classify terrains encountered by mobile robots. Considering the limited computing resource on mobile robots and the requirement for high classification accuracy, the proposed hybrid method combines a convolutional neural network with a support vector machine to keep a high classification accuracy while improve work efficiency. The key idea is that the convolutional neural network is used to finish a multi-class classification and simultaneously the support vector machine is used to make a two-class classification. The two-class classification performed by the support vector machine is aimed at one kind of terrain that users are mostly concerned with. Results of the two classifications will be consolidated to get the final classification result. The convolutional neural network used in this method is modified for the on-board usage of mobile robots. In order to enhance efficiency, the convolutional neural network has a simple architecture. The convolutional neural network and the support vector machine are trained and tested by using RGB images of six kinds of common terrains. Experimental results demonstrate that this method can help robots classify terrains accurately and efficiently. Therefore, the proposed method has a significant potential for being applied to the on-board usage of mobile robots.

Ultrasonic measurement of clamping force for injection molding machine

2019 ◽  
Vol 39 (4) ◽  
pp. 388-396 ◽  
Author(s):  
Peng Zhao ◽  
Yao Zhao ◽  
Jianfeng Zhang ◽  
Junye Huang ◽  
Neng Xia ◽  
...  
Keyword(s):  
Injection Molding ◽  
Measurement Method ◽  
Large Scale ◽  
Force Measurement ◽  
Low Cost ◽  
Ultrasonic Method ◽  
Injection Molding Process ◽  
Propagation Time ◽  
Clamping Force ◽  
Molding Process

AbstractAn online and feasible clamping force measurement method is important in the injection molding process and equipment. Based on the sono-elasticity theory, anin situclamping force measurement method using ultrasonic technology is proposed in this paper. A mathematical model is established to describe the relationship between the ultrasonic propagation time, mold thickness, and clamping force. A series of experiments are performed to verify the proposed method. Experimental findings show that the measurement results of the proposed method agree well with those of the magnetic enclosed-type clamping force tester method, with difference squares less than 2 (MPa)2and errors bars less than 0.7 MPa. The ultrasonic method can be applied in molds of different thickness, injection molding machines of different clamping scales, and large-scale injection cycles. The proposed method offers advantages of being highly accurate, highly stable, simple, feasible, non-destructive, and low-cost, providing significant application prospects in the injection molding industry.

Classification System for Leukemia Cell Images based on Hu Moment Invariants and Support Vector Machines

2021 ◽  
Author(s):  
Yessi Jusman ◽  
Slamet Riyadi ◽  
Amir Faisal ◽  
Siti Nurul Aqmariah Mohd Kanafiah ◽  
Zeehaida Mohamed ◽  
...  
Keyword(s):  
Support Vector Machines ◽  
Classification System ◽  
Leukemia Cell ◽  
Support Vector ◽  
Moment Invariants ◽  
Vector Machines ◽  
Hu Moment ◽  
Hu Moment Invariants

scholarly journals Multi-Classification of Fetal Health Status Using Extreme Learning Machine

2021 ◽  
Vol 5 (2) ◽  
pp. 62-70
Author(s):  
Ömer KASIM
Keyword(s):  
Extreme Learning Machine ◽  
Classification Accuracy ◽  
Binary Classification ◽  
Clinical Decision ◽  
Support Vector ◽  
Data Set ◽  
Multiple Classification ◽  
Learning Machine ◽  
Multi Class Classification ◽  
The University

Cardiotocography (CTG) is used for monitoring the fetal heart rate signals during pregnancy. Evaluation of these signals by specialists provides information about fetal status. When a clinical decision support system is introduced with a system that can automatically classify these signals, it is more sensitive for experts to examine CTG data. In this study, CTG data were analysed with the Extreme Learning Machine (ELM) algorithm and these data were classified as normal, suspicious and pathological as well as benign and malicious. The proposed method is validated with the University of California International CTG data set. The performance of the proposed method is evaluated with accuracy, f1 score, Cohen kappa, precision, and recall metrics. As a result of the experiments, binary classification accuracy was obtained as 99.29%. There was only 1 false positive.  When multi-class classification was performed, the accuracy was obtained as 98.12%.  The amount of false positives was found as 2. The processing time of the training and testing of the ELM algorithm were quite minimized in terms of data processing compared to the support vector machine and multi-layer perceptron. This result proved that a high classification accuracy was obtained by analysing the CTG data both binary and multiple classification.

Sign in / Sign up

Export Citation Format

Share Document