Depth Estimation Based on Pyramid Normalized Cross-Correlation Algorithm for Vergence Control

Mohamed, Abdulla; Culverhouse, Phil F.; Cangelosi, Angelo; Yang, Chenguang

doi:10.1109/access.2018.2877721

“…The Otsu is used for threshold segmentation [17], and the optimal threshold is determined by judging the maximum variance of two parts of pixel values. The normalized cross- correlation template matching method is used for image registration [18] - [19] to find the region matching the template image in the searched image. Finally, by using the imabsdiff function differential-operated to locate defects, the defect image is input into the improved fully convolutional neural networks to classify and identify defects.…”

Section: Algorithm Flowmentioning

confidence: 99%

Printed Circuit Boards Defect Detection Method Based on Improved Fully Convolutional Networks

Zheng

¹

,

Sun

²

,

Zhou

³

et al. 2022

View full text Add to dashboard Cite

Since printed circuit board (PCB) is the key to ensure the reliability of electronic equipment. Therefore, defect detection for PCB is a basic and necessary work. This paper proposes a PCB defect detection method based on an improved fully convolutional neural networks to detect four types of defects: spurs, mouse bites, short circuits and open circuits. The improved neural networks increase the receptive field and solve the problems of ordinary atrous convolution information loss of continuity by introducing the continuous atrous convolution module. At the same time, we introduce the improved skip layer after the upsampling layer to get the characteristics of multi-scale fusion, improve the resolution of the image. Due to the fewer levels of the whole networks, MobileNet-V2 networks model is adopted in the improved fully convolutional neural networks, which is compared with ResNet-50 and Vgg-16 networks model. Comparison of accuracy of four kinds of PCB classification defects. Experimental results show that the improved fully convolutional neural networks based on MobileNet-V2 have an average recognition accuracy of 92.86% for four types of PCB defects, which is better than the ResNet-50 and Vgg-16 networks models, verifying the effectiveness of the PCB defect recognition classification. Moreover, experiments are carried out in a real environment to verify the feasibility of the proposed method.INDEX TERMS Fully convolutional neural networks, PCB defect detection, continuous atrous convolution, skip structure.

show abstract

“…The backbone of the vergence algorithm in this thesis project is cross-correlation. As mentioned previously, various methods have been proposed for the determination of vergence point and ultimately execution of vergence, some of which were reviewed for this thesis project [43]- [50]. Some of these studies make use of disparity and phase difference for vergence but majority of the studies reviewed have a crosscorrelation backbone.…”

Section: Cross-correlation) the Third Category Is What Is Used In This Thesis Projectmentioning

confidence: 99%

Design and Build of an Anthropomorphic Active Vision System

Shaghaghi¹

2021

Preprint

0

View full text Add to dashboard Cite

The aim of this thesis project is to design an anthropomorphic active vision system which builds on biomimicry of the human visual system. The proposed system has some of the characteristics associated with such a visual system, such as degrees of freedom associated with movements of the eyes in addition to foveation and vergence capacities associated with human vision. Through this thesis, novel approaches are proposed in regard to specific elements of system design and system testing. Novel approaches have been proposed regarding foveation and vergence relating to system design and use of Inertial Measurement Unit (IMU) devices incorporating Kalman filters relating to testing of the prototyped device. For foveation, integration of fast Deep Convolutional Neural Networks (DCNN) has been proposed. For vergence, a variation of Cross-correlation has been used. This method has the benefit of being computationally inexpensive which is beneficial for real-time operations of the proposed system.

show abstract

“…In the past few decades, there has been various attempts in the area of active vision systems for creation of vergence solutions. These methods could be divided into three categories [43]: The first category makes use of feature matching for vergence solutions. This is done using methods such as Zero Disparity Filter.…”

Section: Review Of Gaze Vergence and Cross-correlationmentioning

confidence: 99%

“…Cross-correlation methods covered use either cartesian space or log polar space for operation purposes. In the past few years, cross-correlation method variations using cartesian space have had a resurrection [43]. This can likely be attributed, in part, to advancements in processing power which make it possible to run cartesian space cross-correlation, in real time.…”

Section: Cross-correlation) the Third Category Is What Is Used In This Thesis Projectmentioning

confidence: 99%

Design and Build of an Anthropomorphic Active Vision System

Shaghaghi¹

2021

Preprint

0

View full text Add to dashboard Cite

The aim of this thesis project is to design an anthropomorphic active vision system which builds on biomimicry of the human visual system. The proposed system has some of the characteristics associated with such a visual system, such as degrees of freedom associated with movements of the eyes in addition to foveation and vergence capacities associated with human vision. Through this thesis, novel approaches are proposed in regard to specific elements of system design and system testing. Novel approaches have been proposed regarding foveation and vergence relating to system design and use of Inertial Measurement Unit (IMU) devices incorporating Kalman filters relating to testing of the prototyped device. For foveation, integration of fast Deep Convolutional Neural Networks (DCNN) has been proposed. For vergence, a variation of Cross-correlation has been used. This method has the benefit of being computationally inexpensive which is beneficial for real-time operations of the proposed system.

show abstract

Depth Estimation Based on Pyramid Normalized Cross-Correlation Algorithm for Vergence Control

Cited by 3 publications

References 20 publications

Printed Circuit Boards Defect Detection Method Based on Improved Fully Convolutional Networks

Printed Circuit Boards Defect Detection Method Based on Improved Fully Convolutional Networks

Design and Build of an Anthropomorphic Active Vision System

Design and Build of an Anthropomorphic Active Vision System

Contact Info

Product

Resources

About