Image and Video Technology 9th Pacific-Rim Symposium, PSIVT 2019, Sydney, NSW, Australia, November 18-22, 2019, Proceedings

• Main Authors: Lee, Chilwoo, Su, Zhixun, Sugimoto, Akihiro
• Format: eBook
• Language: English
• Published: Netherlands Springer Nature 2019; Springer International Publishing AG; Springer International Publishing
• Edition: 1
• Series: Lecture Notes in Computer Science
• Subjects: Data processing Computer science; Imaging systems-Congresses

This book constitutes the conference proceedings of the 9th Pacific Rim Symposium on Image and Video Technology, PSIVT 2019, held in Sydney, NSW, Australia, in November 2019.A total of 31 papers were carefully reviewed and selected from 55 submissions.The main conference comprises 11 major subject a...

Table of Contents:

• 3.3 Proof of Convergence
• 2.5 2D Histogram Based Local Enhancement -- 3 Experimental Result -- 4 Conclusion -- References -- Grapevine Nutritional Disorder Detection Using Image Processing -- Abstract -- 1 Introduction -- 2 Proposed Grapevine Nutritional Disorder Detection Technique -- 2.1 Development of Nutrient Deficiency/Toxicity Symptoms -- 2.2 Machine Learning and Image Analysis -- 2.3 App Development -- 3 Experimental Results -- 4 Conclusion -- References -- Hierarchical Colour Image Segmentation by Leveraging RGB Channels Independently -- 1 Introduction -- 2 Proposed Technique -- 2.1 Cuboid Segmentation -- 2.2 Merging -- 3 Results and Discussion -- 3.1 Evaluation Metric and Dataset -- 3.2 Performance Evaluation -- 3.3 Complexity Analysis -- 4 Conclusion -- References -- High-Resolution Realistic Image Synthesis from Text Using Iterative Generative Adversarial Network -- 1 Introduction and Related Work -- 2 Proposed Method -- 2.1 Common Techniques -- 2.2 The 1st Iteration Details -- 2.3 The 2nd Iteration Details -- 2.4 The 3rd Iteration Details -- 3 Experiments -- 3.1 Datasets -- 3.2 Results of iGAN All Iterations -- 3.3 Comparisons -- 4 Conclusions -- References -- Human Shape Reconstruction with Loose Clothes from Partially Observed Data by Pose Specific Deformation -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Overview -- 3.2 Non-rigid Registration of the SMPL Model -- 3.3 Sparse Representation of Human Shape Deformation -- 3.4 Shape Deformation Bases and Coefficients Estimation by Neural Network -- 3.5 Handling Partially Observed Data -- 4 Experiment -- 4.1 Non-rigid Registration Using SMPL Model -- 4.2 Inter-frame Interpolation -- 4.3 Intra-frame Interpolation -- 5 Conclusion -- References -- Improved Saliency-Enhanced Multi-cue Correlation-Filter-Based Visual Tracking -- 1 Introduction -- 2 Related Work -- 3 Saliency Feature -- 4 Our Approach
• 4.2 Experimental Setup -- 4.3 Quantitative Results Analysis -- 5 Conclusions -- References -- Body Detection in Spectator Crowd Images Using Partial Heads -- 1 Introduction -- 2 Related Work -- 3 Methodology -- 4 Dataset -- 5 Experiments -- 6 Results -- 7 Conclusion -- References -- Deep Learning for Breast Region and Pectoral Muscle Segmentation in Digital Mammography -- 1 Introduction -- 2 Method -- 2.1 Data -- 2.2 Data Pre-processing -- 2.3 U-Net Architecture -- 2.4 Analysis -- 3 Results -- 4 Conclusion and Discussions -- References -- Detection of Age and Defect of Grapevine Leaves Using Hyper Spectral Imaging -- Abstract -- 1 Introduction -- 2 Experimental Setup -- 3 Results and Discussions -- 3.1 Age Detection -- 3.2 Defect Detection -- 4 Conclusion -- References -- Discrete Cosine Basis Oriented Homogeneous Motion Discovery for 360-Degree Video Coding -- 1 Introduction -- 2 The Discrete Cosine Basis for Motion -- 3 Prediction Using the Proposed Motion Model -- 3.1 Discrete Cosines Basis Oriented Prediction Generation -- 3.2 Discrete Cosine Based Motion Compensated Prediction as a Reference Frame -- 4 Experimental Analysis -- 5 Conclusion -- References -- Double Channel 3D Convolutional Neural Network for Exam Scene Classification of Invigilation Videos -- Abstract -- 1 Introduction -- 2 Problem Definition and Problem Solving Model -- 2.1 Problem Definition -- 2.2 Double Channel 3D Convolutional Neural Network Model (D3DCNN) -- 2.3 Improvement on C3D Model (C1 Channel) -- 2.4 Improvement on Two-Stream Network Model (C2 Channel) -- 2.5 Double Channel Spatial-Temporal Feature Fusion Algorithms -- 3 Experiment and Analysis -- 3.1 Experimental Dataset and Experimental Settings -- 3.2 Optical Flow Experiment of Video Dataset in Exam-Room -- 3.3 Analysis of Experimental Results of EMV-1 Dataset -- 4 Conclusions and Future Work -- References
• Intro -- Preface -- Organization -- Contents -- A Fused Pattern Recognition Model to Detect Glaucoma Using Retinal Nerve Fiber Layer Thickness Measurements -- 1 Introduction -- 2 Materials and Data Analysis -- 2.1 Data Set -- 2.2 Statistical Data Analysis -- 3 Methods -- 3.1 Overview -- 3.2 Proposed Hybrid Supervised Model -- 3.3 Proposed Unsupervised Model -- 3.4 Validation of Our Study -- 4 Results -- 5 Conclusions -- References -- A Robust Face Recognition System for One Sample Problem -- 1 Introduction -- 2 Methodology -- 2.1 Pre-processing of the Input Facial Image -- 2.2 Feature Extraction Using the Proposed Feature Descriptor -- 2.3 Face Recognition Using 2D-Hiden Markov Model (2DHMM) -- 2.4 Optimal Tuning of the 2D-HMM Parameters -- 3 Results -- 4 Discussion -- 5 Conclusion -- References -- Analysis of Motion Patterns in Video Streams for Automatic Health Monitoring in Koi Ponds -- 1 Introduction -- 2 Related Work -- 3 Experimental Setup -- 3.1 Fish Observation -- 3.2 Image Acquisition -- 4 Scene Analysis with Deep Neural Networks -- 4.1 Object Detection -- 4.2 Semantic Segmentation -- 4.3 Instance Segmentation -- 5 Motion Analysis -- 5.1 Visual Assessment -- 5.2 Motion Estimation -- 5.3 Objective Motion Analysis -- 6 Experimental Results -- 7 Conclusion and Future Work -- References -- Attention-Guided Model for Robust Face Detection System -- 1 Introduction -- 2 Related Works -- 3 Attention-Guided Face Detector -- 3.1 System Architecture -- 3.2 Soft Attention Mask and Data Augmentation -- 4 Experiments and Results -- 5 Conclusions -- References -- BackNet: An Enhanced Backbone Network for Accurate Detection of Objects with Large Scale Variations -- 1 Introduction -- 2 Related Works -- 3 BackNet: Proposed Backbone Network -- 3.1 Backbone Network -- 3.2 BackNet Architecture -- 4 Performance Evaluation -- 4.1 Dataset
• Efficient Self-embedding Data Hiding for Image Integrity Verification with Pixel-Wise Recovery Capability -- Abstract -- 1 Introduction -- 2 Proposed Method -- 2.1 Generating and Embedding Watermarked Data -- 2.2 Tamper Detection and Localization -- 2.3 Recovery of Tampered Blocks -- 2.4 Recovery of Tampered Image -- 3 Experimental Results -- 4 Conclusion -- References -- Enhanced Transfer Learning with ImageNet Trained Classification Layer -- 1 Introduction -- 2 Methodology -- 2.1 Traditional and Proposed Fine-Tuning -- 2.2 Proposed Depth Augmented Networks for Fine-Tuning -- 2.3 Layer-Wise Fine-Tuning Scheme -- 3 Performance Study and Analysis -- 3.1 Datasets and Implementation Details -- 3.2 Evaluation and Analysis of Proposed Fine-Tuning -- 3.3 Layers to Be Frozen for Optimal Performance -- 3.4 Performance Analysis of Proposed Depth Augmented Networks -- 3.5 Average Performance Gain -- 4 Conclusion -- References -- Equine Welfare Assessment: Horse Motion Evaluation and Comparison to Manual Pain Measurements -- 1 Introduction -- 2 Related Work -- 2.1 Image Processing -- 2.2 Horse Physiology and Pain Estimation -- 3 Veterinary Horse Welfare Evaluation -- 3.1 Experiment Layout -- 3.2 Manual Measurement Description -- 4 Extraction and Evaluation of Horse Motion Parameters -- 4.1 Experiment and Hardware Setup -- 4.2 Mask Generation for Horse Candidate Extraction -- 4.3 Extracting Horse Orientation and Viewing Direction -- 4.4 Tracking and Motion Parameters -- 5 Results -- 5.1 Interpretation of the Measurements -- 5.2 Interpreting Horses' Motion in Relation to Stress or Pain -- 6 Summary and Discussion -- References -- Exposure Correction and Local Enhancement for Backlit Image Restoration -- 1 Introduction -- 2 Proposed Approach -- 2.1 Motivation -- 2.2 Pseudo Image Generation -- 2.3 Weight Computation for Fusion -- 2.4 Multi-scale Image Fusion
• 4.1 DCF Overview -- 4.2 Tracker Robustness Measurement -- 5 Experimentation and Analysis -- 6 Conclusions and Discussion -- References -- Measuring Apple Size Distribution from a Near Top-Down Image -- 1 Introduction -- 2 COCO Trained Model for Quick Inference -- 3 Custom Training Data and Labelling -- 4 Perspective Transformation -- 5 Object Detection -- 6 Real World Coordinates -- 6.1 Ellipse Fitting and Minor Diameter Measurement -- 6.2 Box Width Measurement -- 6.3 Method Comparison -- 7 Conclusion -- References -- Multi-temporal Registration of Environmental Imagery Using Affine Invariant Convolutional Features -- 1 Introduction -- 2 Prior Work -- 3 The Proposed Deep Image Registration Framework -- 4 Experimental Results and Discussion -- 5 Conclusion -- References -- Multimodal 3D Facade Reconstruction Using 3D LiDAR and Images -- 1 Introduction -- 2 Related Work -- 3 Multimodal 3D Facade Reconstruction -- 3.1 Data Collection and Pre-processing -- 3.2 Dense RGB-D Data Generation -- 4 Evaluations -- 4.1 Evaluation of the Segmentation of the Facade Elements Made of Glass in Images -- 4.2 Evaluation of Facade Mesh Models -- 5 Comparison Study -- 6 Conclusion -- References -- Multiview Dimension Reduction Based on Sparsity Preserving Projections -- 1 Introduction -- 1.1 Organization -- 2 Related Work -- 2.1 Sparse Preserve Projection -- 2.2 Diversity-Induced Multi-view Subspace Clustering -- 3 Multiview Sparsity Preserving Projection -- 3.1 The Construction of MSPP -- 3.2 Optimization of MSPP -- 4 Experiment -- 4.1 Datasets -- 4.2 Experiments on Datasets -- 5 Conclusion -- References -- RETRACTED CHAPTER: Non-peakedDiscriminant Analysis for ImageRepresentation -- Abstract -- 1 Introduction -- 2 Motivation and Proposed Objective Function -- 3 Equivalent Formulation and Solution -- 3.1 An Equivalent Formulation -- 3.2 Solution Algorithm