Image and Graphics Technologies and Applications: 17th Chinese Conference, IGTA 2022, Beijing, China, April 23–24, 2022, Revised Selected Papers ... in Computer and Information Science, 1611)

Preface
Organization
Contents
Image Processing and Enhancement Techniques (Image Information Acquisition, Image/Video Coding, Image/Video Transmission, Image/Video Storage, Compression, Completion, Dehazing, Reconstruction and Display, etc.)
Multi-resolution Parallel Aggregation Network for Single Image Deraining
	1 Introduction
	2 Proposed Method
		2.1 Overall Architecture of MPA-Net
		2.2 Feature Extraction and Aggregation Module
		2.3 Cross-Scale Feature Fusion
		2.4 Loss Function
	3 Experimental Results
		3.1 Experiment Settings
		3.2 Results on Synthetic Datasets
		3.3 Results on Real-World Datasets
		3.4 Ablation Studies
	4 Conclusion
	References
Multi-scale Bézier Filter Based Infrared and Visual Image Fusion
	1 Introduction
	2 Proposed Method
		2.1 Multi-scale Bézier Filter
		2.2 Multi-scale Bézier Filter Based Infrared and Visual Image Fusion
		2.3 Parameter Settings
	3 Experimental Results and Discussions
		3.1 Experimental Settings
		3.2 Qualitative Evaluation Results
		3.3 Quantitative Evaluation Results
	4 Conclusions
	References
Towards Neural Video Compression: A Rate Distortion Modeling Perspective
	1 Introduction
	2 Related Work
	3 R-D Analysis
		3.1 R-D Statistics of Conventional Codec
		3.2 R-D Behavior of NVC
		3.3 Comparative Analysis and R-D Modeling
	4 Proposed RD-NVC
		4.1 Inter Predictive Refinement Network
		4.2 C2F Residual Modeling
		4.3 In-Loop Restoration Network
		4.4 Train for RD-NVC
	5 Experimental Results
		5.1 RD Performance
		5.2 Comparison with H.266/VVC
		5.3 Ablation Study and Analysis
	6 Conclusion
	References
A Fast Line and Ellipse Detection on High Resolution Images
	1 Introduction
	2 Precalculate-Segment-Group-Validate
	3 Segmentation of Arcs
	4 Arc Grouping and Feature Fitting
		4.1 Neighborhood Grouping
		4.2 Arc Global Grouping
		4.3 Feature Parameter Estimation
	5 Validation
	6 Experiments
		6.1 Simulation Images
		6.2 Real Images
	7 Conclusion
	References
Quality Assurance and Verification of GFDM Satellite Imagery
	1 Introduction
	2 Image Quality Analysis and Assurance
		2.1 Large-Angle Imaging
		2.2 Active Imaging Mode
	3 Image Quality Verification of GFDM Satellite
		3.1 High-Resolution Image Atmospheric Correction
		3.2 Active Imaging Results
	4 Conclusion
	References
II Machine Vision and 3D Reconstruction (Visual Information Acquisition, Camera Calibration, Stereo Vision, 3D Reconstruction, and Applications of Machine Vision In Industrial Inspection, etc.)
BSIHKS of Three-Dimensional Point Cloud Model and Model Matching: Bi-temporal Scaling Invariant Heat Kernel Signature
	1 Introduction
		1.1 Related Works
		1.2 Contribution
	2 Fundamentals and Framework
		2.1 Three-Dimensional PCL Laplace-Beltrami Operator
		2.2 Framework
	3 Bi-temporal Scaling Invariant Heat Kernel Signature
		3.1 Scale Invariance Heat Kernel Signature
		3.2 Bi Scaling Invariant Heat Kernel Signature
		3.3 Robutness of BSIHKS
		3.4 Three-Dimensional Point Cloud Models Similarity Measurement
	4 Experiments
		4.1 Effectiveness and Parameter Selection of the BSIHKS
		4.2 Robutness of the BSIHKS
		4.3 The Similarity Results Based on the BSIHKS
	5 Conclusion
	References
A Review of Solutions to Stereo Correspondence Challenges
	1 Introduction
	2 Radiometric Variations
		2.1 Pre-processing
		2.2 Matching Cost
	3 Slanted Surfaces
		3.1 Local Algorithms
		3.2 Global Algorithms
		3.3 Semi-global Algorithms
	4 Texture–Less Regions
	5 Evaluation
	6 Conclusions
	References
Low Overlapping Plant Point Cloud Registration and Splicing Method Based on FPFH
	1 Introduction
	2 Methodology
		2.1 Point Normal Estimation
		2.2 The Calculation of FPFH
		2.3 Find the Matching Point Pairs
		2.4 Strategies to Improve the Accuracy of Matching Pairs
		2.5 Calculation of Transformation
	3 Experiment and Analysis
		3.1 Registration Experiment of Public Dataset
		3.2 Registration Experiment of Noisy Plant Point Cloud
		3.3 Registration Experiment of Multi-source Plant Point Cloud
	4 Conclusion
	References
Integration of Depth Normal Consistency and Depth Map Refinement for MVS Reconstruction
	1 Introduction
	2 Related Work
	3 Main Methods
		3.1 Feature Pyramid
		3.2 Cost Volume Pyramid
		3.3 Depth Normal Consistency
		3.4 Loss Function
	4 Experiment and Result Analysis
		4.1 Datasets and Parameters Setting
		4.2 Results and Analysis
		4.3 Ablation Test
	5 Conclusion
	References
Image/Video Big Data Analysis and Understanding (Object Detection and Recognition, Image/Video Retrieval, Image Segmentation, Matching, Analysis and Understanding.)
Adaptive Registration for Multi-type Remote Sensing Images via Dynamic Feature Selection
	1 Introduction
	2 Methodology
		2.1 Dynamic Threshold Based Prematching
		2.2 Dynamic Feature Selection
		2.3 Parameter Setting
	3 Experiments and Results
		3.1 Experimental Datasets
		3.2 Evaluation Metrics and Implementation Details
	4 Results and Discussion
	5 Conclusions
	References
Deform-CAM: Self-attention Based on Deformable Convolution for Weakly Supervised Semantic Segmentation
	1 Introduction
	2 Methodology
		2.1 Class Activation Maps
		2.2 Deformable Convolution Based Self-attention Module
		2.3 Loss Design of Deform-CAM
	3 Experiments
		3.1 Implementation Details
		3.2 Ablation Studies
		3.3 Comparison with Existing State-of-the-Art Methods
	4 Conclusion
	References
Density-NMS: Cell Detection and Classification in Microscopy Images
	1 Introduce
	2 Related Word
		2.1 Anchor-Based Methods
		2.2 Imbalance Between Foreground and Background
		2.3 Post-processing Methods
	3 Method
		3.1 Guided Anchoring in the Regional Proposal Network
		3.2 Sample Weight Learning
		3.3 Density-NMS
	4 Experiments and Results
		4.1 Dataset
		4.2 Evaluation Metrics
		4.3 Experimental Result
	5 Conclusion
	References
Face Detection in Distorted Images Based on Image Restoration and Meta-learning
	1 Introduction
	2 Related Works
		2.1 Face Detection in Distorted Images
		2.2 Deep Learning for Image Restoration
		2.3 Meta-learning
	3 Method
		3.1 Preliminary
		3.2 Task-Level Data Sampling
		3.3 Meta-training
		3.4 Meta-testing
	4 Experiments
		4.1 Datasets
		4.2 Evaluations on UFDD Dataset
		4.3 Evaluations on DARKFACE Dataset
	5 Conclusion
	References
Person Re-identification Using Multi-branch Cooperative Network
	1 Introduction
	2 Multi-branch Cooperative Network
		2.1 Shade Module
		2.2 Random Erasing Module
		2.3 Stepped Module
		2.4 Multi-scale Branch
		2.5 CAP Network
	3 Experiments
		3.1 Datasets
		3.2 Implementation Details
		3.3 Experimental Results
		3.4 Ablation Experiments
		3.5 Query Results Display
	4 Conclusions
	References
Cross-Domain Object Detection Through Image-Category Features Joint Alignment
	1 Introduction
	2 Related Work
	3 Methods
		3.1 Image-Category DARetinaNet
		3.2 Image-Level Adaptive Module
		3.3 Category-Level Adaptive Module
		3.4 Network Optimization
	4 Experiments
		4.1 Datasets
		4.2 Implementation Details
		4.3 Results
		4.4 Ablation Study
	5 Conclusion
	References
Detection of Abnormal Crowd Behavior Based on ViBE and Optical Flow Methods
	1 Introduction
	2 Foreground Object Extraction
		2.1 ViBE Algorithm
		2.2 Improved Vibe Algorithm
		2.3 Foreground Extraction
	3 Target Feature Extraction Based on Optical Flow Method
		3.1 Optical Flow and Its Estimation Method
		3.2 Normalization of Optical Flow Images
		3.3 Separation of Optical Flow Velocity
	4 Evaluation of Detection Effect
	5 Conclusions
	References
Generating Lymphoma Ultrasound Image Description with Transformer Model
	1 Introduction
	2 Model
		2.1 Visual Extractor
		2.2 Encoder
		2.3 Decoder
		2.4 Deep Stable Learning Network
	3 Experiment and Result
		3.1 Experiment Settings
		3.2 Analysis of Experimental Results
	4 Conclusion
	References
Computer Graphics (Modeling, Rendering, Algorithm Simplification and Acceleration Techniques, Realistic Scene Generation, 3D Reconstruction Algorithm, System and Application, etc.)
Accelerated Photon Mapping for Homogeneous Participating Media Based on Octree
	1 Introduction
	2 Related Work
		2.1 Path Tracing
		2.2 Photon Mapping
		2.3 Caching Data Structure
	3 Light Transport in Participating Media
	4 Render Participating Media Using Octree
		4.1 Establishment of Octree
		4.2 Radiance Calculation
	5 Results
	6 Discussion
	References
Visualization and Visual Analysis (Scientific Visualization, Information Visualization and Visual Analysis.)
A Survey of Literature Analysis Methods Based on Representation Learning
	1 Introduction
	2 Sources and Type of Literature Data
		2.1 Sources of Literature Data
		2.2 Types of Literature Data
	3 RL-Based Literature Analysis Methods
		3.1 Word RL-Based Literature Analysis Methods
		3.2 Graph RL-Based Literature Analysis Methods
	4 Opportunity and Challenge
	5 Conclusion
	References
Applications of Image and Graphics (Image/Video Processing and Transmission, Biomedical Engineering Applications, Information Security, Digital Watermarking, Text Processing and Transmission, Remote Sensing, Telemetering, etc.)
A MobileNet SSDLite Model with Improved FPN for Forest Fire Detection
	1 Introduction
	2 Related Works
	3 Our Method
		3.1 Improved Feature Pyramid Network
		3.2 MobileNetSSDLite with FPN
	4 Evaluation
		4.1 Dataset
		4.2 Experimental Setup
		4.3 Evaluation index
		4.4 Experimental Results
	5 Conclusion
	References
Applications of Photogrammetry and Remote Sensing Technologies in Tianwen-1 Mission
	1 Introduction
	2 Tianwen-1 Landing Site and Data
		2.1 Landing Site
		2.2 Data
	3 Photogrammetry Technology and Applications
		3.1 Landing Site Mapping Using Orbiter Images Based on Rigorous Geometric Model
		3.2 Lander Localization Based on a Two-Step Method
		3.3 Telepresence Reconstruction Based on Multi-source Data
	4 Remote Sensing Technology and Applications
		4.1 Geomorphic Features of the Landing Zoom
		4.2 On-board Application of the in Situ Detection by the Zhurong Rover
	5 Conclusions
	References
MRI Brain Tumor Classification Based on EfficientNet with Non-rigid Transformations
	1 Introduction
	2 Method
		2.1 Overall Architecture
		2.2 EfficientNet
		2.3 Non-rigid Transformations Augmentation
		2.4 Transfer Learning
	3 Experiments and Results
		3.1 Dataset
		3.2 Experiment Settings
		3.3 Experiment Metrics
		3.4 Experiment Results
	4 Conclusion
	References
Design and Implementation of the Search System Based on the Airborne Multi-mode Imaging Cloud Platform
	1 Introduction
	2 System Design
		2.1 Composition of the System
		2.2 Process of the System Operation
	3 Key Technology of the System
		3.1 Design of the Multi-mode Cloud Platform
		3.2 Detection Technology for the Suspected Target Area
		3.3 Target Recognition Technology Based on Deep Learning
	4 Experiments
	5 Summary
	References
Method for Imaging Quality Evaluation of Onboard Camera of Image-Guided Projectile
	1 Introduction
	2 The Establish of Imaging Quality Evaluation Model
		2.1 Quality Evaluation Index of Onboard Camera
		2.2 Imaging Quality Evaluation Model of Onboard Camera
	3 Experiments
		3.1 Experimental Environment
		3.2 Evaluation Results of Measurement
	4 Conclusion
	References
Stroke Based Painting with 3D Perception
	1 Introduction
	2 Relate Work
		2.1 Image Translation and Style Transfer
		2.2 Stroke Based Rendering
		2.3 3D Pose Estimation
	3 Approach
		3.1 Perception Network
		3.2 Interactive Rendering Agent
		3.3 Bi-level Painting Agent
	4 Experiments
		4.1 Dataset
		4.2 Implementation Details
		4.3 Experiments Results and Analysis
		4.4 Ablation Study
		4.5 Gallery
	5 User Study
	6 Conclusion
	References
Author Index