A5068315349- Advanced Image Fusion Techniques
- Inertial Sensor and Navigation
- Advanced Measurement and Detection Methods
- Image Processing Techniques and Applications
- Image and Signal Denoising Methods
- Neural Networks and Applications
- Optical and Acousto-Optic Technologies
- Engineering Technology and Methodologies
- Big Data and Business Intelligence
- Advanced Data Processing Techniques
- Cell Image Analysis Techniques
- Advanced Computational Techniques in Science and Engineering
- Thermography and Photoacoustic Techniques
Ryazan State Radio Engineering University
2020-2022
The paper proposes a novel machine learning-based approach to the pathfinding problem on extremely large graphs. This method leverages diffusion distance estimation via neural network and uses beam search for pathfinding. We demonstrate its efficiency by finding solutions 4x4x4 5x5x5 Rubik's cubes with unprecedentedly short solution lengths, outperforming all available solvers introducing first learning solver beyond 3x3x3 case. In particular, it surpasses every single case of combined best...
We consider the problem of image fusion in a multispectral vision system which quality criteria is peak signal-to-noise ratio. The aim work to develop an algorithm that allows create fused comfortable for subjective observer, even if one channels contains powerful high-frequency noise component. developed provides gain ratio by 3.4 and 4.6 times compared with known methods principal component analysis calculation average respectively.
The article considers algorithms of multiscale decomposition under the influence additive noise in one channels a multispectral vision system. difference between pyramid-based and wavelet-based decomposition-reconstruction methods is shown. Structural schemes for realization different image fusion strategies are presented, their advantages disadvantages described. To estimate fused quality authors applied complex integral-multiplicative index digital grayscale that operates with such partial...
The article discusses the details of implementation in Python programming language methods for optimizing parameters a nonlinear function according to criterion minimum mean-square error N-dimensional parameter space and L-dimensional argument space. SymPy module is used perform symbolic calculations.
The article compares the methods for noise standard deviation estimating on example of an image with a model additive white Gaussian and natural images. For over 10 we obtained approximately similar dependencies all its true value noise. images low (less than 0.8) correlation was various estimations.