To improve the segmentation performance of multi-objective evolutionary clustering algorithms, this paper proposes a parallel dual broad learning surrogate assisted semi-supervised kernel multi-objective evolutionary ...
详细信息
ISBN:
(纸本)9798350349122;9798350349115
To improve the segmentation performance of multi-objective evolutionary clustering algorithms, this paper proposes a parallel dual broad learning surrogate assisted semi-supervised kernel multi-objective evolutionary rough fuzzy clustering algorithm (PDBLS-SKMRFC) for image segmentation. First, the algorithm constructs a parallel dual broad learning surrogate assisted multi-objective evolutionary framework. It uses two broad learning systems as classification and regression surrogate models to evaluate the population in parallel manner. In the framework, a multi-population division strategy guided by dual broad learning system, a dominant individual crossover strategy, a sub-population mutation strategy, and a hierarchical environment selection strategy are designed to obtain more excellent offspring populations. Then, a semi-supervised kernel rough fuzzy intra-class compactness function is constructed, which uses a few labels and pseudo-labels as supervision information to improve the image segmentation performance, and evaluates the clustering quality together with the kernel separability function. Finally, a semi-supervised kernel rough fuzzy clustering validity index is designed to select the optimal solution from the final non-dominated solution set for image segmentation. Experimental results on color images show the effectiveness of the proposed algorithm.
As one of the main carriers of information transmission and storage, image data need to be compressed in some application scenarios. Traditional image compression methods, although reducing the size of image files to ...
详细信息
Artificial intelligence (AI) is a computer technology that has a human way of thinking and can imitate the human brain. The integration of AI into medical imaging has revolutionized diagnosis and treatment, providing ...
详细信息
In this paper, we present a novel transformer-based architecture for end-to-end image compression. Our architecture incorporates blocks that effectively capture local dependencies between tokens, eliminating the need ...
详细信息
The human skin is a remarkable structure, susceptible to numerous known and unknown diseases. Thus, diagnosing skin conditions is one of the most uncertain and complex areas in medical science, making clinical image a...
详细信息
Super-resolution, which is applied to produce sharp and high-resolution images, has become an important research direction in the field of imageprocessing. Interpolation-based super-resolution algorithms are widely u...
详细信息
This work is devoted to the development of a novel deep learning encoder-decoder algorithm for real-time noise and blur elimination in video frames, received from UAV. This work improves on existing algorithms by prov...
详细信息
ISBN:
(纸本)9798350372557
This work is devoted to the development of a novel deep learning encoder-decoder algorithm for real-time noise and blur elimination in video frames, received from UAV. This work improves on existing algorithms by providing a more flexible blind deblurring solution than existing kernel-based methods. The proposed method can be applied to both improve the drone operator's capabilities and to improve the performance of autonomous imageprocessing tasks, such as object identification and visual navigation systems. Different types of blur as well as possible types of noise are presented. A brief overview of existing methods is provided. The problem of frame alignment due to the object's movement and associated noise is considered. Existing deblurring and image restoration methods are reviewed, including state-of-the-art. Their limitations are highlighted. To solve the limitations a method based on a fully convolutional encoder-decoder network with residual connections is presented. Dataset generation and training procedures are discussed. The approach is then compared to existing state-of-the-art deep learning methods. The proposed method enables up to 9 times faster blind image restoration with comparable quality in comparison to existing state-of-the-art image restoration methods.
With 3D reconstruction method, this paper conducts experiments based on the data of a waste incineration plant in eastern China, and analyzes and compares the visual 3D monitoring, laser sensor 3D monitoring, and grab...
详细信息
In recent years, target tracking has been widely studied and applied in intelligent video surveillance, human-machine interaction, radar early warning and tracking, and other fields. With the emergence of the new comb...
详细信息
This article describes the creation and development of a Web Framework for the processing of images and Patterns based on the power of the HTML5 language and the object-oriented capabilities of the Javascript language...
详细信息
ISBN:
(数字)9783031298608
ISBN:
(纸本)9783031298592;9783031298608
This article describes the creation and development of a Web Framework for the processing of images and Patterns based on the power of the HTML5 language and the object-oriented capabilities of the Javascript language, as well as the importance of its functionality on any operating system, thereby improving and simplifying its use. The object-oriented implementation will make it easier to access image variables and characteristics (width, height, color depth and bitmap), the calculation algorithms and matrix processing have been modified to perform convolution and color discretization processes, as well as the use of Gaussian filters and the Sobel, Cannis, Roberts Operator. Leaving the data ready to search for patterns and generate color histograms. As a result, a clear, compact implementation was obtained. to be able to derive and run on any browser that supports HTML5 technology and, more importantly, on any operating system. We provide greater flexibility and portability than other CGI frameworks based on C++ and Python.
暂无评论