The availability of good quality fruits and vegetables is paramount in preventing starvation and minimizing outbreaks of diseases which leads to improving quality of life. One of the major obstacles of the mentioned a...
详细信息
ISBN:
(数字)9783030960407
ISBN:
(纸本)9783030960407;9783030960391
The availability of good quality fruits and vegetables is paramount in preventing starvation and minimizing outbreaks of diseases which leads to improving quality of life. One of the major obstacles of the mentioned availability is plant leaf disease. Although manpower plays a vital role in detecting such problems it is time-intensive, expensive, and very inefficient. Thus, developing a mechanism to vigorously monitor leaf's health and detect diseases of plant leaves at early stages is mandatory so that one can produce plenty. In this contribution, a system that detects leaf disease is developed using imageprocessingalgorithms, the k-nearest neighbor (KNN), support vector machine (SVM), and multilayer perception (MLP) machine learning algorithms are compared based on plant disease detection and classification systems performances. We also developed a prototype of simple-to-install technology that can recognize leaf diseases and allow medicine flow based on the results. This paper presents a smart plant health monitoring system that takes into account humidity, temperature, and soil contents.
Two dimensional 2D convolution is one of the most complex calculations and memory intensive algorithms used in imageprocessing. In our paper, we present the 2D convolution algorithm used in the Gaussian blur which is...
详细信息
ISBN:
(纸本)9789897585111
Two dimensional 2D convolution is one of the most complex calculations and memory intensive algorithms used in imageprocessing. In our paper, we present the 2D convolution algorithm used in the Gaussian blur which is a filter widely used for noise reduction and has high computational requirements. Since, single threaded solutions cannot keep up with the performance and speed needed for imageprocessing techniques. Therefore, parallelizing the image convolution on parallel systems enhances the performance and reduces the processing time. This paper aims to give an overview on the performance enhancement of the parallel systems on image convolution using Gaussian blur algorithm. We compare the speed up of the algorithm on two parallel systems: multi-core central processing unit CPU and graphics processing unit GPU using Google Colaboratory or "colab".
Depth maps obtained by commercial depth sensors are more likely to have missing values due to the occlusion effect, low-reflection objects, etc. Filling holes of depth maps is an important way to meet the demands of d...
详细信息
ISBN:
(数字)9789811910531
ISBN:
(纸本)9789811910531;9789811910524
Depth maps obtained by commercial depth sensors are more likely to have missing values due to the occlusion effect, low-reflection objects, etc. Filling holes of depth maps is an important way to meet the demands of depth related computer vision tasks. In this paper, we propose an efficient end-to-end network that takes RGB image and mask of the hole to jointly guide the depth hole-filling. Previous algorithms indistinguishably treat valid pixels and holes, resulting in inaccurate depth values prediction and blurred boundaries. Nevertheless, the proposed algorithm uses the bidirectional attention mechanism which takes the surrounding valid values as the auxiliary information to focus on the process of depth hole-filling from edge to center. The proposed method achieves competitive performance on existing public datasets.
A Convolutional Neural Network (CNN) is one branch of Deep Learning widely used for image classification. CNN have complex architectures and capable of achieving high accuracy and producing good results. However, CNN ...
详细信息
This paper evaluates the impact of hybrid deep learning approaches on lung tumor segmentation by combining traditional imageprocessing techniques with advanced AI-driven models. The study integrates Convolutional Neu...
详细信息
ISBN:
(数字)9798331506520
ISBN:
(纸本)9798331506537
This paper evaluates the impact of hybrid deep learning approaches on lung tumor segmentation by combining traditional imageprocessing techniques with advanced AI-driven models. The study integrates Convolutional Neural Networks (CNNs) with preprocessing methods such as noise reduction, adaptive thresholding, and contrast enhancement to address challenges associated with complex anatomical structures and variability in medical image quality. A novel hybrid framework is proposed, leveraging traditional methods to preprocess data and improve input quality for deep learning models, ultimately enhancing segmentation accuracy and *** effectiveness of the proposed approach is assessed using quantitative performance metrics, including Dice Similarity Coefficient (DSC), Hausdorff Distance, Jaccard Index, Precision, and Recall. Preliminary results indicate significant improvements in tumor boundary detection and reduced false-positive rates compared to existing methods. By streamlining segmentation workflows and enabling near-realtime applications in clinical settings, this research offers a pathway to improved diagnostic accuracy, treatment planning, and workflow *** implications include the potential for integration into clinical imaging pipelines, fostering advancements in computer-assisted diagnosis and personalized treatment strategies. This study underscores the value of hybrid methodologies in addressing current limitations and paving the way for more precise and efficient medical image segmentation.
This study aims to address two challenging problems that affect the accurate and reliable recognition of ship infrared (IR) images in various scenarios: the interference of radiation highlights from specular reflectio...
详细信息
Generative Adversarial Networks are employed by GAN-based models in image steganography to efficiently conceal information from images while preserving their visual integrity. These models are made up of two primary p...
详细信息
ISBN:
(数字)9798350360660
ISBN:
(纸本)9798350360677
Generative Adversarial Networks are employed by GAN-based models in image steganography to efficiently conceal information from images while preserving their visual integrity. These models are made up of two primary parts: the discriminator which determines if an image is a real image or a stego-image, and the generator which embeds the hidden data into images. The generator learns how to embed data in a way that minimizes obvious distortions through adversarial training, where the discriminator and generator fight to make the hidden data less noticeable to detection algorithms and human observers. High data capacity and versatility are strengths of GAN-based steganography models, which enable efficient data concealment across a variety of image formats and steganographic techniques. They dynamically adjust the embedding procedure to strike a compromise between image quality and data capacity. They do, however, come with certain drawbacks such as the need for sophisticated model training, high processing overhead and hyperparameter sensitivity. Notwithstanding these drawbacks, GAN-based techniques mark a substantial breakthrough in the development of delicate and reliable image steganography techniques.
Constrained multiobjective optimization problems (CMOPs) are prevalent in various real-world applications, presenting a formidable challenge to existing evolutionary algorithms when faced with intricate constraints. W...
详细信息
Chaotic systems are widely used in cryptography due to their traversability, unpredictability and sensitivity to initial values. However, the traditional low-dimensional chaotic system has the disadvantages of small k...
详细信息
ISBN:
(数字)9798350355413
ISBN:
(纸本)9798350355420
Chaotic systems are widely used in cryptography due to their traversability, unpredictability and sensitivity to initial values. However, the traditional low-dimensional chaotic system has the disadvantages of small key space, short iteration period and easy to be constructed, which affects the security of encryption. Therefore, in order to solve this problem, this paper constructs a new one-dimensional chaotic mapping, which is experimentally shown to have wider chaotic range, better unpredictability and larger key space. On this basis, this paper proposes an encryption scheme that can encrypt different images of different sizes by iteratively diffusing the pixels in both directions through chaotic sequences, and the experimental results show the security and effectiveness of the algorithm.
The process of manual delineating is frequently time-consuming and can result in low consistency. Our goal was to design a deep discriminative model (DDM) to mitigate these issues of magnetic resonance imaging (MRI) f...
详细信息
暂无评论