A5035193471- Image Retrieval and Classification Techniques
- Advanced Image and Video Retrieval Techniques
- Neural Networks and Applications
- Medical Image Segmentation Techniques
- Computer Graphics and Visualization Techniques
- 3D Shape Modeling and Analysis
- Handwritten Text Recognition Techniques
- Advanced Neural Network Applications
- Data Management and Algorithms
- Advanced Image Fusion Techniques
- Biometric Identification and Security
- Image and Signal Denoising Methods
- Video Analysis and Summarization
- Brain Tumor Detection and Classification
- Generative Adversarial Networks and Image Synthesis
- Multiculturalism, Politics, Migration, Gender
- Image Processing and 3D Reconstruction
- Industrial Vision Systems and Defect Detection
- Face and Expression Recognition
- Remote-Sensing Image Classification
- Vehicle License Plate Recognition
- Image and Object Detection Techniques
- African Studies and Geopolitics
- Time Series Analysis and Forecasting
- Blind Source Separation Techniques
Cadi Ayyad University
2015-2024
Advancements in multimodal learning have experienced rapid growth over the past decade, particularly within various domains, with a significant emphasis on developments computer vision. Multimodal data fusion has become increasingly prominent realm of image classification, where integration diverse sources enhances overall understanding and performance classification models. This survey delves into recent strides made field classification. Additionally, paper undertakes comparative study,...
Current artificial neural network image recognition techniques use all the pixels of an as input. In this paper, we present efficient method for handwritten digit that involves extracting characteristics a by coding each row decimal value, i.e., transforming binary representation into value. This is called rows. The set values calculated from initial arranged vector and normalized; these represent inputs to network. approach proposed in work uses multilayer perceptron classification,...
With the development of information technology and coming period large data, image signals play an increasingly more significant role in our life because phenomenal system correspondence innovation, comparing high proficiency handling strategies are requested earnestly. The Fourier transform is important processing tool, which used a wide range applications, such as filtering, analysis, compression reconstruction. It 's simplest among other transformation method mathematics. real time...
Abstract In this paper, a novel method for binary image comparison is presented. We suppose that the set of transactions and items. The proposed applies along rows columns an image; represented by all frequent itemset. Firstly, are considered as Secondly, we items transactions. Besides, also apply our technique to color firstly segment each segmented region image. tested on MPEG7 database compared with moment’s show its efficiency.
In this work, a new method is presented for the representation of 3D objects with binary matrix. This based on two stages: normalization and quantization. allows us to compare by computing similarity between them. fact our algorithm compute matrix, frequency matrix cluster coordinates. So we can identify an object comparing those representations.
In this work we discuss the problems of template matching and propose some solutions. Those are: 1) Template image search differ by a scale, 2) or is object rotation, 3) an affinity. The well known method NCC (Normalized Cross Correlation); can not handle affinity occlusion. Also preferred for binary image. So here to use index similarity example Jaccard index.
In this work, we propose to compare affine shape using Hausdorff distance (HD), Dynamic Time Warping (DTW), Frechet (DF), and Earth Mover (EMD). Where there is only a change in resolution are computed between coordinates because the not invariant under rotation or affinity. case of transformation, distances calculated but Arc length Affine length. while The main advantage invariance resolution, rotation,