Iris texture is one of the most secure biometric characteristics used for person recognition, where the most significant step in the iris identification process is effective features extraction. Deep Convolutional Neural network models have been achieved massive success in the features extraction field in recent years, but several of these models have tens to hundreds of millions of parameters, which affect the computational time and resources. A lot of systems proposed in the iris recognition field extract features from normalized iris images after applying many pre-processing steps. These steps affect the quality and computational efficiency of these systems; also, occlusion, reflections, blur, and illumination variation affect the quality of features extracted. This paper proposed a new robust approach for iris recognition that locates the iris region based on the YOLOv4-tiny, then it extracts features without using iris images’ pre-processing, which is a delicate task. In addition, we have proposed an effective model that accelerated the feature extraction process by reducing the architecture of the Inception-v3 model. The obtained results on four benchmark datasets validate the robustness of our approach, where we achieved average accuracy rates of 99.91%, 99.60%, 99.91%, and 99.19% on the IITD, CASIA-Iris-V1, CASIA-Iris-Interval, and CASIA-Iris-Thousand datasets, respectively.

Automated plant classification using tree trunk has attracted increasing interest in the computer vision community as a contributed solution for the management of biodiversity. It is based on the description of the texture information of the bark surface. The multi-scale variants of the local binary patterns have achieved prominent performance in bark texture description. However, these approaches encode the scale levels of the macrostructure separately from each other. In this paper, a novel handcrafted texture descriptor termed multi-scale Statistical Macro Binary Patterns (ms-SMBP) is proposed to encode the characterizing macro pattern of different bark species. The proposed approach consists of defining a sampling scheme at high scale levels and summarizing the intensity distribution using statistical measures. The characterizing macro pattern is encoded by an in-depth gradient that describes the relationship between the scale levels and their adaptive statistical prototype. Besides this handcrafted feature descriptor, a learning-based description is performed with the ResNet34 model for bark classification. Extensive and comprehensive experiments on challenging and large-scale bark datasets demonstrate the effectiveness of ms-SMBP to identify bark species and outperforming different multi-scale LBP approaches. The tree trunk classification with ResNet34 shows interesting results on a very large-scale dataset.

Image classification has been one of the most widely topic in artificial intelligence, deep models need larger datasets and powerful hardware to improve the highperformance classification. ImageNet Challenge was started in 2010 to classify 100,000 test images into 1000 different classes. Tiny ImageNet challenge is similar to ImageNet challenge, where images are taken from the standard ImageNet and resized to be 64x64. In this paper a fine-tuned Xception to classify images into the 200 classes is presented using the standard Tiny ImageNet dataset, the down-sampling (64x64) of images and the low similarity inter-class makes feature extraction and classification difficult and more challenging. We used a transfer learning algorithm to fine-tune the Xception architecture using the Extreme version of the Inception module to achieve a high validation accuracy of 65.14%.

With the huge increase of large-scale multimedia over Internet, especially images, building Content-Based Image Retrieval (CBIR) systems for large-scale images has become a big challenge. One of the drawbacks associated with CBIR is the very long execution time. In this article, we propose a fast Content-Based Image Retrieval system using Spark (CBIR-S) targeting large-scale images. Our system is composed of two steps. (i) image indexation step, in which we use MapReduce distributed model on Spark in order to speed up the indexation process. We also use a memory-centric distributed storage system, called Tachyon, to enhance the write operation (ii) image retrieving step which we speed up by using a parallel k-Nearest Neighbors (k-NN) search method based on MapReduce model implemented under Apache Spark, in addition to exploiting the cache method of spark framework. We have showed, through a wide set of experiments, the effectiveness of our approach in terms of processing time.

With the huge increase of large-scale multimedia over Internet, especially images, building Content-Based Image Retrieval (CBIR) systems for large-scale images has become a big challenge. One of the drawbacks associated with CBIR is the very long execution time. In this article, we propose a fast Content-Based Image Retrieval system using Spark (CBIR-S) targeting large-scale images. Our system is composed of two steps. (i) image indexation step, in which we use MapReduce distributed model on Spark in order to speed up the indexation process. We also use a memory-centric distributed storage system, called Tachyon, to enhance the write operation (ii) image retrieving step which we speed up by using a parallel k-Nearest Neighbors (k-NN) search method based on MapReduce model implemented under Apache Spark, in addition to exploiting the cache method of spark framework. We have showed, through a wide set of experiments, the effectiveness of our approach in terms of processing time.

This article presents a computer vision system for real-time detection and robust recognition of speed limit signs, specially designed for intelligent vehicles. First, a new segmentation method is proposed to segment the image, and the CHT transformation (circle hog transform) is used to detect circles. Then, a new method based on local binary patterns is proposed to filter segmented images in order to reduce false alarms. In the classification phase, a cascading architecture of two linear support vector machines is proposed. The first is trained on the GTSRB dataset to decide whether the detected region is a speed limit sign or not, and the second is trained on the MNIST dataset to recognize the sign numbers. The system achieves a classification recall of 99.81% with a precision of 99.08% on the GTSRB dataset; in addition, the system is also tested on the BTSD and STS datasets, and it achieves a classification recall of 99.39% and 98.82% with a precision of 99.05% and 98.78%, respectively, within a processing time of 11.22 ms.

The security of multimedia documents becomes an urgent need, especially with the increasing image falsifications provided by the easy access and use of image manipulation tools. Hence, usage of image authentication techniques fulfills this need. In this paper, we propose an effective self-embedding fragile watermarking scheme for color images tamper detection and restoration. To decrease the capacity of insertion, a Bayer pattern is used to reduce the color host image into a gray-level watermark, to further improve the security Torus Automorphism permutation is used to scramble the gray-level watermark. In our algorithm, three copies of the watermark are inserted over three components (R, G, and B channels) of the color host image, providing a high probability of detection accuracy and recovery if one copy is destroyed. In the tamper detection process, a majority voting technique is used to determine the legitimacy of the image and recover the tampered regions after interpolating the extracted gray-level watermark. Using our proposed method, tampering rate can achieve 25% with a high visual quality of recovered image and PSNR values greater than 34 (dB). Experimental results demonstrate that the proposed method affords three major properties: the high quality of watermarked image, the sensitive tamper detection and high localization accuracy besides the high-quality of recovered image.