Color recognition in vehicles is a topic widely discussed in the literature given that color is one of the key features that define a vehicle identity. Recently, the need to recognize and describe a vehicle's appearance in traffic surveillance has grown in demand, as a result of the need for efficient traffic monitoring systems. In this work, we present a different approach to recognize the predominant color of a vehicle, with minimal computational resources when compared to other methods. The goal is to segment the car body from the original image and, then, recognize the predominant color in the segmented image. To accomplish the segmentation task, we use Genetic Expression Programming (GEP) to evolve a mathematical expression to filter the original image leaving only the body of the vehicle. Another objective of this work is to create a dataset, annotated at the pixel level, for car body segmentation. Our results showed that the proposed approach was efficient for vehicle color recognition, possibly for a real-time implementation.
Birds of prey play an essential role in maintaining the health of their ecosystems. In Brazil, where there is a vast amount of biodiversity, the identification and monitoring of predatory birds are essential for maintaining the ecosystem. However, developing computational methods for the classification of predatory birds based on images is not trivial, given the many possible variations, such as angles, lighting, birds camouflage, and others. Nowadays, Transfer Learning (TL) approaches have gained popularity for many applications due to a large amount of knowledge previously acquired by models from huge datasets, which can be leveraged for other similar problems. In this paper, we present a dataset of birds of prey images and also introduce a baseline classification benchmark using the TL approach. The experiments were divided into two subcategories: families and species classification. The proposed dataset contains 42,475 samples, from 6 families and 41 species. The samples of the dataset contain birds in different positions and angles, with great variety with respect to background and illumination. Baseline results achieved an F1-Score of 92% in family and 80% in species classification.
Car make and model classification is an issue frequently discussed in the literature due to its several applications in security, traffic control, and urban planning, especially in the context of smart cities. Currently, deep learning methods are the state-of-the-art for image and video classification. This work it is presented a method for classifying cars at the level of make and model in a simple and effective way using deep learning methods. To accomplish this task, the Inception-v3 neural network was used to train and evaluate the model. Another objective of this work is to create a high-quality dataset of images of cars produced by the Brazilian industry. The full dataset has 24319 images distributed into 10 makes and 50 models, with an average of 500 images per class. The average classification accuracy reached 82.36% and 94,87%, when considering the top-3 results. Our results showed that the proposed approach was very successful for classification purposes and encourages further development.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.