RESUMEN
Websites can improve their security and protect against harmful Internet attacks by incorporating CAPTCHA verification, which assists in distinguishing between human users and robots. Among the various types of CAPTCHA, the most prevalent variant involves text-based challenges that are intentionally designed to be easily understandable by humans while presenting a difficulty for machines or robots in recognizing them. Nevertheless, due to significant advancements in deep learning, constructing convolutional neural network (CNN)-based models that possess the capability of effectively recognizing text-based CAPTCHAs has become considerably simpler. In this regard, we present a CAPTCHA recognition method that entails creating multiple duplicates of the original CAPTCHA images and generating separate binary images that encode the exact locations of each group of CAPTCHA characters. These replicated images are subsequently fed into a well-trained CNN, one after another, for obtaining the final output characters. The model possesses a straightforward architecture with a relatively small storage in system, eliminating the need for CAPTCHA segmentation into individual characters. Following the training and testing of the suggested CNN model for CAPTCHA recognition, the experimental results demonstrate the model's effectiveness in accurately recognizing CAPTCHA characters.
RESUMEN
Due to the ever-increasing number and diversity of data sources, and the continuous flow of data that are inevitably redundant and unused to the cloud, the Internet of Things (IoT) brings several problems including network bandwidth, the consumption of network energy, cloud storage, especially for paid volume, and I/O throughput as well as handling huge amount of stored data in the cloud. These call for data pre-processing at the network edge before data transmission over the network takes place. Data reduction is a method for mitigating such problems. Most state-of-the-art data reduction approaches employ a single tier, such as gateways, or two tiers, such gateways and the cloud data center or sensor nodes and base station. In this paper, an approach for IoT data reduction is proposed using in-networking data filtering and fusion. The proposed approach consists of two layers that can be adapted at either a single tier or two tiers. The first layer of the proposed approach is the data filtering layer that is based on two techniques, namely data change detection and the deviation of real observations from their estimated values. The second layer is the data fusion layer. It is based on a minimum square error criterion and fuses the data of the same time domain for specific sensors deployed in a specific area. The proposed approach was implemented using Python and the evaluation of the approach was conducted based on a real-world dataset. The obtained results demonstrate that the proposed approach is efficient in terms of data reduction in comparison with Least Mean Squares filter and Papageorgiou's (CLONE) method.
