Machine vision is increasingly replacing manual steel surface inspection. The automatic inspection of steel surface defects makes it possible to ensure the quality of products in the steel industry with high accuracy. However, the optimization of inspection time presents a great challenge for the integration of machine vision in high-speed production lines. In this context, compressing the collected images before transmission is essential to save bandwidth and energy, and improve the latency of vision applications. The aim of this paper was to study the impact of quality degradation resulting from image compression on the classification performance of steel surface defects with a CNN. Image compression was applied to the Northeastern University (NEU) surface-defect database with various compression ratios. Three different models were trained and tested with these images to classify surface defects using three different approaches. The obtained results showed that trained and tested models on the same compression qualities maintained approximately the same classification performance for all used compression grades. In addition, the findings clearly indicated that the classification efficiency was affected when the training and test datasets were compressed using different parameters. This impact was more obvious when there was a large difference between these compression parameters, and for models that achieved very high accuracy. Finally, it was found that compression-based data augmentation significantly increased the classification precision to perfect scores (98–100%), and thus improved the generalization of models when tested on different compression qualities. The importance of this work lies in exploiting the obtained results to successfully integrate image compression into machine vision systems, and as appropriately as possible.
The future Intelligent Transport System "ITS" is one of the major challenges of the smart city. It requires fast and efficient communication between vehicles (vehicle-to-vehicle “V2V”), to ensure information exchange in order to improve safety, which reduces accidents and consequently save lives, hence the need of the Vehicular Ad Hoc Network “VANET”, which makes possible the inter-vehicle communication. This network is characterized by a variable topology. Therefore, MANET (Mobile Adhoc NETwork) routing protocols need a few tweaks to be available for the vehicle environment. In this paper, we start by exposing some works related to the evaluation of the most well-known protocols. After a comparative study, we deduce that the OLSR (Optimized Link State Routing) protocol outperforms other routing protocols in terms of End-to-End Delay (EED) and Packet Delivery Ratio (PDR). In addition, we note that the Greedy forwarding “GF” mechanism is suited for the VANET environment, which has been improved and called Greedy forwarding Advanced “GFA”, to overcome the stationary node problem. Our approach improves the OLSR protocol to be more suitable and efficient for VANET by introducing the GFA mechanism. Moreover, we compare our approach to the OLSR classic version. In this work, we use a realistic scenario from Open Street Map (OSM), and simulations are performed using SUMO (Simulation of Urban MObility). The trace files generated from SUMO are used for further simulation in NS-3 (Network Simulator) to validate our proposition. The simulation results are analyzed and discussed. Our approach performs best compared to OLSR in terms of EED and PDR, especially for dense traffic.
20 November 2020 in Wireless Personal Communications