Automated surface defect detection is a challenging problem that has attracted major attention for decades. Traditional methods were designed using a pipeline of carefully designed operations. The resulting methods were complex systems, which were difficult to tune and adapt to different problems or data. A new approach to solving this problem has emerged recently: deep learning. This trend is motivated by two main factors: the increasing digitization of society, which makes it possible to record large datasets of labeled samples; and the availability of a large pool of computational resources. This work evaluates state-of-the-art deep-learning methods in object detection and semantic segmentation in the field of automated surface inspection in metals. Images acquired in the industry are affected by the conditions of the environment, including noise, dust, and vibrations, which is an additional challenge. Moreover, industrial inspection requires accuracy, but also robustness and speed. The selected methods are applied to different datasets of images that include the most common defects in metals and the performance is compared in terms of accuracy and speed. Results show exceptional accuracy at a fraction of the required processing time.

The need for improved quality control in industry makes object detection crucial. This work addresses the challenging problem of subsurface defect detections using a combination of active thermography and deep learning. The novel contribution of this work is to pose the problem as one of object detection rather than semantic segmentation or classification. The images used as input for the deep learning algorithms are three-channel color images obtained using principal component thermography (PCT). The use of these images improves the signal-to-noise ratio (SNR). A framework to label ground truths automatically is also created. The most widely used deep learning detector algorithms were evaluated, and the fifth version of you only look once (YOLO), i.e., YOLOv5, was selected because of its excellent average precision (AP) and its low inference time. The resulting combination of this algorithm and active thermography is effective and accurate in detecting subsurface defects.

Land use classification using aerial imagery can be complex. Characteristics such as ground sampling distance, resolution, number of bands and the information these bands convey are the keys to its accuracy. Random Forest is the most widely used approach but better and more modern alternatives do exist. In this paper, state-of-the-art methods are evaluated, consisting of semantic segmentation networks such as UNet and DeepLabV3+. In addition, two datasets based on aircraft and satellite imagery are generated as a new state of the art to test land use classification. These datasets, called UOPNOA and UOS2, are publicly available. In this work, the performance of these networks and the two datasets generated are evaluated. This paper demonstrates that ground sampling distance is the most important factor in obtaining good semantic segmentation results, but a suitable number of bands can be as important. This proves that both aircraft and satellite imagery can produce good results, although for different reasons. Finally, cost performance for an inference prototype is evaluated, comparing various Microsoft Azure architectures. The evaluation concludes that using a GPU is unnecessarily costly for deployment. A GPU need only be used for training.

The recognition of livestock activity is essential to be eligible for subsides, to automatically supervise critical activities and to locate stray animals. In recent decades, research has been carried out into animal detection, but this paper also analyzes the detection of other key elements that can be used to verify the presence of livestock activity in a given terrain: manure piles, feeders, silage balls, silage storage areas, and slurry pits. In recent years, the trend is to apply Convolutional Neuronal Networks (CNN) as they offer significantly better results than those obtained by traditional techniques. To implement a livestock activity detection service, the following object detection algorithms have been evaluated: YOLOv2, YOLOv4, YOLOv5, SSD, and Azure Custom Vision. Since YOLOv5 offers the best results, producing a mean average precision (mAP) of 0.94, this detector is selected for the creation of a livestock activity recognition service. In order to deploy the service in the best infrastructure, the performance/cost ratio of various Azure cloud infrastructures are analyzed and compared with a local solution. The result is an efficient and accurate service that can help to identify the presence of livestock activity in a specified terrain.

Quality control and defect detection are major challenges in industrial environments. The localization of these defects is of crucial importance, as they can affect the performance of manufactured products and even the safety of people. Defect detection methods based on visual sensors and image processing are nowadays the most common approaches. To detect subsurface defects it is common to combine active thermography and deep learning. In this paper, PCT (Principal Components Thermography) is employed to enhance the SNR (signal-to-noise ratio), leading to a significant improvement in the results. Recent developments in this field apply deep learning models for semantic segmentation and object detection. However, the quality of the predictions is not enough to ensure that all quality controls are met. In this paper, a combination of semantic segmentation and object detection is proposed to increase the reliability of predictions. To carry out this combination, a working methodology and two new ensembling strategies are proposed. After comparing the results of the proposed combination with the results of using only semantic segmentation methods in a real industrial scenario, where carbon fiber sheets are used, it is found that the proposal improves the segmentation metrics by a 5% to a 24%. Thus the reliability of the predictions is improved.