Search

As the importance of hygiene and safety management in food manufacturing has been increasingly emphasized, research on non-destructive and non-contact inspection technologies has become more active. This study proposes a real-time and non-destructive food inspection system with sub-terahertz waves which penetrates non-conducting materials by using a frequency of 0.1 THz. The proposed system detects not only the presence of foreign matter, but also the degree of depth to which it is mixed in foods. In addition, the system estimates water activity levels, which serves as the basis for assessing the freshness of seaweed by analyzing the transmittance of signals within the sub-terahertz image. The system employs YOLOv8n, which is one of the newest lightweight object detection models. This lightweight model utilizes the feature pyramid network (FPN) to effectively detect objects of various sizes while maintaining a fast processing speed and high performance. In particular, to validate the performance in real manufacturing facilities, we implemented a hardware platform, which accurately inspects seaweed products while cooperating with a conveyor device moving at a speed of 45 cm/s. For the validation of the estimation performance against various water activities and the degree of depth of foreign matter, we gathered and annotated a total of 9659 sub-terahertz images and optimized the learning model. The final results show that the precision rate is 0.91, recall rate is 0.95, F1-score is 0.93, and mAP is 0.97, respectively. Overall, the proposed system demonstrates an excellent performance in the detection of foreign matter and in freshness estimation, and can be applied in several applications regarding food safety.

The consumption of seaweed is increasing year by year worldwide. Therefore, the foreign object inspection of seaweed is becoming increasingly important. Seaweed is mixed with various materials such as laver and sargassum fusiforme. So it has various colors even in the same seaweed. In addition, the surface is uneven and greasy, causing diffuse reflections frequently. For these reasons, it is difficult to detect foreign objects in seaweed, so the accuracy of conventional foreign object detectors used in real manufacturing sites is less than 80%. Supporting real-time inspection should also be considered when inspecting foreign objects. Since seaweed requires mass production, rapid inspection is essential. However, hyperspectral imaging techniques are generally not suitable for high-speed inspection. In this study, we overcome this limitation by using dimensionality reduction and using simplified operations. For accuracy improvement, the proposed algorithm is carried out in 2 stages. Firstly, the subtraction method is used to clearly distinguish seaweed and conveyor belts, and also detect some relatively easy to detect foreign objects. Secondly, a standardization inspection is performed based on the result of the subtraction method. During this process, the proposed scheme adopts simplified and burdenless calculations such as subtraction, division, and one-by-one matching, which achieves both accuracy and low latency performance. In the experiment to evaluate the performance, 60 normal seaweeds and 60 seaweeds containing foreign objects were used, and the accuracy of the proposed algorithm is 95%. Finally, by implementing the proposed algorithm as a foreign object detection platform, it was confirmed that real-time operation in rapid inspection was possible, and the possibility of deployment in real manufacturing sites was confirmed.

Modern image processing techniques use three-dimensional (3D) images, which contain spatial information such as depth and scale, in addition to visual information. These images are indispensable in virtual reality, augmented reality (AR), and autonomous driving applications. We propose a novel method to estimate monocular depth using a cycle generative adversarial network (GAN) and segmentation. In this paper, we propose a method for estimating depth information by combining segmentation. It uses three processes: segmentation and depth estimation, adversarial loss calculations, and cycle consistency loss calculations. The cycle consistency loss calculation process evaluates the similarity of two images when they are restored to their original forms after being estimated separately from two adversarial losses. To evaluate the objective reliability of the proposed method, we compared our proposed method with other monocular depth estimation (MDE) methods using the NYU Depth Dataset V2. Our results show that the benchmark value for our proposed method is better than other methods. Therefore, we demonstrated that our proposed method is more efficient in determining depth estimation.