Search

Image white-balancing is an integral part of every camera's processing pipeline. White-balancing is used to remove illumination chromaticity from an image. Most research in this field has been limited to images with a single uniform illuminant. In this paper, we introduce a novel method for illumination estimation for situations where the scene is illuminated by a variable number of different illuminants and where the illumination in the scene can be non-uniform. The proposed method uses a lightweight convolutional neural network that achieves state-of-the-art results. The method performs illumination estimation on a patch-by-patch basis. We use the assumption that only one illuminant affects each patch since they are so small. Unlike other such methods, our method uses features extracted from the entire image to perform patch illumination estimation. The paper also shows how the image features improve method accuracy with a minimal increase in complexity. The proposed method has around 42 k parameters, and it was tested on three different cameras from the Large-Scale Multi-Illuminant dataset. [ABSTRACT FROM AUTHOR]

Spray strategies contribute to the overall engine efficiency, combustion process, and reduction of pollutant formation in internal combustion engines with direct rail injection systems. Spray shape is determined by several parameters such as nozzle diameter, injection pressure, injector geometry, and cylinder type, which influence spray macroscopic parameters. The spray macroscopic parameters are commonly used to describe the input parameters of numerical simulations. Three main spray macroscopic parameters are spray cone angle, penetration length, and spray area. In this paper, we propose a method for spray macroscopic parameter estimation that achieves the state-of-the-art results. To obtain these results, image segmentation was performed to separate the spray from the rest of the image. Then the spray macroscopic parameters are estimated from the segmented image. To perform image segmentation Min U-Net is used. Min U-Net is a novel lightweight deep learning neural network based on U-Net. Min U-Net achieves the state-of-the-art segmentation results while having more than 500 times fewer parameters and being at least twice as fast as other learning-based methods. To evaluate the proposed method, an available dataset containing a variety of images with various spray shapes and orientations. The experiments performed on the dataset showed that Min U-Net achieved a mean dice coefficient of 0.95 with an inference time of 11.94 ms/image. The spray macroscopic parameters estimation is also highly accurate, with spray cone angle having an error of 1.08 ∘ , spray penetration length having a relative error of 5.95%, and spray area having a relative error of 4.05%. [ABSTRACT FROM AUTHOR]

Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. B Correction to: Neural Computing and Applications b https://doi.org/10.1007/s00521-023-08487-z In this article the author name Donik Vrsnak was incorrectly written as Donik Vrsnjak. [Extracted from the article]