RT Journal Article
T1 Adaptive image enhancement for robust CNN classification under low illumination
A1 Rodríguez-Rodríguez, José A.
A1 López-Rubio, Ezequiel
A1 Jiménez-Segura, Salvador
A1 Molina-Cabello, Miguel Ángel
K1 Redes neuronales (Informática)
K1 Imágenes
AB Convolutional Neural Networks (CNNs) are widely used in image classification tasks, but their performance candegrade significantly under poor illumination conditions. Although numerous image enhancement methods havebeen developed to mitigate this issue, identifying the most appropriate technique for a specific image remainschallenging. This work aims to determine the most effective image enhancement algorithm for a potentiallydimmed input image to enhance the classification prediction performance. To do this, a trained regressor, whichevaluates different features of an image, ascertains the increase or decrease in classification prediction perfor-mance between the input image and the enhanced image. These predictions are then used to select the mostsuitable image enhancement algorithm to be applied to the input image for optimal CNN classification. Exper-imental results using various CNN architectures and enhancement techniques demonstrate that the proposedstrategy consistently improves classification performance under challenging lighting conditions.
PB Elsevier
YR 2026
FD 2026-02
LK https://hdl.handle.net/10630/45662
UL https://hdl.handle.net/10630/45662
LA eng
NO Rodríguez-Rodríguez, José A., López-Rubio, Ezequiel, Jiménez-Segura, Salvador,  Molina-Cabello, Miguel A. (2026). Adaptive image enhancement for robust CNN classification under low illumination. Expert Systems with Applications. Elsevier. Vol. 315, 10 de junio.
NO Funding for open access charge: Universidad de Málaga / CBUA
DS RIUMA. Repositorio Institucional de la Universidad de Málaga
RD 2 mar 2026