<![CDATA[The research is based on the DeepSun method, a system that uses Convolution Artificial Neural Networks (CNNs) to classify the phases of sunlight by color. Light phases such as the Golden Hour, Blue Hour, and Pink Hour have distinctive visual characteristics, and identifying these light phases automatically can provide a better understanding of the mood and aesthetics of an image. The proposed approach uses a dataset consisting of images collected during various sunlight conditions. The data is annotated with the appropriate light phase label. CNNs are used to extract important features from these images. Then, those features are used as inputs for classifiers trained using machine learning algorithms. Experiments were conducted to evaluate the performance of the DeepSun system. The results obtained show that this system is able to classify the phases of sunlight with a high degree of accuracy. Misclassification mainly occurs when light conditions are very similar between certain phases. However, by increasing the amount of training data and improving the CNN architecture, the accuracy rate can be further improved. With the ability to classify the phases of sunlight. With the ability to classify the phases of sunlight automatically, DeepSun can help users to choose the right time to take quality pictures. In addition, the system can also be used to improve automatic image processing and editing based on the desired light phase.]]>
Copyrights © 2023
