Sorting flower is important for farming, plant studies, and nature. A good way to do this is by using models that have already learned from many images before. This study looks at how to use these models to recognize flowers. It uses deep learning methods like AlexNet and VGG16 with a special math rule called a logistic map. Both AlexNet and VGG16 can identify flowers well, but the choice depends on the task. The dataset has many flower types with different colors, shapes, and textures. First, the pre-trained models are adjusted to learn more about flowers. Then, different models are compared to see which one works best in terms of accuracy and speed. Extra steps, like changing the images and adding rules, help the models work better. Tests show that transfer learning helps a lot in sorting flowers better than training from the beginning. The best results are 90.12% accuracy with AlexNet and 95.45% with VGG16, both using the logistic map. The best model does a great job of telling flowers apart, showing that this method works well even when there are not many labeled images.