Search

This paper examines a deep convolutional neural network (Deep CNN) for plant recognition in the natural environment. The primary objective was to compare 4 CNN architectures including LeNet-5, AlexNet, GoogLeNet, and VGGNet on three plant datasets; PNE, 102 Flower, and Folio. The images in the PNE and 102 Flower dataset include a complicated background because they were taken in a natural environment. On the other hand, the images in the Folid dataset are only leaf images that were taken in a laboratory environment using a white background. The comparison of deep CNN using GoogLeNet and VGGNet Architecture show that GoogLeNet outperformed while working on the PNE and 102 Flower dataset when using a training time with iterations of 10,000 epochs. also faster than the VGGNet architecture. However, the experiment showed that the VGGNet architecture outperforms the other CNN architectures on the Folio dataset and used only 1,000 epochs for training. In our experiment, we can create a model from the deep CNN using GoogleNet architecture, and this is because it showed better results with the plant images that were taken in the natural environment. [ABSTRACT FROM AUTHOR]

This paper aims to do a comparative study of local feature descriptor techniques and convolutional neural networks (CNN) for retrieving Thai silk pattern images. Two feature descriptor techniques, the histogram of oriented gradients and the scale-invariant feature transform, are compared to extract feature vectors from the silk pattern images. We combined the feature vectors extracted from feature descriptor techniques with k-nearest neighbors (KNN) and support vector machine. Then we modified CNN architectures: LeNet-5 and AlexNet. The LeNet-5 was modified by increasing the number of neurons in each layer of the fully connected layers. The AlexNet architecture was modified by reducing the neurons in each layer of the fully connected layers. Finally, we evaluated the local descriptor techniques, the existing CNN architectures and our modified CNN architectures on Thai silk pattern dataset. The results of the study showed that the local descriptor techniques combined with KNN algorithm significantly outperform the CNN methods. [ABSTRACT FROM AUTHOR]

This paper examines a deep convolutional neural network (Deep CNN) for plant recognition in the natural environment. The primary objective was to compare 4 CNN architectures including LeNet-5, AlexNet, GoogLeNet, and VGGNet on three plant datasets; PNE, 102 Flower, and Folio. The images in the PNE and 102 Flower dataset include a complicated background because they were taken in a natural environment. On the other hand, the images in the Folid dataset are only leaf images that were taken in a laboratory environment using a white background. The comparison of deep CNN using GoogLeNet and VGGNet Architecture show that GoogLeNet outperformed while working on the PNE and 102 Flower dataset when using a training time with iterations of 10,000 epochs. GoogLeNet also faster than the VGGNet architecture. However, the experiment showed that the VGGNet architecture outperforms the other CNN architectures on the Folio dataset and used only 1,000 epochs for training. In our experiment, we can create a model from the deep CNN using GoogleNet architecture, and this is because it showed better results with the plant images that were taken in the natural environment.