CT image segmentation of foxtail millet seeds based on semantic segmentation model VGG16-UNet

Abstract

Foxtail millet is an important minor cereal crop rich in nutrients. Due to the small size of its seeds, there is little information on the diversity of its seed structure among germplasms, limiting the identification of genes controlling seed development and germination. This paper utilized X-ray computed tomography (CT) scanning technology and deep learning models to reveal the microstructure of foxtail millet seeds, gaining insights into their internal features, distribution, and composition. A total of 100 foxtail millet varieties were scanned with X-ray computed tomography to obtain 3D reconstruction images and slices. Pre-processing steps were adopted to improve image segmentation accuracy, including noise reduction, rotation, contrast enhancement, and brightness enhancement. The experiment revealed that traditional OpenCV image processing methods failed to achieve precise segmentation, whereas deep learning models exhibited outstanding performance in segmenting seed CT slice images. We compared UNet, PSPNet, and DeepLabV3 models, selected different backbones and optimizers based on the dataset, and continuously adjusted learning rates and maximum training epochs to train the models. Results demonstrated that VGG16-UNet achieved an accuracy of 99.19% on the foxtail millet seed CT slice image dataset, outperforming PSPNet and DeepLabV3 models. Compared to ResNet-UNet, VGG16-UNet shows an improvement of approximately 3.18% in accuracy, demonstrating superior performance in accurately segmenting the inner glume, outer glume, embryo, and endosperm under various adhesion conditions. Accurate segmentation of foxtail millet CT images enables analysis of embryo size, endosperm size, and glume thickness, which impact germination, growth, and nutrition. This study fills a gap in small grain structure research, offering insights to optimize agriculture and molecular breeding for improved yield and quality.

Keywords: Computed tomography (CT) images; Foxtail millet; Seed structure; UNet network.

Fig. 1

The architectural diagram of the UNet model

Algorithm 1

Foxtail millet segmentation using OpenCV

Fig. 2

VGG16-UNet model architecture diagram

Fig. 3

X-ray CT system equipment configuration

Fig. 4

3D reconstructed slice of foxtail millet seed

Fig. 5

XY plane section at different positions

Fig. 6

Diagram of data preprocessing process

Fig. 7

Data set diagram

Fig. 8

Labelme data labeling map

Fig. 9

Diagram of dataset types

Fig. 10

Different models Loss curves

Fig. 11

Different models mIoU curves

Fig. 12

Comparison of OpenCV and VGG16-UNet segmentation results

Fig. 13

Segmentation effect of different models

Fig. 14

Comparison of model test inner glume area with the actual value

Fig. 15

Comparison between model calculated and actual data