Image segmentation has made great progress in recent years, but the annotation required for image segmentation is usually expensive, especially for remote sensing images. To solve this problem, we explore semi-supervised learning methods and appropriately utilize a large amount of unlabeled data to improve the performance of remote sensing image segmentation. This paper proposes a method for remote sensing image segmentation based on semi-supervised learning. We first design a Consistency Regularization (CR) training method for semi-supervised training, then employ the new learned model for Average Update of Pseudo-label (AUP), and finally combine pseudo labels and strong labels to train semantic segmentation network. We demonstrate the effectiveness of the proposed method on three remote sensing datasets, achieving better performance without more labeled data. Extensive experiments show that our semi-supervised method can learn the latent information from the unlabeled data to improve the segmentation performance.

No abstract

No abstract

Tobacco is be sensitively affected by chilling injury in the vigorous growth period, which can easily lead to tobacco leaf browning during flue-curing and quality loss, however, the physiological response of tobacco in the prosperous period under low temperature stress is unclear. The physiological response parameters of two tobacco varieties to low temperature stress were determined. The main results were as follows: ① For tobacco in the vigorous growing period subjected to low-temperature stress at 4–16 °C, the tissue structure of chloroplast changed and photosynthetic pigments significantly decreased compared with each control with the increase of intensity of low-temperature stress. ② For tobacco in the vigorous growing period at 10–16 °C, antioxidant capacity of the protective enzyme system, osmotic adjustment capacity of the osmotic adjusting system and polyphenol metabolism in plants gradually increased due to induction of low temperature with the increase of intensity of low-temperature stress. ③ Under low-temperature stress at 4 °C, the protective enzyme system, osmotic adjusting system and polyphenol metabolism of the plants played an insignificant role in stress tolerance, which cannot be constantly enhanced based on low-temperature resistance at 10 °C. This study confirmed that under the temperature stress of 10–16 °C, the self-regulation ability of tobacco will be enhanced with the deepening of low temperature stress, but there is a critical temperature between 4 and 10 °C. The self-regulation ability of plants under low temperature stress will be inhibited.

The purpose of this work is to construct a tribo-fatigue damage map of high-speed railway wheel material under different tangential forces and contact pressure conditions through JD-1 testing equipment. The results indicate that the wear rate of the wheel material varies with tangential force and contact pressure. The wear mapping of the wheel material is constructed and divided into three regions: slight wear, severe wear, and destructive wear, based on the wear rate under each test condition. With an increase in tangential force and contact pressure, the maximum crack length and average crack length of the wheel material increases. According to the surface damage morphologies and corresponding statistical results of average crack length of wheel material under each experiment condition, a tribo-fatigue damage map is constructed and divided into three regions: slight fatigue damage region, fatigue damage region, and severe fatigue damage region. Fatigue cracks initiate on the wheel specimen surface. Some cracks may propagate into material and fracture under cyclic rolling contact; some cracks may grow into inner material with a certain depth, and then turn toward the surface to form material flaking; some cracks may always propagate parallel to the wheel roller surface.