Search

IntroductionLung image segmentation plays an important role in computer-aid pulmonary disease diagnosis and treatment.MethodsThis paper explores the lung CT image segmentation method by generative adversarial networks. We employ a variety of generative adversarial networks and used their capability of image translation to perform image segmentation. The generative adversarial network is employed to translate the original lung image into the segmented image.ResultsThe generative adversarial networks-based segmentation method is tested on real lung image data set. Experimental results show that the proposed method outperforms the state-of-the-art method.DiscussionThe generative adversarial networks-based method is effective for lung image segmentation.

Bent branches are commonly utilized in cut roses to enhance commercial flower quality through increased nutrient production and biomass. However, the ideal number of bent branches concerning crop water productivity remains unknown. The experiment was conducted in Jinning, Yunnan, in 2022 with total of three growth periods with 1–3 bent branches. Sap flow of a commercial branch was measured using the heat ratio method to assess rose transpiration and crop water productivity during each growth period. Across three harvests, cut roses with three bent branches increased plant height by 87.9% compared to those with one bent branch (217 cm). The Fresh weight of cut roses with three bent branches was 63.6 g shoot−1, 135% more than with one bent branch. The roses with three branches produced 56% more high-grade (grade A and grade B) shoots than in other treatments. Daily transpiration in roses with three bent branches was 4.09 mm d−1 across three harvests, which was 115% higher than in one bent branch. The economic value in three bent branches was 54% higher than in two bent branches and 120% higher than in one bent branch. Crop water productivity in three bent branches across three periods was 14.7 kg m−3, which was 84.7% higher than in two bent branches and 167% than in one bent branch. Our results help farmers to optimize irrigation of cut rose in greenhouse conditions and could be used in improving rose models by quantifying transpiration at crop level.

Abstract This paper considers the problem of target tracking in complex manoeuvering scenarios with a lack of relevant prior knowledge. This is a challenge for classical model‐based manoeuvering target tracking algorithms because they rely heavily on accurate domain and prior knowledge of target motion. To address this problem, we propose a hybrid‐driven continuous‐time filter algorithm in this paper, which combines the advantages of the model‐driven and data‐driven. We use the stochastic differential equation (SDE) with the acceleration model as the basic framework of the proposed algorithm. In order to deal with unpredictable manoeuvres and unknown perturbations, we adopt neural networks as data‐driven to estimate target accelerations and compensations of composite perturbations in real time using historical and current measurements. And the direction of gradient descent of the neural network is constrained by the motion model, thus the learning efficiency of the network and the interpretability of the proposed algorithm is improved. The proposed algorithm can simultaneously utilise historical trajectory information and domain knowledge as hybrid‐driven to achieve complex manoeuvering target tracking with little prior information. Experimental results demonstrate the superiority of our algorithm in tracking accuracy, robustness and environmental adaptability.

Abstract The traditional track correlation methods have problems such as limitation of the application scenarios, unstable performances and poor practicalities. To solve those problems, a new adaptive track correlation method for multiple scenarios was proposed in this paper using the theories and methods from machine learning. Through interpreting and translating the track correlation problem in the field of information fusion into the classification recognition problem in the field of machine learning, the new method was derived based on deep convolutional neural networks. The association performance and adaptation capabilities of the proposed method had been validated by simulation experiments. The results show that the proposed method is better than the traditional methods with respect to association performance and adaptation capabilities, and can solve the track association problems for multiple scenarios, for example sensors have systematic errors and targets are dense or in formation. Thus, it can be predicted that the proposed method would have a well‐applied foreground.