The existing object detection algorithm based on the deep convolution neural network needs to carry out multilevel convolution and pooling operations to the entire image in order to extract a deep semantic features of the image. The detection models can get better results for big object. However, those models fail to detect small objects that have low resolution and are greatly influenced by noise because the features after repeated convolution operations of existing models do not fully represent the essential characteristics of the small objects. In this paper, we can achieve good detection accuracy by extracting the features at different convolution levels of the object and using the multiscale features to detect small objects. For our detection model, we extract the features of the image from their third, fourth, and 5th convolutions, respectively, and then these three scales features are concatenated into a one-dimensional vector. The vector is used to classify objects by classifiers and locate position information of objects by regression of bounding box. Through testing, the detection accuracy of our model for small objects is 11% higher than the state-of-the-art models. In addition, we also used the model to detect aircraft in remote sensing images and achieved good results.

No abstract

No abstract

No abstract

The dynamic gas injection method with high-speed horizontal oscillation could reduce the cell size and improve the cell quality of aluminum foams, so it is of interest to study the compressive performance and deformation mechanism of the foams produced with this process. In-situ compression in X-ray tomography was used for this research. Results have shown that the plateau stresses of bulk aluminum foams prepared by the dynamic gas injection method are in the range of 0.3 ~ 11 MPa. When the cell size is reduced to around 1 mm, the plateau stress could reach 22 MPa. In addition, the brittle deformation characteristics of aluminum foams in the quasi-static compression process are obvious. The dynamic gas injection method could greatly improve the mechanical properties of aluminum foams, and aluminum foams prepared by this method have a better compressive performance compared to that prepared by static method even at the same relative density. The uniformity of the cell size and sphericity also affects the mechanical properties. The result of in-situ compression shows that there are two main failure modes for cell walls of aluminum foams: the fracture after buckles of the cell walls and the direct fracture of the cell walls. The aluminum foams prepared by the dynamic gas injection method could have a wide application prospect due to its superior compressive performance.

No abstract

We report the design and fabrication of double-network polyurethane (PU)/nanoporous cellulose gel (NCG) nanocomposites with excellent mechanical properties, multistimuli-responsive shape-memory effects, and solvent resistance using NCG as a 3D reinforcement nanofiller for the PU network. The interconnected nanofibrillar cellulose networks of the NCG are finely distributed and preserved well in the PU network after polymerization. The modified percolation model agrees well with the mechanical properties of the PU/NCG nanocomposites. The remarkable reinforcement effect on the PU network is most probably due to the incorporation of the permanent, rigid, three-dimensional percolating network of the NCG that successfully transfers mechanical stresses through the covalent cross-linking, hydrogen bonds, and chain entanglements between the NCG and PU networks. The PU/NCG nanocomposites have excellent shape-memory properties with good thermal-and water-stimuli responsiveness, good dimensional stability, excellent solvent resistance, and outstanding mechanical properties in organic solvents, and they have considerable potential applications in switchable devices, sensors, biomaterials, and many other fields.

No abstract