Overview: In the stereo matching process, unique pixel features are extracted by aggregating local and global context information. The pixel features on the pole lines of the left and right images are then matched. With the rapid application of deep learning (DL) methods in the field of image processing, end-to-end neural networks of DL are used to estimate the disparity maps. Although CNN-based algorithms have excellent feature representation capabilities, they often exhibit limitations in modeling explicit long-range relationships due to the inherent locality of the convolution operations. For objects with weak textures and large differences in shape and size, the results of using CNN alone are often unsatisfactory. To solve this problem, an improved model STransMNet stereo matching network based on the Swin Transformer is proposed in this paper. We analyze the necessity of the aggregated local and global context information. Then the difference in matching features during the stereo matching process is discussed. The feature extraction module is improved by replacing the CNN-based algorithm with the Transformer-based Swin Transformer algorithm. The rectified left and right images are fed into Swin Transformer module to generate multi-scale features. Then the multi-scale features are fed into the patch expanding module, the transformation of the linear layer, to make them the same size. Finally, the multi-scale features are fused in the channel dimension. The additional multi-scale fusion module makes the features output by the improved Swin Transformer fuse shallow and deep semantic information. The Swin Transformer used to extract the left and right image features is partially shared by the weights. Although weight sharing makes the model converge faster, our proposed feature differentiation loss can only supervise left or right images. If the full weights are shared, it is equivalent to supervising the left and right images at the same time. Partial weight sharing speeds up the convergence of the model to a certain extent. In addition, partial weight sharing enables the model to extract not only the commonalities of left and right image but also the differences. Furthermore, a feature differentiation loss is proposed in this work to improve the model's ability to pay attention to details. The loss is trained by forcing the classification of pixel features on the epipolar line of the left image, which makes each pixel feature unique. The experimental results on the Sceneflow and KITTI datasets show that our algorithm reduces the 3 px error and EPE compared to the previous algorithms. Experiments show that the proposed STransMNet model reduces the matching error and improves the quality of the disparity maps. It shows that the excellent performance of the improved Swin Transformer in capturing long-distance context information is beneficial to improving the accuracy of stereo matching; feature differentiation loss helps to enhance the detailed information of the disparity maps.

In this paper, a high-power and high-efficiency 4.3µm mid-infrared (MIR) optical parametric oscillator (OPO) based on ZnGeP2 (ZGP) crystal is demonstrated. An acousto-optically Q-switched Ho:Y3Al5O12 laser operating at 2.1µm with a maximum average output power of 35 W and pulse width of 38 ns at a repetition rate of 15 kHz is established and employed as the pump source. A doubly resonant OPO is designed and realized with the total MIR output power of 13.27 W, including the signal and idler output power of 2.65 W at 4.07µm and 10.62 W at 4.3µm. The corresponding total optical-to-optical and slope efficiencies are 37.9% and 67.1%, respectively. The shortest pulse width, beam quality factor, and output power instability are measured to be 36 ns, Mx2=1.8, My2=2.0, and RMS<1.9% at 8 h, respectively. Our results pave a way for designing high-power and high-efficiency 4–5µm MIR laser sources.

We propose an effective way to achieve an enhanced optical absorption surface of titanium alloy 7 (Ti7) fabricated by a femtosecond (fs) laser assisted with airflow pressure. The effect of laser scanning speed and laser power on the surfaces’ morphology and average reflectivity was studied. In order to further reduce the surface’s reflectivity, different airflow pressure was introduced during the fabrication of Ti7 by a fs laser. Furthermore, the average reflectivity of samples fabricated under different laser parameters assisted with airflow was presented. In addition, the high and low temperature tests of all samples were performed to test the stability performance of the hybrid micro/nanostructures in extreme environments. It is demonstrated that the airflow pressure has an important influence on the micro/nanostructures for light trapping, the average reflectivity of which could be as low as 2.31% over a broad band of 250–2300 nm before high and low temperature tests, and the reflection for specific wavelengths can go below 1.5%.

An effective and simple method is proposed for fabricating the micro/nano hybrid structures on metal surfaces by adjusting femtosecond laser fluence, scanning interval, and polarization. The evolution of surface morphology with the micro/nano structures is discussed in detail. Also, the mechanism of light absorption by the micro/nano hybrid structures is revealed. Compared with the typical periodic light-absorbing structures, this type of micro/nano hybrid structures has an ultralow average reflectivity of 2% in the 250–2300 nm spectral band and the minimum 1.5% reflectivity in UV band. By employing this method, large areas of the micro/nano hybrid structures with high consistency could be achieved.