The dark-field defect inspection system occupies 70% of the market in the field of unpatterned wafer inspection. But the detection limit is still restrained by the haze signals. Signal-to-noise ratio (SNR) enhancement could effectively decrease the detection limit by decreasing the influence of the haze signals on the defect signals. The existing method of optimizing the inspection conditions, including beam path and collection channel, can enhance the SNR, but the effect is restrained by the system structure. The empirically designed aperture has been attempted to be used by blocking the scattering signals in a certain azimuth angle range. However, the performance is restrained, as a signal with a large SNR exists in the blocked scattering signals. In this paper, we propose a novel (we believe) aperture design method in the light of scattering field analysis to reduce the influence of the haze signals caused by the wafer surface roughness on the particle signals. On the basis of the bidirectional reflectance distribution function, apertures are designed according to the ratio field of the particle signal to haze and verified by the scattering model developed based on the tools of the National Institute of Standards and Technology. Additionally, incident conditions are optimized according to their influence on the SNR. It is noteworthy that the aperture designed under specific conditions cannot be used for all particles. Three aperture combination schemes are proposed in this paper, which can ensure the scattering characteristics such as intensity and sensitivity to meet the system requirements while improving the contrast. Simulation results verify that the detection limit decreases from 48 to 22 nm by introducing a well-designed aperture, with the case of p -polarized incident light when the threshold is 3 and the incident angle is 72°. Multiaperture schemes have better performance over others, especially the one-to-one scheme.

Retinal vessels are the only deep micro vessels that can be observed in human body, the accurate identification of which has great significance on the diagnosis of hypertension, diabetes and other diseases. To this end, a retinal vessel segmentation algorithm based on residual convolution neural network is proposed according to the characteristics of the retinal vessels on fundus images. Improved residual attention module and deep supervision module are utilized, in which the low-level and high-level feature graphs are joined to construct the encoder-decoder network structure, and atrous convolution is introduced to the pyramid pooling. The experiments result on the fundus image data set DRIVE and STARE show that this algorithm can obtain complete retinal vessel segmentation as well as connected vessel stems and terminals. The average accuracy on DRIVE and STARE reaches 95.90 and 96.88%, and the average specificity is 98.85 and 97.85%, which shows superior performance compared to other methods. This algorithm is verified feasible and effective for retinal vessel segmentation of fundus images and has the ability to detect more capillaries.

As an important part of lab-on-a-chip and micro-total analysis systems, micromixers have a wide range of applications in biochemical analysis, pharmaceutical preparation and material synthesis. In the work, a novel rhombic separation and recombination micromixer with baffles was presented to further improve the performance of the micromixer and study the effect of multiple structural parameters on mixing. The effects of the rhombic angle, the width ratio of sub-channel and the size and relative positions of baffles on the mixing index were studied numerically at different Reynolds numbers (Re), and the sensitivity of the mixing index to various structures was also investigated. The results showed that the mixing index increased with the subchannel’s width ratio and slowly decreased after reaching the peak value in the range of Re from 0.1 to 60. The maximum mixing index appeared when the width ratio was 6.5. The pressure drops in the microchannel were proportional to the width ratio. The mixing effect can be further improved by adding baffle structure to asymmetric rhombus micromixer, and more baffle quantity and larger baffle height were beneficial to the improvement of the mixing index. The research results can provide reference and new ideas for the structure design of passive micromixers.

Active hand prostheses controlled using surface electromyography (SEMG) signals have been used for decades to restore the grasping function. Amputees with myoelectric hands wish to control the prostheses according to their own will and act like human hands as much as possible. Therefore, substantial research efforts have been put forth to advance the control of myoelectric hands. However, the tactile sensing and feedback of the myoelectric hands are still missing, thus limit hand grasp capabilities. In fact, integration of tactile sensing and feedback with hand prostheses plays an important role in improving the manipulation performance and enhancing perceptual embodiment for users. This paper reviews current state-of-the-art of tactile sensing technologies, including tactile sensor types and integration methods. Then, it introduces the basic theory of SEMG signals and presents an overview of the sensory feedback employed to prosthetic hand. The paper concludes with a detailed discussion of challenging issues and future developments.