Simultaneous localization and mapping (SLAM) are fundamental elements for many emerging technologies, such as autonomous driving and augmented reality. For this paper, to get more information, we developed an improved monocular visual SLAM system by using omnidirectional cameras. Our method extends the ORB-SLAM framework with the enhanced unified camera model as a projection function, which can be applied to catadioptric systems and wide-angle fisheye cameras with 195 degrees field-of-view. The proposed system can use the full area of the images even with strong distortion. For omnidirectional cameras, a map initialization method is proposed. We analytically derive the Jacobian matrices of the reprojection errors with respect to the camera pose and 3D position of points. The proposed SLAM has been extensively tested in real-world datasets. The results show positioning error is less than 0.1% in a small indoor environment and is less than 1.5% in a large environment. The results demonstrate that our method is real-time, and increases its accuracy and robustness over the normal systems based on the pinhole model.

Background: Malnutrition is the most common nutritional issue in Alzheimer’s disease (AD) patients, but there is still a lack of a comprehensive evaluation of the nutritional status in AD patients. This study aimed to determine the potential association of various nutritional indices with AD at different stages.Methods: Subjects, including individuals with normal cognition (NC) and patients diagnosed with AD, were consecutively enrolled in this cross-sectional study. Demographics, body composition, dietary patterns, nutritional assessment scales and nutrition-related laboratory variables were collected. Binary logistics regression analyses and receiver operating characteristic (ROC) curves were used to indicate the association between nutrition-related variables and AD at different stages.Results: Totals of 266 subjects, including 73 subjects with NC, 72 subjects with mild cognitive impairment due to AD (AD-MCI) and 121 subjects with dementia due to AD (AD-D) were included. There was no significant difference in dietary patterns, including Mediterranean diet and Mediterranean-DASH diet intervention for neurodegenerative delay (MIND) diet between the three groups. Lower BMI value, smaller hip and calf circumferences, lower Mini Nutritional Assessment (MNA) and Geriatric Nutritional Risk Index (GNRI) scores, and lower levels of total protein, albumin, globulin, and apolipoprotein A1 were associated with AD (all p < 0.05). Total protein and albumin levels had the greatest ability to distinguish AD from non-AD (AUC 0.80, 95% CI 0.74–0.84, p < 0.001), increased by combining calf circumference, MNA score and albumin level (AUC 0.83, 95% CI 0.77–0.88, p < 0.001). Albumin level had the greatest ability to distinguish NC from AD-MCI (AUC 0.75, 95% CI 0.67–0.82, p < 0.001), and MNA score greatest ability to distinguish AD-MCI from AD-D (AUC 0.72, 95% CI 0.65–0.78, p < 0.001).Conclusion: Nutritional status of AD patients is significantly compromised compared with normal controls, and tends to be worsened with AD progresses. Early identification and intervention of individuals with nutritional risk or malnutrition may be significantly beneficial for reducing the risk, development, and progression of AD.

A fiber nonlinearity compensation scheme based on a complex-valued dimension-reduced neural network is proposed. The proposed scheme performs all calculations in complex values and employs a dimension-reduced triplet feature vector to reduce the size of the input layer. Simulation and experiment results show that the proposed neural network needed only 20% of computational complexity to reach the saturated performance gain of the real-valued triplet-input neural network, and had a similar saturated gain to the one-step-per-span digital backpropagation. In addition, the proposed scheme was 1.7 dB more robust to the noise from training data and required less bit precision for quantizing trained weights, compared with the real-valued triplet-input neural network.

A cost-effective chromatic dispersion (CD) estimation method is proposed, which is based on the delay-tap sampling (DTS) technique in combination with periodic training sequences (PTS). By performing equivalent sampling of PTS, DTS-based CD estimation at ultra-low sampling rate can be realized, significantly reducing the cost and high-speed requirements of analog to digital converters (ADCs). In contrast to the existing Nyquist DTS-based CD estimation method, we construct the CD lookup table in the CD over-compensated region, which can be suitable for a dynamic range of OSNR conditions. The CD lookup table built from the simulation can be used in the experimental system with similar configurations directly, indicating that the suggested method is reasonable and highly feasible. Besides, with only 1.25 GSa/s sampling rate of ADCs (about 1/30 of the Nyquist sampling rate which is 40.28 GSa/s), both 20.14 GBaud quadrature-phase-shift-keying (QPSK) simulation and experimental systems demonstrate the feasibility of the method, with a 28 ps/nm maximum CD estimation error for the simulation and a 35 ps/nm maximum CD estimation error for the experiment when OSNR is fixed to 18 dB. Furthermore, we also investigate the robustness of our method under the condition of different amplifier spontaneous emission (ASE) noise and nonlinear noise. The proposed method shows great potential for making a remarkable cost reduction in CD monitoring.