Search

Background: Notoginseng stem-leaf (NGL) ginsenosides have not been well used. To improve their utilization, the biotransformation of NGL ginsenosides was studied using ginsenosidase type-I from Aspergillus niger g.848. Methods: NGL ginsenosides were reacted with a crude enzyme in the RAT-5D bioreactor, and the dynamic changes of multi-ginsenosides of NGL were recognized by HPLC. The reaction products were separated using a silica gel column and identified by HPLC and NMR. Results: All the NGL ginsenosides are protopanaxadiol-type ginsenosides; the main ginsenoside contents are 27.1% Rb3, 15.7% C-Mx1, 13.8% Rc, 11.1% Fc, 7.10% Fa, 6.44% C-Mc, 5.08% Rb2, and 4.31% Rb1. In the reaction of NGL ginsenosides with crude enzyme, the main reaction of Rb3 and C-Mx1 occurred through Rb3→C-Mx1→C-Mx; when reacted for 1 h, Rb3 decreased from 27.1% to 9.82 %, C-Mx1 increased from 15.5% to 32.3%, C-Mx was produced to 6.46%, finally into C-Mx and a small amount of C-K. When reacted for 1.5 h, all the Rb1, Rd, and Gyp17 were completely reacted, and the reaction intermediate F2 was produced to 8.25%, finally into C-K. The main reaction of Rc (13.8%) occurred through Rc→C-Mc1→C-Mc→C-K. The enzyme barely hydrolyzed the terminal xyloside on 3-O or 20-O-sugar-moiety of the substrate; therefore, 9.43 g C-Mx, 6.85 g C-K, 4.50 g R7, and 4.71 g Fc (hardly separating from the substrate) were obtained from 50 g NGL ginsenosides by the crude enzyme reaction. Conclusion: Four monomer ginsenosides were successfully produced and separated from NGL ginsenosides by the enzyme reaction. Keywords: dynamic changes, enzyme reaction, ginsenosidase type-I, notoginseng, notoginseng stem-leaf ginsenosides

In recent years, semantic segmentation methods based on fully convolutional networks (FCNs) have achieved great success. However, these methods tend to produce inconsistency and isolated class labels, mainly because the end-to-end mapping of FCN essentially treats each pixel independently. As a post-processing approach, the conditional random field (CRF) has been widely used to alleviate this problem. However, the inference of CRF is usually very time-consuming in computation. To solve this problem, we present a new method, called locally shared features (LSFs), to model local dependence between pixels. The LSF encourages adjacent pixels to have similar features by making them share certain properties with each other. This is achieved by concatenating features around a pixel, including the pixel itself. Our experimental results indicate—the LSF approach delivers comparable or better performance than the CRF method with respect to the accuracy and local smoothness in segmentation output, while obtaining a significant gain in computational efficiency.

Background: Ginsenoside F1, a pharmaceutical component of ginseng, is known to have antiaging, antioxidant, anticancer, and keratinocyte protective effects. However, the usage of ginsenoside F1 is restricted owing to the small amount found in Korean ginseng. Methods: To enhance the production of ginsenoside F1 as a 10 g unit with high specificity, yield, and purity, an enzymatic bioconversion method was developed to adopt the commercial enzyme Cellulase KN from Aspergillus niger with food grade, which has ginsenoside-transforming ability. The proposed optimum reaction conditions of Cellulase KN were pH 5.0 and 50°C. Results: Cellulase KN could effectively transform the ginsenosides Re and Rg1 into F1. A scaled-up biotransformation reaction was performed in a 10 L jar fermenter at pH 5.0 and 50°C for 48 h with protopanaxatriol-type ginsenoside mixture (at a concentration of 10 mg/mL) from ginseng roots. Finally, 13.0 g of F1 was produced from 50 g of protopanaxatriol-type ginsenoside mixture with 91.5 ± 1.1% chromatographic purity. Conclusion: The results suggest that this enzymatic method could be exploited usefully for the preparation of ginsenoside F1 to be used in cosmetic, functional food, and pharmaceutical industries.

Inspired by the human 3D visual perception system, we present an obstacle detection and classification method based on the use of Time-of-Flight (ToF) cameras for robotic navigation in unstructured environments. The ToF camera provides 3D sensing by capturing an image along with per-pixel 3D space information. Based on this valuable feature and human knowledge of navigation, the proposed method first removes irrelevant regions which do not affect robot’s movement from the scene. In the second step, regions of interest are detected and clustered as possible obstacles using both 3D information and intensity image obtained by the ToF camera. Consequently, a multiple relevance vector machine (RVM) classifier is designed to classify obstacles into four possible classes based on the terrain traversability and geometrical features of the obstacles. Finally, experimental results in various unstructured environments are presented to verify the robustness and performance of the proposed approach. We have found that, compared with the existing obstacle recognition methods, the new approach is more accurate and efficient.

Camera calibration plays a critical role in 3D computer vision tasks. The most commonly used calibration method utilizes a planar checkerboard and can be done nearly fully automatically. However, it requires the user to move either the camera or the checkerboard during the capture step. This manual operation is time consuming and makes the calibration results unstable. In order to solve the above problems caused by manual operation, this paper presents a full-automatic camera calibration method using a virtual pattern instead of a physical one. The virtual pattern is actively transformed and displayed on a screen so that the control points of the pattern can be uniformly observed in the camera view. The proposed method estimates the camera parameters from point correspondences between 2D image points and the virtual pattern. The camera and the screen are fixed during the whole process; therefore, the proposed method does not require any manual operations. Performance of the proposed method is evaluated through experiments on both synthetic and real data. Experimental results show that the proposed method can achieve stable results and its accuracy is comparable to the standard method by Zhang.