Search

In this paper, cellulose nanofibers (CNF) were prepared from wheat bran (WB) and the structure of CNF was determined. Fourier transform infrared spectra and X-ray diffractograms showed the groups such as hydroxyl and carboxyl groups and cellulose type 1 structure possessed by CNF, respectively. Scanning electron microscopy exhibited that CNF was filamentous and intertwined. In addition, Pickering emulsions were prepared using CNF and the physicochemical properties of the emulsions were characterized. The results showed that CNF was able to increase the zeta potential and viscosity of the emulsions, thus improving the stability of the emulsions. Moreover, CNF formed a physical barrier by adsorbing at the oil-water interface and near the oil droplets and CNF dispersed in the aqueous phase formed a network structure to restrict the movement of oil droplets, thus improving the stability of the emulsions. These findings may provide some new insights for the potential applications of WB.

Hyperspectral image (HSI) classification based on deep learning has demonstrated promising performance. In general, using patch-wise samples helps to extract the spatial relationship between pixels and local contextual information. However, the presence of background or other category information in an image patch that is inconsistent with the central target category has a negative effect on classification. To solve this issue, a patch confidence-enhanced transformer (PCET) approach for HSI classification is proposed. To be specific, we design a patch quality assessment (PQA) branch model to evaluate the input patches during training process, which effectively filters out the intrusive non-central pixels. The output confidence of the branch model serves as a quantitative indicator of the contribution degree of the input patch to the overall training efficacy, which is subsequently weighted in the loss function, thereby endowing the model with the capability to dynamically adjust its learning focus based on the qualitative of the inputs. Second, a spectral–spatial multi-feature fusion (SSMF) module is devised to procure scores of representative information simultaneously and fully exploit the potential of multi-scale feature HSI data. Finally, to enhance feature discrimination, global context is efficiently modeled using the efficient additive attention transformer (EA2T) module, which streamlines the attention process and allows the model to learn efficient and robust global representations for accurate classification of the central pixel. A series of experimental results executed on real HSI datasets have substantiated that the proposed PCET can achieve outstanding performance, even when only 10 samples per category are used for training.

Transplant treatment with chlorantraniliprole (CAP) is a proactive approach to protect transplanted plants from pests during early establishment and has been comprehensively applied in tobacco fields in Guangdong Province, China. However, it is not known whether the high dose of CAP in transplant treatments has lethal or sublethal effects on the generalist predator Rhynocoris fuscipes Fabricius (Hemiptera: Reduviidae). To address this concern, the mortalities of R. fuscipes were assessed when 2nd instar larvae of R. fuscipes were in direct contact with or consuming CAP and when their eggs were exposed to CAP. Furthermore, 2nd instar nymphs R. fuscipes were long-term exposed to CAP until they reached adulthood, and their life table parameters were determined. After exposure to CAP, the activity of detoxification enzymes (P450, CaeE and GST) and the functional respond of R. fuscipes to their preys Agrotis ipsilon larvae were determined. In this study, CAP at all concentrations did not significantly increase the mortality of 2nd instar of R. fuscipes nymphs in comparison with the control. The detoxification enzyme (P450, CarE and GST) activities and the number of A. ipsilon larvae consumed by R. fuscipes in the transplant treatment were not affected by CAP after 3-d or long-term exposure. These results indicated that CAP was harmless to R. fuscipes according to IOBC protocols. However, during the treatment of 2nd instar nymphs with a label rate of 15 g AI/ha and a 5× label rate of 75 g AI/ha, CAP significantly prolonged the pre-adult and pre-oviposition periods, and treated adults had lower oviposition. Attention should be given to the time interval between transplant treatment and the release of this biocontrol agent into the field to minimize the impact of CAP on the predator R. fuscipes.

Bacterial soft rot caused by coinfection with Dickeya spp. and Pectobacterium spp. in hosts can cause successive changes in fields, and it is difficult to prevent the spread of and control the infection. Pectobacterium spp. are prevalent in the growing areas of tuberous crops, including taro and potato. Recently, Dickeya fangzhongdai has emerged as a virulent pathogen in taro. To determine the prevalence status of the causal agents and evaluate the potential spreading risks of D. fangzhongdai, screening and taxonomic classification were performed on phytopathogenic bacteria collected from different taro-growing areas in Guangdong Province, China, and biological and genomic characteristics were further compared among typical strains from all defined species. The causative agents were verified to be phytobacterial strains of D. fangzhongdai, Pectobacterium aroidearum and Pectobacterium colocasium. P. aroidearum and P. colocasium were found to form a complex preferring Araceae plants and show intensive genomic differentiation, indicating their ancestor had adapted to taro a long time prior. Compared with Pectobacterium spp., D. fangzhongdai was more virulent to taro corms under conditions of exogenous infection and more adaptable at elevated temperatures. D. fangzhongdai strains isolated from taro possessed genomic components of additional T4SSs, which were accompanied by additional copies of the hcp-vgrG genes of the T6SS, and these contributed to the expansion of their genomes. More gene clusters encoding secondary metabolites were found within the D. fangzhongdai strains than within the Pectobacterium complex; interestingly, distinct gene clusters encoding zeamine and arylpolyene were both most similar to those in D. solani that caused potato soft rot. These comparisons provided genomic evidences for that the newly emerging pathogen was potentially equipped to compete with other pathogens. Diagnostic qPCR verified that D. fangzhongdai was prevalent in most of the taro-growing areas and coexisted with the Pectobacterium complex, while the plants enriching D. fangzhongdai were frequently symptomatic at developing corms and adjacent pseudostems and caused severe symptoms. Thus, the emerging need for intensive monitoring on D. fangzhongdai to prevent it from spreading to other taro-growing areas and to other tuberous crops like potato; the adjustment of control strategies based on different pathopoiesis characteristics is recommended.

As some of the richest sources of natural antioxidants, small berry fruits have attractive colors and special tastes, with recognized benefits for human health. However, sour tastes in small berry juices result in a poor flavor and low acceptance among consumers, greatly limiting their marketability. Among the most commonly used deacidification methods, chemical deacidification methods can neutralize fruit juice via the addition of a deacidification agent, while physical deacidification methods include freezing deacidification, ion-exchange resin deacidification, electrodialysis deacidification, and chitosan deacidification. All of these methods can markedly improve the pH of fruit juice, but they introduce new substances into the juice that may have an influence on its color, taste, and stability. Biological deacidification can effectively remove malic acid from fruit juice, reducing the content from 15 g/L to 3 g/L; additionally, it maintains the taste and stability of the juice. Therefore, it is widely applied for fruit juice deacidification. On this basis, some compound deacidification technologies have also emerged, but they also present problems such as high costs and complicated working procedures. This review of deacidification methods for small berry juice provides a foundation for the industrial development of such juices.

The fall armyworm (FAW), Spodoptera frugiperda, is a serious pest that threatens a range of important crops worldwide. It originated in America and rapidly dispersed throughout Africa and Asia in 2018. There are two subtypes, corn-strain (C-strain) and rice-strain (R-strain), that have different host plant preferences, and the individuals damaging maize in China were identified as C-strain. In the present study, we found FAW individuals damaging rice plants in the field of Guangdong Province, China. FAW larvae and male adults were collected, and the majority of FAWs were characterized as CO I R-strain Tpi C-strain, which is similar to the FAWs damaging maize in China. The FAW adults preferred laying eggs on maize plants more than on rice plants. Compared to those that were fed maize leaves, the FAW larvae were unable to survive when fed 4-week-old rice plants, whereas they could complete their life cycle on 2-week-old rice plants, for which the total survival rate was 8%. The pre-adult- and pupal-stage durations were prolonged, and the fecundity of adult females decreased. Thus, the FAWs found in paddy fields showed better fitness on maize than on rice in the laboratory. Owing to their low survival rate on rice plants, they were unlikely to damage paddy fields in large areas, but populations of FAWs in paddy fields should be monitored.

In this study, 11 varieties of potatoes were selected as raw materials to process whole powder under the same conditions. The quality differences of snowflake whole powder prepared by different varieties of potatoes were studied by measuring components, iodine blue value, chroma, water holding capacity (WHC), oil holding capacity (OHC), freeze-thaw stability and other indicators, so as to clarify the influence of different varieties of potatoes on the quality of whole powder. The results showed that the Longshu 6 had the strongest WHC (9.84 g/g), OHC (1.63 g/g) and coagulation capacity, and the attenuation value (315.2 RVU) was significantlyhigher than that of the other varieties (P

The choice of appropriate proteases and pretreatment methods significantly influences the preparation of bioactive peptides. This study aimed to investigate the effects of different pretreatment methods on the hydrolytic performance of diverse proteases during the production of dipeptidyl peptidase-IV (DPP-IV) inhibitory peptides derived from wheat and their foaming and emulsion properties. Dry heating, aqueous heating, and ultrasound treatment were employed as pretreatments for the protein prior to the enzymatic hydrolysis of wheat gluten. FTIR analysis results indicated that all pretreatment methods altered the secondary structure of the protein; however, the effects of dry heating treatment on the secondary structure content were opposite to those of aqueous heating and ultrasound treatment. Nevertheless, all three methods enhanced the protein solubility and surface hydrophobicity. By using pretreated proteins as substrates, five different types of proteases were employed for DPP-IV inhibitory peptide production. The analysis of the DPP-IV inhibitory activity, degree of hydrolysis, and TCA-soluble peptide content revealed that the specific pretreatments had a promoting or inhibiting effect on DPP-IV inhibitory peptide production depending on the protease used. Furthermore, the pretreatment method and the selected type of protease collectively influenced the foaming and emulsifying properties of the prepared peptides.

The study aimed to investigate the effect of germination conditions on the content of polyphenol extract in mung bean and to further investigate the effect of polyphenol extract in germinated mung bean on diabetic mice. Through single factor experiment and response surface experiment, the effects of soaking temperature, soaking time, germination temperature, germination time and soaking liquid CaCl2 concentration on the polyphenol content of mung bean were analyzed. The optimal germination conditions of mung bean were determined as soaking temperature 25°C, soaking time 11 h, germination temperature 28°C, germination time 3 days and CaCl2 concentration 2 mM. Under these conditions, the content of polyphenol extract in germinated mung bean was 4.878 ± 0.30 mg/g, which was 3.07 times higher than that in ungerminated mung bean. The structure and content of purified polyphenols in germinated mung bean were determined by HPLC-MS/MS. Quinic acid, Quercetin, Rutin, Vitexin, Isovitexin and other substances were identified, and the content of polyphenols was 65.19%. In addition, through the in vivo and in vitro hypoglycemic activity experimental study of germinated mung bean polyphenols extract, the results showed that germinated mung bean polyphenols had an in vitro inhibitory effect on α-glucosidase, IC50 was 44.45 mg/ml. In vitro inhibitory activity was stronger after digestion. Polyphenol extract can significantly reduce blood sugar and improve insulin resistance in Type 2 diabetic mice (T2DM). According to the results, germination treatment is an effective way to increase the content of polyphenols in mung bean, and the polyphenols extract has hypoglycemic activity.

Rapid and accurate acquisition of glacier regional changes is of great significance to the study of glaciers. Among all satellite images, Synthetic Aperture Radar (SAR) data has a great advantage in monitoring the glaciers in harsh weather conditions. Conventionally, glacier boundaries are manually delineated on images. However, this is a time-consuming process, especially in the batch process of large-area data. In this paper, we propose a Multiscale Joint Deep Neural Network (MJ-DNN) for large-scale glaciers contour extraction using single-polarimetric SAR intensity images. Based on U-Net, the proposed method has been improved in three aspects. First, Atrous Separable Convolution is used instead of convolution with the down-sampling part. Second, we propose a multiscale joint convolution layer to obtain information at multiple scales. Third, we deepen the network with the residual connection structure for higher-level features. At the final layer, we optimize the network result with the conditional random field method. To validate our approach, we test it on three glaciers and we compare the segmentation results of four different methods in parallel. The results show that the intersection over the union of the proposed method is the most efficient.

The number of labeled samples has a great impact on the classification results of a very-high-resolution (VHR) remote sensing image. However, the acquisition of available labeled samples is difficult and time consuming. Faced with the limited labeled samples on a high-resolution remote sensing image, a semisupervised method becomes an effective way. In semisupervised learning, an accurate similarity prediction between unlabeled and labeled samples is very important. However, a reliable similarity prediction between high-dimensional features is difficult. For more reliable similarity prediction for the high-dimensional feature, a novel semisupervised classification framework via improved metric learning with a convolutional neural network is proposed. In the proposed method, a novel trainable metric learning network is designed to accurately evaluate the similarity between high-dimensional features. The vector distance parameter solving problem is transformed into a neural network design problem, which can automatically calculate parameters by the back-propagation algorithm. Finally, the pixel constraint mechanism is introduced to select the unlabeled samples. Experimental results conducted on three VHR remote sensing images, including Aerial, Xi'an, and Pavia University, and the results present that the proposed method performs better than the compared state-of-the-art methods.

Bacterial soft rot of banana, caused by Dickeya zeae, is spreading rapidly in important banana growing areas in China and seriously threatens banana production. In this study, we sequenced the high-quality complete genomes of three typical banana strains, MS1 (size: 4,831,702-bp; genome coverages: 538x), MS_2014 (size: 4,740,000-bp; genome coverages: 586x) and MS_2018 (size: 4,787,201-bp; genome coverages: 583x), isolated in 2009, 2014, and 2018, respectively. To determine their genomic and phenotypic diversity with respect to their hosts of origin, they were compared with other D. zeae strains, including another representative banana strain MS2 from China. The sequenced strains were similar in utilization of carbon source and chemical substrates, and general genomic features of GC content, and tRNA and rRNA regions. They were also conserved in most virulence determinants, including gene-encoding secretion systems, plant cell wall degrading enzymes, and exopolysaccharides. We further explored their genomic diversity in the predicted genomic islands (GIs). These GIs were rich in integrases and transposases, where some genomic dissimilarity was observed in the flagellar gene cluster and several secondary metabolite gene clusters. Different constituents of core biosynthetic modules were found within the bacteriocin and aryl polyene (APE) pigment gene clusters, and the strains from banana showed different phenotypes with respect to antibiosis effects and colony pigmentation. Additionally, clustered regularly interspaced short palindromic repeat (CRISPR) and prophage elements, such as type I-F and III-A CRISPR arrays and an intact prophage of MS1-P5, contributed to bacterial diversity. Phylogenetic tree analysis and genome-genome nucleotide comparison confirmed the genomic divergence among the strains isolated from banana. Considering these characteristics, MS2 and MS_2014 probably diverged later than MS1, while MS_2018 was different and more similar to foreign strains isolated from other hosts in several characteristics. Strain MS_2018 caused severe symptoms on banana varieties previously considered moderately resistant or moderately susceptible, including varieties of Cavendish (Musa AAA) and Plantain (Musa ABB). Our study of genomic and phenotypic diversity raises public attention to the risk of spreading new pathogenic variants within banana growing regions and supports development of predictive strategies for disease control.

Accurate land-use classification is essential for the management and supervision of urban development, land resources and environment sustainability. Feature extractor and classifier are the important modules of land-use classification. Deep convolutional neural network has been proved to be able to learn more robust and discriminative features from images. In this paper, we increase the diversity and discriminative of features by fusing features extracted by three deep convolutional neural networks with different architectures, which are obtained by fine-tuning the pre-trained models with land-use image dataset. In order to make the classification faster and have excellent generalization performance, we select constrained extreme learning machine instead of fully connected layer or support vector machine. Experimental results show that the proposed method can achieve a better performance with the overall classification accuracy of 98.35%, compared with other state-of-the-art methods.

Because land covers always have different scales, multiscale methods are widely used in very-high-resolution (VHR) remote sensing image classification. Traditional multiscale methods usually capture multiscale information by using rectangular windows of different sizes. Each scale contains the same number of training samples and is independently trained. Hence, the training process is time-consuming. In this article, a novel convolutional neural network with a class-driven loss (CNNs-CDL) model is proposed for multiscale VHR remote sensing image classification. First, a multiscale sample construction method is proposed to select a training sample and capture the relationships among different scale samples. The lowest-scale samples are selected on the lowest-resolution image and are mapped to the higher-resolution image without additional label information. Then, a CNN with class-driven loss is trained with the lowest-scale training samples. Class-driven loss can effectively learn the spatial dependence between the nonadjacent samples to promote classification accuracy. Finally, the CNN model is fine-tuned with the higher-scale samples. Although the number of higher-scale training samples increases, the fine-tuning process requires only a small number of iterations to converge. Hence, the proposed model can effectively reduce the training time. The experimental results for three VHR remote sensing images show that the proposed method performs better than several recently proposed methods.

Synthetic Aperture Radar technology is highly convenient for monitoring the glacier surface motion in unfavorable areas due to its advantages of being independent of time and weather conditions. A novel glacier motion monitoring method based on the deep matching network (DMN) is proposed in this paper. The network learns the relationship between the glacier SAR image patch-pairs and the corresponding matching labels in an end-to-end manner. Unlike conventional methods that utilize shallow feature tracking, the DMN performs a similarity measurement of deep features, which comprises feature extraction and a metric network. Feature extraction adopts the framework of a Siamese neural network to improve the training efficiency and dense connection blocks to increase the feature utilization. In addition, a self-sample learning method is introduced to generate training samples with matching labels. The experiments are performed on simulated SAR images and real SAR intensity images of the Taku Glacier and the Yanong Glacier, respectively. The results confirm the superiority of the DMN presented in the paper over other methods, even in case of strong noise. Furthermore, a quantitative 2D velocity field of real glaciers is obtained to provide reliable support for high-precision, long-term and large-scale automatic glacier motion monitoring.

Adzuki beans are grown in several countries around the world and are widely popular in Asia, where they are often prepared in various food forms. Adzuki beans are rich in starch, and their proteins contain a balanced variety of amino acids with high lysine content, making up for the lack of protein content of cereals in the daily diet. Therefore, the research on adzuki beans and the development of their products have broad prospects for development. The starch, protein, fat, polysaccharide, and polyphenol contents and compositions of adzuki beans vary greatly among different varieties. The processing characteristic components of adzuki beans, such as starch, isolated protein, and heated flavor, are reported with a view to further promote the processing and development of adzuki bean foods. In addition to favorable edibility, the human health benefits of adzuki beans include antioxidant, antibacterial, and anti-inflammatory properties. Furtherly, adzuki beans and extracts have positive effects on the prevention and treatment of diseases, including diabetes, diabetes-induced kidney disease or kidney damage, obesity, and high-fat-induced cognitive decline. This also makes a case for the dual use of adzuki beans for food and medicine and contributes to the promotion of adzuki beans as a healthy, edible legume.

Abstract Background Dickeya sp. strain PA1 is the causal agent of bacterial soft rot in Phalaenopsis, an important indoor orchid in China. PA1 and a few other strains were grouped into a novel species, Dickeya fangzhongdai, and only the orchid-associated strains have been shown to cause soft rot symptoms. Methods We constructed the complete PA1 genome sequence and used comparative genomics to explore the differences in genomic features between D. fangzhongdai and other Dickeya species. Results PA1 has a 4,979,223-bp circular genome with 4269 predicted protein-coding genes. D. fangzhongdai was phylogenetically similar to Dickeya solani and Dickeya dadantii. The type I to type VI secretion systems (T1SS–T6SS), except for the stt-type T2SS, were identified in D. fangzhongdai. The three phylogenetically similar species varied significantly in terms of their T5SSs and T6SSs, as did the different D. fangzhongdai strains. Genomic island (GI) prediction and synteny analysis (compared to D. fangzhongdai strains) of PA1 also indicated the presence of T5SSs and T6SSs in strain-specific regions. Two typical CRISPR arrays were identified in D. fangzhongdai and in most other Dickeya species, except for D. solani. CRISPR-1 was present in all of these Dickeya species, while the presence of CRISPR-2 varied due to species differentiation. A large polyketide/nonribosomal peptide (PK/NRP) cluster, similar to the zeamine biosynthetic gene cluster in Dickeya zeae rice strains, was discovered in D. fangzhongdai and D. solani. The D. fangzhongdai and D. solani strains might recently have acquired this gene cluster by horizontal gene transfer (HGT). Conclusions Orchid-associated strains are the typical members of D. fangzhongdai. Genomic analysis of PA1 suggested that this strain presents the genomic characteristics of this novel species. Considering the absence of the stt-type T2SS, the presence of CRISPR loci and the zeamine biosynthetic gene cluster, D. fangzhongdai is likely a transitional form between D. dadantii and D. solani. This is supported by the later acquisition of the zeamine cluster and the loss of CRISPR arrays by D. solani. Comparisons of phylogenetic positions and virulence determinants could be helpful for the effective quarantine and control of this emerging species.

Baked adzuki beans are rich in tantalizing odor and nutritional components, such as protein, dietary fiber, vitamin B, and minerals. To analyze the final quality of baked beans, the acrylamide and volatile formation of adzuki beans were investigated under the conditions of microwave baking and drum roasting. The results indicate that the acrylamide formation in baked adzuki beans obeys the exponential growth function during the baking process, where a rapid increase in acrylamide content occurs at a critical temperature and low moisture content. The critical temperature that leads to a sudden increase in acrylamide content is 116.5 °C for the moisture content of 5.6% (w.b.) in microwave baking and 91.6 °C for the moisture content of 6.1% (w.b.) in drum roasting. The microwave-baked adzuki beans had a higher formation of the kinetics of acrylamide than that of drum-roasted beans due to the microwave volumetric heating mode. The acrylamide content in baked adzuki beans had a significant correlation with their color due to the Maillard reaction. A color difference of 11.1 and 3.6 may be introduced to evaluate the starting point of the increase in acrylamide content under microwave baking and drum roasting, respectively. Heating processes, including microwave baking and drum roasting, for adzuki beans generate characteristic volatile compounds such as furan, pyrazine, ketone, alcohols, aldehydes, esters, pyrroles, sulfocompound, phenols, and pyridine. Regarding flavor formation, beans baked via drum roasting showed better flavor quality than microwave-baked beans.

Multi-sensor image can provide supplementary information, usually leading to better performance in classification tasks. However, the general deep neural network-based multi-sensor classification method learns each sensor image separately, followed by a stacked concentrate for feature fusion. This way requires a large time cost for network training, and insufficient feature fusion may cause. Considering efficient multi-sensor feature extraction and fusion with a lightweight network, this paper proposes an attention-guided classification method (AGCNet), especially for multispectral (MS) and panchromatic (PAN) image classification. In the proposed method, a share-split network (SSNet) including a shared branch and multiple split branches performs feature extraction for each sensor image, where the shared branch learns basis features of MS and PAN images with fewer learn-able parameters, and the split branch extracts the privileged features of each sensor image via multiple task-specific attention units. Furthermore, a selective classification network (SCNet) with a selective kernel unit is used for adaptive feature fusion. The proposed AGCNet can be trained by an end-to-end fashion without manual intervention. The experimental results are reported on four MS and PAN datasets, and compared with state-of-the-art methods. The classification maps and accuracies show the superiority of the proposed AGCNet model.

Rice foot rot caused by Dickeya zeae is an important bacterial disease of rice worldwide. In this study, we identified a new strain EC2 from rice in Guangdong province, China. This strain differed from the previously identified strain from rice in its biochemical characteristics, pathogenicity, and genomic constituents. To explore genomic discrepancies between EC2 and previously identified strains from rice, a complete genome sequence of EC2 was obtained and used for comparative genomic analyses. The complete genome sequence of EC2 is 4,575,125 bp in length. EC2 was phylogenetically closest to previously identified Dickeya strains from rice, but not within their subgroup. In terms of secretion systems, genomic comparisons revealed that EC2 harbored only type I (T1SS), typeⅡ (T2SS), and type VI (T6SS) secretion systems. The flagella cluster of this strain possessed specific genomic characteristics like other D. zeae strains from Guangdong and from rice; within this locus, the genetic diversity among strains from rice was much lower than that of within strains from non-rice hosts. Unlike other strains from rice, EC2 lost the zeamine cluster, but retained the clustered regularly interspaced short palindromic repeats-1 (CRISPR-1) array. Compared to the other D. zeae strains containing both exopolysaccharide (EPS) and capsular polysaccharide (CPS) clusters, EC2 harbored only the CPS cluster, while the other strains from rice carried only the EPS cluster. Furthermore, we found strain MS1 from banana, carrying both EPS and CPS clusters, produced significantly more EPS than the strains from rice, and exhibited different biofilm-associated phenotypes. Comparative genomics analyses suggest EC2 likely evolved through a pathway different from the other D. zeae strains from rice, producing a new type of rice foot rot pathogen. These findings emphasize the emergence of a new type of D. zeae strain causing rice foot rot, an essential step in the early prevention of this rice bacterial disease.

To explore better methods of natural protein modification for black soybean, comparisons among the effects of different modified methods on structural changes of the modified products of black soybean protein isolate (BSPI) were carried out in this study. The modified products used in this study included enzymatic crossing-link black soybean protein isolate (ECBSPI), wet heating treatment glycosylation black soybean protein isolate (WHTGBSPI) and especially enzymatic glycosylation black soybean protein isolate catalyzed by transglutaminase (EGBSPI). The effects of the modification methods on structural changes were analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), amino acid content and circular dichroism (CD) analysis. Moreover, the processing properties changes caused by structural changes of BSPI were detected by thermogravimetric analysis, particle size analysis, zeta-potential, surface hydrophobicity, solubility, emulsification, gelation, and rheological properties. The results show that the modified BSPI products were protein polymers, and among them, EGBSP and WHTGBSPI are covalently bonded glycation products. Products modified by Maillard reactions and transglutaminase (TG) display partly destroyed α-helix and β-sheet structures that form more open secondary BSPI structures. For ECBSPI, the proportion of irregular crimp structure reduces to form a high order secondary structure. All the modified products form fine aggregations in dispersion, except WHTGBSPI has most negative zeta-potential and least molecular stability due to the hydrophobic amino acids embedded in the protein molecules. The zeta-potentials of BSPI, ECBSPI, WHTGBSPI and EGBSPI are respectively −21.5, −23.8, −18.1 and −20.2 mV. The surface hydrophobicity of EGBSPI (5.07 ± 0.07) and WHTGBSPI (7.02 ± 0.05) decrease, while the surface hydrophobicity of ECBSPI (19.5 ± 0.06) increases. The solubility and rheological properties of EGBSPI, ECBSPI and WHTGBSPI after modification are all better than those of BSPI, especially EGBSPI. Emulsification of EGBSPI and WHTGBSPI increase (by 24.5% and 12.2%, respectively) while ECBSPI decrease (by 17.0), and there is similar emulsion stability trend. Moreover, the properties of ECBSPI increase except cohesiveness compared to BSPI. In conclusion, as a safe and efficient method for natural protein modification, enzymatic glycosylation catalyzed by TG has great potential in improving food processing characteristics.