Automatic freezing-tolerant rapeseed material recognition using UAV images and deep learning

Background Freezing injury is a devastating yet common damage that occurs to winter rapeseed during the overwintering period which directly reduces the yield and causes heavy economic loss. Thus, it is an important and urgent task for crop breeders to find the freezing-tolerant rapeseed materials in the process of breeding. Existing large-scale freezing-tolerant rapeseed material recognition methods mainly rely on the field investigation conducted by the agricultural experts using some professional equipments. These methods are time-consuming, inefficient and laborious. In addition, the accuracy of these traditional methods depends heavily on the knowledge and experience of the experts. Methods To solve these problems of existing methods, we propose a low-cost freezing-tolerant rapeseed material recognition approach using deep learning and unmanned aerial vehicle (UAV) images captured by a consumer UAV. We formulate the problem of freezing-tolerant material recognition as a binary classification problem, which can be solved well using deep learning. The proposed method can automatically and efficiently recognize the freezing-tolerant rapeseed materials from a large number of crop candidates. To train the deep learning network, we first manually construct the real dataset using the UAV images of rapeseed materials captured by the DJI Phantom 4 Pro V2.0. Then, five classic deep learning networks (AlexNet, VGGNet16, ResNet18, ResNet50 and GoogLeNet) are selected to perform the freezing-tolerant rapeseed material recognition. Result and conclusion The accuracy of the five deep learning networks used in our work is all over 92%. Especially, ResNet50 provides the best accuracy (93.33$$\%$$ % ) in this task. In addition, we also compare deep learning networks with traditional machine learning methods. The comparison results show that the deep learning-based methods significantly outperform the traditional machine learning-based methods in our task. The experimental results show that it is feasible to recognize the freezing-tolerant rapeseed using UAV images and deep learning.

A method for obtaining field wheat freezing injury phenotype based on RGB camera and software control

Background Low temperature freezing stress has adverse effects on wheat seedling growth and final yield. The traditional method to evaluate the wheat injury caused by the freezing stress is by visual observations, which is time-consuming and laborious. Therefore, a more efficient and accurate method for freezing damage identification is urgently needed. Results A high-throughput phenotyping system was developed in this paper, namely, RGB freezing injury system, to effectively and efficiently quantify the wheat freezing injury in the field environments. The system is able to automatically collect, processing, and analyze the wheat images collected using a mobile phenotype cabin in the field conditions. A data management system was also developed to store and manage the original images and the calculated phenotypic data in the system. In this experiment, a total of 128 wheat varieties were planted, three nitrogen concentrations were applied and two biological and technical replicates were performed. And wheat canopy images were collected at the seedling pulling stage and three image features were extracted for each wheat samples, including ExG, ExR and ExV. We compared different test parameters and found that the coverage had a greater impact on freezing injury. Therefore, we preliminarily divided four grades of freezing injury according to the test results to evaluate the freezing injury of different varieties of wheat at the seedling stage. Conclusions The automatic phenotypic analysis method of freezing injury provides an alternative solution for high-throughput freezing damage analysis of field crops and it can be used to quantify freezing stress and has guiding significance for accelerating the selection of wheat excellent frost resistance genotypes.

Hyperbaric Oxygen Therapy with Iloprost Improves Digit Salvage in Severe Frostbite Compared to Iloprost Alone

Background and Objectives: Frostbite is a freezing injury that can lead to amputation. Current treatments include tissue rewarming followed by thrombolytic or vasodilators. Hyperbaric oxygen (HBO) therapy might decrease the rate of amputation by increasing cellular oxygen availability to the damaged tissues. The SOS-Frostbite study was implemented in a cross-border program among the hyperbaric centers of Geneva, Lyon, and the Mont-Blanc hospitals. The objective was to assess the efficacy of HBO + iloprost among patients with severe frostbite. Materials and Methods: We conducted a multicenter prospective single-arm study from 2013 to 2019. All patients received early HBO in addition to standard care with iloprost. Outcomes were compared to a historical cohort in which all patients received iloprost alone between 2000 and 2012. Inclusion criteria were stage 3 or 4 frostbite and initiation of medical care <72 h from frostbite injury. Outcomes were the number of preserved segments and the rate of amputated segments. Results: Thirty patients from the historical cohort were eligible and satisfied the inclusion criteria, and 28 patients were prospectively included. The number of preserved segments per patient was significantly higher in the prospective cohort (mean 13 ± SD, 10) compared to the historical group (6 ± 5, p = 0.006); the odds ratio was significantly higher by 45-fold (95%CI: 6-335, p < 0.001) in the prospective cohort compared to the historical cohort after adjustment for age and delay between signs of freezing and treatment start. Conclusions: This study demonstrates that the combination of HBO and iloprost was associated with higher benefit in patients with severe frostbite. The number of preserved segments was two-fold higher in the prospective cohort compared to the historical group (mean of 13 preserved segments vs. 6), and the reduction of amputation was greater in patients treated by HBO + iloprost compared with the iloprost only.

Physiological and biochemical responses of strawberry crown and leaf tissues to freezing stress

Background In northern Iran and other cold regions, winter freezing injury and resultant yield instability are major limitations to strawberry production. However, there is scarcity of information on the physiological and biochemical responses of strawberry cultivars to freezing stress. This study aimed to investigate the physiological and biochemical responses of strawberry cultivars (Tennessee Beauty, Blakemore, Kurdistan, Queen Elisa, Chandler, Krasnyy Bereg, and Yalova) to different freezing temperature treatments (− 5, − 10, − 15, − 20, and − 25 °C) under controlled conditions. Results All measured physiological and biochemical features were significantly affected by the interaction effect between low temperatures and cultivars. Tennessee Beauty showed the highest RWC at − 25 °C. The highest Fv/Fm was observed in Queen Elisa. Krasnyy Bereg had the least freezing injury (FI) in crown and leaf, while Yalova and Chandler showed the highest crown and leaf FI, respectively. At − 20 to − 25 °C, the highest carbohydrates contents of crown and leaf were noted in Blakemore and Krasnyy Bereg cultivars, respectively. The Yalova showed the highest protein content in both crown and leaf tissues at − 25 °C. The Tennessee Beauty and Blackmore cultivars showed the highest proline in crowns and leaves at − 15 °C, respectively. The highest ThioBarbituric Acid Reactive Substances (TBARS) contents in the crown and leaf were observed in Kurdistan and Queen Elisa, respectively. Queen Elisa and Krasnyy Bereg cultivars showed SOD and POD peaks in the crown at − 15 °C, respectively. Conclusion Freezing stress was characterized by decreased Fv/Fm and RWC, and increased FI, TBARS, total carbohydrates, total proteins, proline content, and antioxidant enzyme activity. The extent of changes in above mentioned traits was cultivar dependent. FI and TBARS were the best traits among destructive parameters for evaluating freezing tolerance. Moreover, maximum quantum yield of PSII (Fv/Fm index), as non-destructive parameters, showed a significant efficiency in rapid assessment for screening of freezing tolerant strawberry cultivars. The cultivars Krasnyy Bereg, Queen Elisa, and Kurdistan were the most tolerant cultivars to freezing stress. These cultivars can be used as parents in breeding programs to develop new freezing tolerant cultivars.

Identification and Regionalization of Cold Resistance of Wine Grape Germplasms (V. vinifera)

With the extreme changes of the global climate, winter freezing injury has become an important limiting factor for the development of the global grape industry. Therefore, there is a significant need for the screening of cold-resistant wine grape germplasms and cold regionalization for cold-resistant breeding and the development of grapevine cultivation in cold regions. In this study, the low-temperature half-lethal temperature (LT50) values were determined for the annual dormant branches of 124 wine grape germplasms (V. vinifera) to evaluate their cold resistance. The LT50 values of the 124 tested germplasms ranged from −22.01 °C to −13.18 °C, with six cold-resistant germplasms below −20 °C. Based on the LT50 values, the 124 germplasms were clustered into four types, with cold resistance from strong to weak in the order of type Ⅱ > type Ⅰ > type Ⅳ > type Ⅲ, corresponding to the four cold hardiness zones. Zones 1, 2, 3, and 4 included 6, 22, 68, and 28 germplasms, respectively, with decreasing cold resistance. The number of germplasms in different hardiness zones followed a normal distribution, with the most in zone 3. In Type Ⅱ, the fruit skin color of germplasms was positively correlated with cold hardiness, while the temperature of origin was negatively correlated with cold hardiness. The average LT50 of germplasms in different origin regions ranged from −17.44 °C to −16.26 °C, with differences among some regions. The cold regionalization analysis resulted in the distribution of 124 germplasms in four temperature regions in China with six germplasms in region A (−22 °C ≤ LT50 ≤ −20 °C), 30 germplasms in region B (−20°C ≤ LT50 ≤ −18°C), 71 germplasms in region C (−18 °C ≤ LT50 ≤ −15 °C), and 17 germplasms in region D (−15 °C ≤ LT50 ≤ −13 °C). Strong cold-resistant wine grape germplasms (V. vinifera) were identified, and these could be used as parental material for cold-resistant breeding. In some areas in China, soil-burial over-wintering strategies are used, but our results suggest that some wine grapes could be cultivated without requiring winter burial during overwintering. The results of this study should provide guidance for the selection of promising strains for cold-resistant breeding for expanded cultivation of improved varieties for wine grape production in China.

Automatic Freezing-Tolerant Rapeseed Material Recognition Using UAV Images and Deep Learning

Background: Freezing injury is a serious and common damage that occurs to winter rapeseed during the overwintering period. The freezing injury directly reduces the rapeseed yield and causes serious economic loss. Thus, it is an important and urgent task for crop breeders to find the freezing-tolerant rapeseed materials in the process of breeding. Existing large-scale freezing-tolerant rapeseed material recognition methods mainly rely on the field investigation conducted by the agricultural experts using some professional equipment. These methods are time-consuming, inefficient and laborious. In addition, the accuracy of these traditional methods depends heavily on the knowledge and experience of experts. Methods: To solve these problems of existing methods, we propose a low-cost freezing-tolerant rapeseed material recognition approach using deep learning technology and unmanned aerial vehicle (UAV) images captured by a consumer drone. We formulate the problem of freezing-tolerant material recognition as a binary classification problem, which can be solved well using deep learning technology. The proposed method can automatically and efficiently recognize the freezing-tolerant rapeseed materials from a large number of candidates. To train the deep learning network, we first manually construct the real dataset using the UAV images of rapeseed materials collected by the Phantom 4 Pro. Then, five classic deep learning networks (AlexNet, VGGNet16, ResNet18, ResNet50 and GoogLeNet) are selected to perform the freezing-tolerant rapeseed material recognition. Result and Conclusion: The accuracy of the five deep learning networks used in our work is all over 92%. Especially, ResNet50 provides the best accuracy (93.33%) in this task. In addition, we also compare deep learning networks with traditional machine learning methods. The comparison results show that the deep learning-based approach significantly outperforms the traditional machine learning-based methods in our task. The experimental results show that it is feasible to recognize the freezing-tolerant rapeseed using UAV images and deep learning.

Loading equine oocytes with cryoprotective agents captured with a finite element method model

Cryopreservation can be used to store equine oocytes for extended periods so that they can be used in artificial reproduction technologies at a desired time point. It requires use of cryoprotective agents (CPAs) to protect the oocytes against freezing injury. The intracellular introduction of CPAs, however, may cause irreversible osmotic damage. The response of cells exposed to CPA solutions is governed by the permeability of the cellular membrane towards water and the CPAs. In this study, a mathematical mass transport model describing the permeation of water and CPAs across an oocyte membrane was used to simulate oocyte volume responses and concomitant intracellular CPA concentrations during the exposure of oocytes to CPA solutions. The results of the analytical simulations were subsequently used to develop a phenomenological finite element method (FEM) continuum model to capture the response of oocytes exposed to CPA solutions with spatial information. FEM simulations were used to depict spatial differences in CPA concentration during CPA permeation, namely at locations near the membrane surface and towards the middle of the cell, and to capture corresponding changes in deformation and hydrostatic pressure. FEM simulations of the multiple processes occurring during CPA loading of oocytes are a valuable tool to increase our understanding of the mechanisms underlying cryopreservation outcome.

Risk assessment of freezing injury during overwintering of wheat in the northern boundary of the Winter Wheat Region in China

Freezing injury is one of the main restriction factors for winter wheat production, especially in the northern part of the Winter Wheat Region in China. It is very important to assess the risk of winter wheat-freezing injury. However, most of the existing climate models are complex and cannot be widely used. In this study, Zunhua which is located in the northern boundary of Winter Wheat Region in China is selected as research region, based on the winter meteorological data of Zunhua from 1956 to 2016, seven freezing disaster-causing factors related to freezing injury were extracted to formulated the freezing injury index (FII) of wheat. Referring to the historical wheat-freezing injury in Zunhua and combining with the cold resistance identification data of the National Winter Wheat Variety Regional Test (NWWVRT), consistency between the FII and the actual freezing injury situation was tested. Furthermore, the occurrence law of freezing injury in Zunhua during the past 60 years was analyzed by Morlet wavelet analyze, and the risk of freezing injury in the short term was evaluated. Results showed that the FII can reflect the occurrence of winter wheat-freezing injury in Zunhua to a certain extent and had a significant linear correlation with the dead tiller rate of wheat (P = 0.014). The interannual variation of the FII in Zunhua also showed a significant downward trend (R2 = 0.7412). There are two cycles of freezing injury in 60 years, and it showed that there’s still exist a high risk in the short term. This study provides reference information for the rational use of meteorological data for winter wheat-freezing injury risk assessment.

A Method for Obtaining Field Wheat Freezing Injury Phenotype Based on RGB Camera and Software Control

Low temperature freezing stress has adverse effects on wheat seedling growth and final yield. The traditional method to evaluate the wheat injury caused by the freezing stress is by visual observations, which is time-consuming and laborious. Therefore, to effectively and efficiently quantify the wheat freezing injury in the field environments, a high-throughput phenotyping system was developed in this paper , namely, RGB FREEZING INJURY SYSTEM. The system is able to automatically collect, processing, and analyze the wheat images collected using a mobile phenotype cabin in the field conditions. A data management system was also developed to store and manage the original images and the calculated phenotypic data in the system. A group of 128 wheat varieties were planted with replicates under a freezing environment. Canopy images of the wheat were collected at the seedling stage and three image features were extracted for each wheat samples, including ExG, ExR and ExV. The results show that the developed methods can clearly distinguish wheat samples with different wheat freezing injury scores. The automatic phenotypic analysis method of freezing injury provides a solution for high-throughput phenotypic analysis of field wheat and can quantify the stress caused by freezing injury at the seedling stage. The method has a certain guiding significance for wheat breeding.