جستجوی مقالات مرتبط با کلیدواژه « یادگیری عمیق » در نشریات گروه « کشاورزی »

In this research, field experiments were carried out in two types of soil (sandy clay loam and clay) to model the traction performance of the tractor considering drawbar power, rolling resistance, and traction efficiency, using deep learning and convolutional neural network while having some parameters such as soil type and conditions, tool parameters, and operation parameters. The tests were conducted within each soil texture in the form of factorial tests based on the randomized complete block design (RCBD) in triplicates. The tests were done in various moisture levels (8-17% for dry soils and 18-40% for moist soils), tractor forward speed (1.2, 1.6, 1.8, and 2.2 km h-1), working depth (30 and 50 cm), the number of pass (2 and 6 times), and tire inflation pressure (20 and 25 psi). The cone index, dynamic load, draft force, and moisture content were measured in each tests. The networks designed to model the drawbar power, rolling resistance, and traction efficiency were of convolutional neural network type. Various algorithms such as Sgdm, Adam, and Rmsprop were utilized to train the network. The results showed that the neural network developed by Sgdm algorithm outperformed the others. Therefore, this algorithm was utilized for the modeling process. Statistical criteria such as R2 and MSE were also employed to evaluate the performance of the network. For the drawbar power, the 8-499-499-1 architecture showed the best performance with R2=0.9953 and MSE=0.0016. Concerning the rolling resistance, the best performance was observed in 8-301-305-1 architecture with R2=0.9903 and MSE=0.0039. The best performance for the traction efficiency was obtained by 8-371-371-1 architecture with R2=0.9888 and MSE=0.003. The results showed that these networks can be used to model parameters by removing convolution layers and reducing dimensions.

Maintaining the well-being of individuals is greatly influenced by a healthy lifestyle and balanced diet. The identification and segmentation of food items can be improved by utilizing a mobile-based system in this era of rapid lifestyle changes and technology. This article introduces a novel system that, upon receiving input images, detects and segmentation the food items within the images. The system utilizes deep learning techniques and models, employing the YOLO algorithm. By incorporating regression-based simple methods, the system achieves the capability to detect and categorize food items in a single pass through the network, aiming to enhance accuracy and speed in the detection process. YOLOv7 was employed for food detection and YOLOv5, YOLOv7, and YOLOv8 was utilized for image segmentation. Based on the results, the accuracy, recall, and average precision values for YOLOv7 were 0.844, 0.924, and 0.932, respectively. Furthermore, the instance segmentation performance of YOLOv7 outperformed YOLOv5 and YOLOv8, with precision, recall, and mean average precision values of 0.959, 0.943, and 0.906, respectively. These findings underscore the high accuracy in detecting Iranian foods and the remarkable speed and precision in food image segmentation attainable through advanced deep-learning algorithms. Consequently, this study establishes that accurate detection of Iranian foods can be accomplished through the utilization of sophisticated deep-learning techniques. This research focuses on promoting a healthy lifestyle through intelligent technology and novel deep learning algorithms in Iran.

In many countries, on average, more than 50% of people's food comes from grains, and nearly 70% of the cultivated area of one billion hectares of the world is dedicated to grains. A variety of weeds grow along with cereals in the fields, which can reduce crop yield due to competition for light, water and nutrients. To eliminate weeds accurately and with minimal problems, timely detection with high accuracy and speed is required. be done.
In the field of agriculture, it is controlling and eliminating weeds in grain fields. Weeds are one of the most important factors affecting the production of agricultural products, which are their most important competitors in conventional agriculture, they spray the entire field to eliminate weeds, while weeds appear scattered and patchy in the field. which shows the necessity of using precise agriculture to solve this type of heterogeneity. In addition to causing economic damage, the conventional method of fighting can cause pollution of the environment and even the human food chain. Research shows that the losses caused by pests, diseases and weeds can reach 40% of the global crop every year and it is predicted that this percentage will increase significantly in the coming years. Besides, according to the research of Goktoan et al., the annual cost of weeds for The Australian economy is estimated to be around $4 billion as a loss in agricultural income.

Among the new methods in this field is the use of machine vision technology and related methods such as deep learning object detection algorithms and convolutional neural networks (CNN). The steps related to the implementation of the project include preparing data for training and evaluating networks, using new object detection algorithms, using different convolutional neural networks with different characteristics to extract image features in algorithms, and using the Feature Pyramid Network (FPN) method in object detection algorithms. Was. The output of the networks was evaluated in terms of the number of detections, the exact location of detection and the time of detection in the field. ViTs is based on the Transformer architecture that was originally developed for NLP tasks. Transformers use self-awareness mechanisms that allow the model to capture complex relationships between elements in a sequence. In the case of ViTs, sequence elements are image patches. In using the transformer architecture for visual data, it is dividing the image into small and non-interfering parts. Each patch typically consists of a grid of pixels. These patches are considered the "words" of the image sequence. Spatial embeddings are added to image patches to provide spatial information to the model. Spatial embeddings are necessary because transformers do not have built-in notions of order or spatial relationships. ViTs use multi-series self-awareness mechanisms to capture relationships between different image patches, and the representation of each patch is updated by attention to other patches. Data separation is very important in data watch transformers for two reasons a) the model needs data to learn and b) we need data to measure the model because the model may not be able to extract the information correctly.

The best network in terms of positioning accuracy was the transform model (ViTs) with an average accuracy of 0.95. In addition to this, the network considered in this research managed to recognize 503 of the 535 target weeds, and this means that our network is able to recognize 95% of these weeds. The presented method has been able to reach the highest accuracy compared to other existing methods and has been able to detect existing grasses in a much shorter period of time. Compared to other methods, the reset50 algorithm has been able to detect more than 88%, although its execution time is about 2.5 times that of the proposed method.
In comparing the efficiency of algorithms, execution time is as important as accuracy. By making comparisons and considering 70% of the data as training data and 30% as test data, the presented algorithm has been able to detect the weeds in the field with an accuracy of over 90% in just 13 seconds.

Today, deep learning methods are much more efficient than other methods, so we can use the new methods available in deep learning in the field of agriculture.

Floods are one of the natural hazards that occur in many parts of the world and cause irreparable financial and human losses. One of the most important aspects of controlling this crisis is the accurate identification of flooded areas. Also, it is important to predict flood-prone areas in order to prevent and reduce losses and casualties related to flooding. In this article, with the help of Sentinel-1 satellite images and a deep learning encoder-decoder network, the flood phenomenon in the images has been identified and segmented. These images belong to the regions of Nebraska, North Alabama, Bangladesh, Red River North, and Florence, and the ground truth map of each image, in which the target and non-target classes are shown as 0 and 1, was provided by NASA in 2021. In this article, flood-affected areas have been identified and segmented using encoder-decoder convolutional neural networks and the aforementioned satellite images. This network consists of encoder and decoder paths each containing convolutional layers which are responsible for extracting features and recovering these features, respectively. Various evaluation criteria were used to evaluate the performance of this method, including accuracy, IoU, F1-Score, and Kappa. This method has shown very good performance in the process of identifying and segmenting flooded areas based on the results obtained. The IoU obtained during the evaluation process was 96.04%, which is higher than the highest IoU obtained in other comparable studies (76.81%).

Accurate detection of apple leaf diseases is essential to prevent the reduction in both the quantity and quality of crop yield. With advancements in deep learning methods, the diagnosis of these diseases is improving, but data limitations remain a significant barrier. This research evaluates pretrained deep learning models with precise settings and demonstrates that high accuracy in disease detection is possible even with limited data. The selected models perform better than conventional methods and introduce transfer learning as an effective strategy to combat limited data and the diversity of diseases. These models aid farmers and horticulture specialists in identifying and managing apple leaf diseases more efficiently and rapidly. Additionally, these models are not only beneficial for farmers but also for agricultural consultants, students of agricultural sciences, and researchers in this field. Moreover, these methods are adaptable to other diseases and plants, promising advancements in plant disease detection systems in the future. The present study has the potential to revolutionize horticultural processes through optimization and automation, leading to a transformation in plant disease management.

The use of modern technologies for detecting and measuring pest population density can be an important step in facilitating the implementation of integrated pest management programs and achieving more precise and effective control. In this study, the deep learning technique and convolutional neural network with AlexNet architecture were used for the automatic detection and counting of the tomato leaf miner, Tuta absoluta (Myrick) (Lepidoptera: Gelechiidae), which is one of the key pests of tomato plants in Iran. To collect images of adult T. absoluta insects, 15 delta traps were installed in two hectares of tomato farms at the Campus of Agricultural and Natural Resources, Razi University. The Sony DSC-WX100 camera with an effective sensor resolution of 18 megapixels was used to capture the images. The performance of the convolutional neural network with the AlexNet architecture was evaluated using the parameters of average accuracy, accuracy, and recall. For counting performance, the linear regression curve and coefficient of determination were used. The average accuracy (98.0), accuracy (100), and recall (100) parameters indicate the high performance of the convolutional neural network in detecting the tomato leaf miner, and the coefficient of determination (0.98) indicates the network's high accuracy in counting this pest. Overall, the results demonstrate that the neural network can provide a practical solution for the accurate detection and counting of this pest on tomato plants using captured images

Identifying the growth stages of fruits in orchards is an important factor in improving the quantity and quality of the final product. Having such information helps the growers to apply the appropriate treatment for each growth stage and also to get a proper understanding of the fruit harvesting time, which may change due to changing weather conditions. Therefore, in the present study, color images of golden apples were used to estimate the weeks remaining to the harvest time. The EfficientNetB1 model was used to classify images taken from different weeks of apple fruit development using deep learning technology and convolutional neural networks. The data were divided into three categories: training (60%), validation (20%), and test (20%). Also, two pre-processing processes, i.e. data normalization and data augmentation, were used to obtain better results. Finally, Nadam optimizer and categorical_crossentropy cost function were considered in creating the model. The results showed that the developed model would have a good ability to classify input images. The value of the correlation coefficient (R) for training, validation, and test data was 0.86, 0.88, and 0.87 respectively. Also, the ability of the model to classify different classes was presented using precision, recall, and f1-score parameters for each class, according to which some classes achieved 100% accuracy. Consequently, the obtained results can be used as a platform for the development of harvesting robots, mobile apps, as well as aerial imagery systems using drones, etc. to fulfill various purposes in precision agriculture, and in particular, precision horticulture.

Predicting the maximum temperature changes is very important and has become increasingly important due to the many effects it has on water resources, agriculture and the environment. By forecasting the temperature, one can be aware of future changes and make the necessary arrangements to adjust its negative effects on water resources, agriculture and the environment. Therefore, modeling and forecasting the maximum temperature can be used as an important tool in the planning and management of natural, economic and industrial resources.

In this article, the maximum temperature was modeled using the Long-Short-Term Memory (LSTM) method based on Discrete Wavelet Transform (DWT) and Complete Experimental Mode Decomposition (CEEMD) methods in two different climates (humid and semi-arid). For this purpose, the daily data of maximum temperature, minimum temperature, precipitation, and solar radiation were used from 2001 to 2020 of the synopic stations located in Siyahbisheh, Amol City in Mazandaran Province and Urmia City airport in West Azarbaijan Province. It was determined that in the semi-arid region, the parameters of maximum and minimum temperature two days before, and maximum and minimum temperature one day before, as well as the minimum temperature and solar radiation of the same day, and in the humid region, the parameters of maximum temperature two days before, and maximum and minimum temperature one day before, as well as the minimum temperature and solar radiation of the same day were recognized as the superior model.

The results of the analysis of the models showed the capability and high efficiency of the method used in estimating the maximum temperature. On the other hand, the pre-processor methods improved the results. In the investigations, it was observed that the results of analysis based on wavelet transformation led to better results so that the DC evaluation criterion for the superior model in the semi-arid region of Urmia City went from 0.965 to 0.993 and in the humid area of Amol City increased from 0.926 to 0.970 and the RMSE criterion in Urmia Airport decreased from 1.943 to 0.896 and in Siyahbisheh from 2.595 to 1.648.

The results showed an increase in DC evaluation criteria and a decrease in RMSE for the synoptic station of Urmia Airport by 2.74% and 53.87%, respectively, and by 4.80% and 35.50% for the Siyahbisheh Amol Synoptic Station, respectively. This again shows that wavelet conversion has the greatest effect in improving the performance of the LSTM model and the selected models have high capability and efficiency in estimating the maximum temperature. According to the results of the sensitivity analysis, it was determined that the temperature parameter of the previous day is the most influential in estimating the maximum daily temperature for two regions with different climates (humid and semi-arid).

Plant pathogens profoundly impact crop yields and pose a significant barrier to achieving environmentally friendly goals regarding farming, cultivating food, and mitigating hunger-related concerns. The utilization of traditional methods for generalized seasonal pesticide application has been found to cause significant harm to both the surroundings and human well-being. In contrast, nanomaterial biosensors have emerged as cost-effective, highly efficient, specific, rapid, and accurate approaches for diagnosing plant pathogens and diseases. The Internet of Bio-Nano Things (IoBNT) is an emerging networking paradigm that utilizes small, biodegradable, and nonintrusive devices to gather and detect biological signals in the surrounding environment. Integrating the biological and digital realms of the Internet is facilitated by a technologically advanced device known as the Bio Cyber Interface (BCI). The BCI is a technology that facilitates the translation of biochemical messages from nanonetworks within the human body into electromagnetic radiation and, conversely, the translation of electromagnetic radiation into biochemical information. The security of BCIs is a significant concern regarding their practical application. This is because connecting BCIs to the Internet (5G) reveals their interfaces to potential external threats. To effectively address this concern, it is necessary to classify abnormal patterns in BCI traffic accurately. The utilization of current Machine Learning (ML) anomaly detection techniques is hindered by the complexity of parameters and the intricate correlations that exist among BBI traffic parameters. These methods often necessitate the manual design of features. To achieve this objective, the current study examines the utilization of a hybrid ensemble consisting of Convolutional Neural Networks and Long Short-Term Memory (CNN + LSTM) that enable flexible and scalable feature design to distinguish between conventional and abnormal BCI traffic. Through rigorous validation utilizing singular and multi-dimensional models on the generated dataset, our hybrid ensemble Deep Learning (DL) of CNN+LSTM achieved an accuracy of approximately 94.6%, surpassing other DL architectures.

Iran is faced with a dry and semi-dry climate with destructive floods, droughts, and water shortages. Droughts and floods can affect the environment, economic, and social activities. Therefore, examining and predicting river discharge and planning management to control it, especially for future water consumption, is very valuable. In this study, changes in river discharge were modeled using statistical data from 2001 to 2020.Statistical data from synoptic and hydrometric stations in a semi-arid region in Urmia city of West Azerbaijan province and a humid region in Amol city of Mazandaran province were used. Out of twelve time-series models defined for the Long Short-Term Memory (LSTM) network, the best model was identified. Then, LSTM modeling was performed based on pre-processing methods of Discrete Wavelet Transform (DWT) and Complementary Ensemble Empirical Mode Decomposition (CEEMD). The results showed that the selected model has high ability and efficiency in estimating the amount of river discharge. On the other hand, pre-processing methods improved the results such that the DC evaluation criterion in the wavelet transform increased from 0.93 to 0.95 in the Nazloo River and from 0.83 to 0.90 in the Chalous River. The best evaluation results for test data using wavelet transform for the Nazloo River in the semi-arid climate with evaluation criteria of R=0.977, DC=0.954, and RMSE=0.018 were obtained. Furthermore, the results of the sensitivity analysis indicated that the discharge parameter of one day before is the most effective in daily discharge estimation.

Maize is one of the most important cereal crops worldwide, providing staple food for people globally. Counting maize tassels provides essential information about yield prediction, growth status, and plant phenotyping, but traditional manual approaches are expensive and time-consuming. Recent developments in technology, including high-resolution RGB imagery acquired by unmanned aerial vehicles (UAVs) and advanced machine-learning techniques such as deep learning (DL), have been used to analyze genotypes, phenotypes, and crops.In this study, we modified the YOLOv5s single-stage object detection technique based on a deep convolutional neural network and named it MYOLOv5s. We incorporated BottleneckCSP structures, Hardswish activation function, and two-dimensional spatial dropout layers to increase tassel detection accuracy and reduce overfitting. Our method's performance was compared with three state-of-the-art algorithms: Tasselnetv2+, RetinaNet, and Faster R-CNN. The results obtained from our proposed method demonstrate the effectiveness of MYOLOv5s in detecting and counting maize tassels. 

The High-Intensity Phenotyping Site (HIPS) dataset was collected from the large field at the Agronomy Center for Research and Education (ACRE) of Purdue University, located in West Lafayette, Indiana, USA during the 2020 growing season. A Sony Alpha 7R-III RGB camera mounted on a UAV at a 20m altitude captured high-resolution orthophotos with a pixel resolution of 0.25 cm. The dataset consisted of two replications of 22 entries each for hybrids and inbreds, planted on May 12 using a two-row segment plot layout with a plant population of 30,000 per acre. The hybrids and inbreds in this dataset had varying flowering dates, ranging from 20 days between the first and last variety.This article uses orthophotos taken on July 20th and 24th to train and test the proposed deep network "MYOLOv5s." These orthophotos were divided into 15 images (3670×2150) and then cropped to obtain 150 images (608 × 2048) for each date. Three modifications were applied to the original YOLOv5s to form MYOLOv5s: BottleneckCSP structures were added to the neck part of the YOLOv5s, replacing some C3 modules; two-dimensional spatial dropout layers were used in the defect layer; and the Hardswish activation function was utilized in the convolution structures. These modifications improved tassel detection accuracy. MYOLOv5s was implemented in the Pytorch framework, and the Adam algorithm was applied to optimize it. Hyper-parameters such as the number of epochs, batch size, and learning rates were also optimized to increase tassel detection accuracy.

In this study, we first compared the original and modified YOLOv5s techniques, and our results show that MYOLOv5s improved tassel detection accuracy by approximately 2.80%. We then compared MYOLOv5s performance to the counting-based approach TasselNetv2+ and two detection-based techniques: Faster R-CNN and RetinaNet. Our results demonstrated the superiority of MYOLOv5s in terms of both accuracy and inference time. The proposed method achieved an AP value of 95.30% and an RMSE of 1.9% at 84 FPS, making it about 1.4 times faster than the other techniques. Additionally, MYOLOv5s correctly detected the highest number of maize tassels and showed at least a 17.64% improvement in AP value compared to Faster R-CNN and RetinaNet, respectively. Furthermore, our technique had the lowest false positive and false negative values. The regression plots show that MYOLOv5s provided slightly higher fidelity counts than other methods.Finally, we investigated the effect of score values on the performance of detection-based models and calculated the optimal values of hyperparameters.

The MYOLOv5s technique outperformed other state-of-the-art models in detecting maize tassels, achieving the highest precision, recall, and average precision (AP) values.The MYOLOv5s method had the lowest root mean square error (RMSE) value in the error counting metric, demonstrating its accuracy in detecting and counting maize tassels.We evaluated the correlation between predicted and ground-truth values of maize tassels using the R2 score, and for the MYOLOv5s method, the R2 score was approximately 99.28%, indicating a strong correlation between predicted and actual values.The MYOLOv5s method performed exceptionally well in detecting tassels, even in highly overlapping areas. It accurately distinguished and detected tassels, regardless of their proximity or overlap with other objects.When compared to the counting-based approach TasselNetv2+, our proposed MYOLOv5s method showed faster inference times. This suggests that the MYOLOv5s method is computationally efficient while maintaining accurate tassel detection capabilities.

In the past decades, due to excessive extraction of underground water resources, decrease in rainfall and increase in air temperature, the level of underground water has decreased drastically. According to previous researches, Iran has 130 billion cubic meters of underground water resources; but in the last 20 and six years, renewable water resources have decreased to 110 and less than 100 billion cubic meters, respectively. Therefore, the issue of underground water level changes and the prediction it, is of particular importance. Therefore, in this research, a model was developed to predict these changes using the data absorption algorithm. In addition, a deep learning model was also developed as a competing model to compare the results of the proposed model with it. South Khorasan province was selected as a case study for modeling. The comparison between the proposed model and the competing model showed that the proposed model has a very high prediction ability and its accuracy is close to the accuracy of the competitor model. Based on this evaluation, for the proposed model and the competing model, (R2) was equal to 0.91 and 0.95, and the root mean square error (RMSE) was equal to 0.18 and 0.20, respectively. Also, explicit presentation of equations and parameters of the model along with providing uncertainties and a confidence interval are other advantages of proposed model that can provide a wide future for data absorption algorithms. Meanwhile, machine learning and deep learning models, that are widely used today, do not provide such benefits.

This study was performed to intelligent and rapid assessment of the status of colonies in terms of honey production efficiency during foraging period, and presenting a method based on machine vision system. Using deep learning method, at first the comb frame and then the geometric, textural and color pattern of honey were identified. After that, the percentage of honey area was calculated. To do this, the imaging test of bee colonies using digital camera was designed and performed in such a way that different states of cells were present on the combs. In image analysis stage, the convolutional neural network with YOLOv5 algorithm and semantic segmentation method were used. The results showed that the present intelligent system has the ability to detect the comb frame from the surrounding environment of the image with an accuracy of more than 88%. Also, honey-related areas in each comb were identified with almost 83% accuracy and about 240 times quicker that of an expert beekeeper. These results were simultaneously confirmed with manual counting by a skilled beekeeper. Due to increase in the estimation speed, reduction of human error and consequently reduction of disruption time in colony activity, the proposed method can be a proper alternative to the traditional method of using framing technique for regular visits and evaluation of honey production efficiency.

Identification and mapping of tree species at single-tree levels using remotely sensed data is important in sustainable forest management. On the other hand, UAVs provide possibilities to acquire remotely sensed data with high spatial and temporal resolution that facilitate monitoring single tree and assessing quantitative and qualitative characteristics of trees such as species types. Methods based on machine learning can identify species types on UAV colour images but not with high accuracies. Additionally, similarity of trees in visible wavelengths registered on UAV colour images causes errors in species recognition. Therefore, this study was aimed to evaluate combination of UAV-based colour imagery and digital surface model (DSM) and deep learning algorithms in species recognition of a Pistacia-Amygdalus stand.

A part of Pistacia-Amygdalus stands in the wild pistachio research forest with an area of 24 ha was selected for this study. The study area was covered by 649 colour images with spatial resolution of 3.5 cm in 16 flight lines acquired by a Phantom 4Pro UAV. In addition to orthophoto, the DSM with similar spatial resolution was obtained by inverse distance weighted (IDW) method. The colour images and the combination of colour images and DSM were then used by Convolutional Neural Networks (CNNs) to identify Pistacia trees and Amygdalus shrubs. Results were evaluated by indices of accuracy assessment (e.g., accuracy, area under curve (AUC) of receiver operating characteristics).

The results on colour images showed that Amygdalus shrubs (accuracy=0.77, AUC=0.82) were identified with slightly higher accuracy than Pistacia trees (accuracy=0.76, AUC=0.80). Moreover, the results on combination of colour images and DSM showed that Pistacia trees (accuracy=0.85, AUC=0.85) were identified with higher accuracy than Amygdalus shrubs (accuracy=0.81, AUC=0.83). The final map was consisted of 455 Pistacia trees and 1951 Amygdalus shrubs. In addition, the visual interpretation of results revealed that species recognition on the combination of colour images and DSM had higher accuracy despite almost similar values of the indices of accuracy assessment.

In general, the present study explored that combination of colour images and DSM can facilitate species recognition of Pistacia trees and Amygdalus shrubs. Furthermore, this study confirms the capability of CNN in mapping of species at the individual level.

Today, the implementation of precision agriculture to manage and control citrus pests can effectively optimize pesticide use, reducing adverse environmental effects and ensuring human health. But managing individual trees on a large scale is a big challenge. Therefore, using a machine vision system seems necessary to monitor and identify pests in different partsof the trees at different times. In this study, an Unmanned Aerial Vehicle (UAV) equipped with a camera was used to identify pests in other parts of the citrus orchard. For the optimal selection of the UAV linear speed, three speeds in the range of 10, 20, and 30 cm/s were considered. After framing and formatting, the recorded videos were trained in three pretrained models: AlexNet, VGG-16, and GoogleNet. Three optimization algorithms were used in the network training process: SGDm, RMSProp, and Adam. The evaluation results showed that the AlexNet model, with the help of SGDm algorithm, had the best performance in terms of detection accuracy. The highest pest detection accuracy was 96.43% at a velocity of 10 cm/s, so increasing the linear velocity to 30 cm/s reduced the detection accuracy by 13%. The results of this study show that using a combination of UAV technology and artificial intelligence methods can help professionals and farmers manage and control citrus orchard pests.

Assessing and estimating sediment transport from a long time ago is one of the major issues for hydraulic and river engineers. Determining the amount of bed load carried in rivers depends on various factors, and this factor has complicated this phenomenon. Studies on different rivers show that the amount of bed load in different hydraulic and hydrological conditions is different. In addition, the physical properties of bed load particles have a significant effect on the accuracy of model prediction, On the other hand, despite the emphasis on the unreliability of experimental equations that have been extended over a specific area, unfortunately, limited studies have been conducted on temporary changes in bed load. Therefore, Investigating the predictability of this phenomenon is of great importance. In this study, we will try to estimate the bed load in gravel bed rivers using classical and intelligent methods.

Machine learning methods due to their high accuracy in predicting various issues have been noted in recent years . Therefore, in the present study, two methods of classical artificial neural network (ANN) and deep learning of long short-term memory (LSTM), which is a kind of artificial neural network with layers and amplification algorithms to improve network performance; have been used to predict the bed load of 19 gravel-bed rivers. To define suitable models for networks, the results of 10 experimental formulas in bed load prediction have been evaluated and the parameters of superior formulas have been used as the input of intelligent networks.

The results showed that all experimental formulas had very poor results; As most formulas have predicted the bed load with a Discrepancy index of r greater than 100. However, machine methods with input parameters obtained from experimental formulas have acceptable accuracy in predicting bed load. and in comparison with machine methods, LSTM method has provided more accurate results than ANN method. Finally, the model related to the parameters of Begnold formula in LSTM method with DC= 0.900 and RMSE= 0.024 for the test data is the best model obtained from this research and The average diameter of sediment particles (D50), which is a common parameter of the top three models, has been selected as the most effective parameter in predicting bed load.

Despite the very poor performance of experimental formulas in predicting sediment transport, intelligent networks with input parameters derived from experimental formulas have had good results. Also, LSTM network is more efficient than artificial neural network (ANN) in predicting bed load transfer, which indicates that Maintaining training memory during the training process and adding reinforcement layers to the network improves network performance and increases network accuracy in subsequent training.

Soil is one of the most important sources of production in agriculture. Therefore, with the determination of soil and its important characteristics, proper management and sustainable use of agricultural lands can be achieved. The current study aimed to predict the soil texture using a machine vision system and deep convolutional neural network (DCNN) algorithm. The proposed CNN model was composed of two blocks, including convolutional layers, max pooling layers, a dropout layer, batch normalization layers, fully connected layers, and a support vector machine classifier. This model was trained and tested on the images of different soil samples (11 types of soil texture and a total of 790 soil sample images). The data is prepared by a machine vision system and a smartphone camera (Galaxy A8). Using the confusion matrix, important statistical parameters such as accuracy, precision, specificity, sensitivity, and area under the curve were obtained at 99.65%, 98.75%, 99.8%, 98.75, and 99.27%, respectively. The suggested model successfully and correctly classified the soil sample images with 98.1% accuracy. The obtained results indicated that this study's implemented deep learning model can be a proper alternative to costly and time-consuming laboratory methods for determining soil texture.

In recent years, due to drought in the country, the issue of management of available water resources is extremely important, and this attention is increasingly to the management of reservoirs and forecasting the volume of water in order to provide appropriate exploitation policies. On the other hand, seasonal and excessive rainfall caused dramatic changes in the bedding of rivers and catchments, which examines the forecasting models in the event of heavy rains, which in addition to preventing damage in addition to preventing damage. Due to the occurrence of floods, surplus water can also be used in the desired direction. Therefore, not developing a proper operation policy, especially in drought conditions, can cause a lot of damage to water-consuming sectors. Proper forecasting of water flows and reservoir inventories leads to the use of control curves for the optimal use of dams and reservoir systems. In this paper, due to the importance of the subject, a model based on deep learning and Mann-Kendall experimental test was used to estimate the flood rate in the Kan-Sulqan area. The results showed that the monthly difference in flood forecast for the convolution neural network is 0.00654 and for the Men-Kendall method is 0.19532. Also, the error rates of MSE, RMSE, MAE and MPE for the neural network were equal to 0.0019, 0.0439, 0.0239, and 0.0159, respectively, which shows the high accuracy of this method in estimating the flood rate in the region.

Agriculture has been recognized as the main motive for economic growth and development in different countries of the world. In the meantime, mapping croplands through the classification of remote sensing images is one of the effective solutions in decision making and providing food security to the community. In this research, croplands are classified into different classes of agricultural products (including wheat, barley, corn, alfalfa, potatoes, and Sugar beets) using multi-temporal optical (Sentinel-2) and synthetic aperture radar (Sentinel-1) satellite images. All the steps related to the preparation of satellite images, have been conducted in the Google Earth Engine online processing platform. A novel three-dimensional deep convolutional neural network is used as the classifier. The designed network, in addition to three-dimensional kernels with the ability to extract spatial and temporal information of each pixel simultaneously, uses some escape connections of the previous layers. These connections, contrary to the feed-forward convolutional networks, feed the output of the previous layers to the new layers. After dividing the ground truth data into two categories of training and evaluation and assessing the performance of the network with 50 different training and evaluation data, the network’s overall accuracy was calculated 91.6% on average. According to the final results, the designed escape connections increased the overall accuracy of classification by 2%. The proposed network was also compared with temporal and spatial-temporal Random Forests and Support Vector Machines which showed a better performance with a difference of at least 2.4%.