Yield Of Agricultural Products Research Articles

Capsular attention Conv-LSTM network (CACN): A deep learning structure for crop yield estimation based on multispectral imagery

Precise prediction of agricultural production output is crucial for farmers, policymakers, and the Farming-related industry. This article introduces a novel methodology to crop yield forecasting using a capsular neural network equipped with Conv-LSTM and attention mechanism. Our model combines the strengths of 3DCNN, and Conv-LSTM, which can capture the temporal dependencies and 3D features of crop yield data, and attention mechanism, which Can prioritize the most significant characteristics for making predictions. We evaluated CACN on a sizable collection of data of soybean crop yield in the United States from 2003 to 2019 and evaluated against various cutting-edge deep learning models. The outcomes indicate that our suggested approach surpasses other models in performance in terms of RMSE, correlation coefficient, and prediction error map. Specifically, our model achieved approximately 14 % improvement in terms of RMSE, compared to the state-of-the-art model Deep-Yield. Our model also demonstrated the ability to extract more meaningful features and capture the complex relationships between crop yield data and meteorological variables. Overall, our proposed method shows great potential for accurate and efficient crop yield forecasting and can be applied to other crops and regions.

Visual information guided multi-modal model for plant disease anomaly detection

Plant diseases significantly impact the quality and yield of agricultural products, leading to considerable economic losses. Most existing plant disease recognition systems are limited to identifying categories within the training set, which poses potential systemic risks. Rejecting unknown samples is crucial for the safety and reliability of practical applications. This study aims to harness the strong generalization capabilities of vision-language models to address plant disease anomaly detection. To this end, we comprehensively explore prompt tuning paradigms based on vision-language models. We observe that anomaly detection methods guided by textual concepts perform poorly in the fine-grained task of plant disease due to their focus on concept matching. We argue that visual information is crucial for anomaly detection in plant diseases. Therefore, we propose guiding the vision-language model with visual information to address this issue. Additionally, we find that utilizing the general knowledge extracted by the original vision-language model can further enhance anomaly detection performance. Extensive experimental results demonstrate that our method significantly improves the current baseline methods by incorporating visual information. Notably, deploying our method under vision-language prompt tuning achieved an AUROC score of 99.85% in the all-shot setting. Even in a challenge 2-shot setting, our approach achieves an AUROC score of 93.81%, significantly outperforming CoCoOp fine-tuned on the entire dataset (88.61%). We believe that our study will contribute to the community and, to fuel the field, our code will be released.

Bacillus Species’ Gaseous Metabolites Prolong the Shelf Life of Apples and Grapes

In view of the threat to the environment posed by the widely used plant and fruit preservation chemicals in current agricultural practices, and the indispensability of fruit and plant preservatives in ensuring the yield of agricultural products, it is urgent to explore environmentally friendly and efficient preservative alternatives. Against this backdrop, bacterial preservatives stand out as a potential green solution. In this study, 12 strains of Bacillus derived from different ecological niches were carefully screened to evaluate the inhibitory efficacy of their gaseous metabolites against Anthracnose colloidospora, a common fruit pathogen, by bisectional method. The results showed that the bacterial suspension of strain B4 was the most prominent in inhibiting the growth of Colloidosporum anthracnose, and it significantly narrowed the growth circle of pathogenic bacteria, so it was selected as the object for further optimization and application. Subsequently, we designed a practical application test for fruit preservation, in which the bacteria of strain B4 were directly placed in the same environment as the fruit to be preserved, and the observation was continued for five days. By comparing and analyzing the fruit state before and after treatment, we were pleased to find that strain B4 showed a positive effect on inhibiting fruit rot, indicating its great potential as a natural fruit preservative. This discovery not only provides a scientific basis for reducing the use of chemical preservatives, but also contributes to the sustainable development of agriculture and the protection of the ecological environment.

A Smart Farm Monitoring and Control System using Wireless Sensor Network

Agriculture is one of the most important sector that accounts for economic growth in the developing countries of the world. Many developing countries are now focusing on agricultural development, yet the sector is not without challenges including climate changes, drought, flooding to mention but a few. These result in poor yield of agricultural products. In this work, we l developed a robust smart system to enhance both irrigation and flooding monitoring and control, leveraging on Wireless Sensor Networks (WSNs) to boost agricultural production. In the system’s implementation, Arduino based instrumentation, integrated with temperature, soil moisture and water level sensor shall be adopted to monitor the agricultural environment, while reporting the status wirelessly through the Radio Frequency (RF) modules to the base station. The base station will evaluate the received data and either activates or deactivates the irrigation or drainage pump using specified threshold values. The remote reporting the state of farm shall be done by deploying a 900Mhz transmitter interfaced with the system’s controller. This work shall be validated using an experimental test bed, which shall be used for field experiments, data collection and evaluation. The result of the work shall highlight the potentials of WSN technology in monitoring and control technology of both irrigation and flooding within the farm and in turn boost productivity

Fuzzy logic-based control for robot-guided strawberry harvesting: visual servoing and image segmentation approach

The concept of digital farming can help farmers increase their agricultural production yield. One of the technologies to support digital farming is robotics, which can be utilized to complete a redundant task efficiently for 24 hours. This paper presents a simple and effective harvesting robot that is applied to harvest a ripe strawberry. The mechanical and electrical design is kept simple to ensure it is reproducible. The input from a proximity sensor and image detection by a Pi camera is utilized by FLC (Fuzzy Logic Controller) to improve the effectiveness of the harvesting task. The image processing method in this study is image segmentation, which fits with the limited source of the microcontroller available in the market. The experiment included 60 times (20 times center, left, and right position) harvesting using the FLC algorithm and 60 times without FLC to show the effectiveness of the proposed method. From 60 experiments without an FLC experiment, there is an 80% hit rate for strawberries positioned in the middle of an image plane and 55% for left and right strawberries. From 60 times of FLC experiment, 95% hit rate for strawberries positioned in the middle of an image plane, 80% for left and right strawberries. The average time required to finish the task without FLC for strawberries in the middle is 13.51 s, the left is 11.04 s, and the right is 17.28 s. While the average time required to finish the task with FLC for strawberry in the middle is 12.90 s, the left side is 11.71 s, and the right side is 10.93 s. This study is intended to show that simple designs can be helpful and affordable when applied to greenhouse farming in Indonesia.

Characteristics of Rapeseed (Brassica rapa L.) genome DREB family demonstrate their roles in stress

Dehydration-responsive element-binding (DREB) proteins are crucial transcription factors in plant regulatory mechanisms, playing a key role in responding to environmental stresses. Dehydration-responsive element-binding (DREB) proteins are crucial transcription factors in plant regulatory mechanisms, playing a key role in responding to environmental stresses. This study employs bioinformatics research methods to identify and characterize the DREB gene family in rapeseed (Brassica rapa L.) under abiotic stress conditions. The approach facilitates a detailed exploration of the genomic architecture and potential regulatory mechanisms that these genes may exhibit in response to environmental stresses. A whole-genome survey identified 75 DREB genes in rapeseed, which were classified into six subgroups (Ⅰ-Ⅵ) based on phylogenetic comparison with Arabidopsis homologs. The amino acid lengths of these genes ranged from 144 to 530, with predicted molecular weights between 15.63 kDa and 60.36 kDa, and isoelectric points from 4.48 to 9.32. Chromosomal localization revealed that these genes are distributed across 10 chromosomes, predominantly near telomeric regions, with more than 10 DREB genes located on Chr3 and Chr9 respectively. Structural domain analysis indicated that 57 DREB genes possess a single exon. Promoter cis-element analyses suggested that BrDREB genes could respond to diverse stressors, including drought, heat, hormones, and light. Transcriptomic analysis under cold stress revealed significant upregulation of BrDREB17, 24, 49, 50, and 51 in response to low temperature, while five BrDREB genes were not expressed. Specifically, BrDREB17, 24, 49, 50, and 51 showed high expression levels at 3 h of cold stress, whereas BrDREB30, 42, and 67 were highly expressed at 24 h of cold stress. Gene Ontology (GO) analysis demonstrated that the majority of biological processes mediated by BrDREB are associated with plant stress resistance. Protein interaction network analysis revealed that BrDREB9 and BrDREB26 interact with 15 proteins each, underscoring their significant biological roles within the DREB family in rapeseed. These findings provide a theoretical framework for further examination of the physiological and biochemical properties of the DREB gene family components in rapeseed. Our results may help to screen candidate genes for cold resistance in rapeseed and select rapeseed varieties better suited to the Tibetan locality to improve quality and yield in agricultural production.

Exploration of the Application of Agricultural Intelligent Robots in the Qinghai-Tibet Plateau

In order to improve agricultural efficiency, reduce labor intensity, and improve the quality and yield of agricultural products on the Tibetan Plateau, this paper takes the application of traditional agricultural robots as an example and proposes the application of agricultural robots to agriculture on the Qinghai-Tibet Plateau based on the special breeding environment of the Qinghai-Tibet Plateau combined with the current situation of traditional agriculture. Through the application practice of robots in traditional agriculture, the possibility of the application of agricultural robots in the Qinghai-Tibet Plateau is pointed out, the working methods and advantages of robots in the Qinghai-Tibet Plateau are expounded, and the application of intelligent agricultural robots in the agriculture of the Qinghai-Tibet Plateau is discussed and analyzed. The results show that with the progress of science and technology, intelligent robots have gradually played an important role in various fields, and intelligent robots have broad application prospects in plateau agriculture. The application of agricultural robots in the Qinghai-Tibet Plateau should be aimed at the Qinghai-Tibet Plateau to make some adaptive research and development, intelligent upgrading and demonstration project promotion in order to promote the Qinghai-Tibet Plateau agricultural automation.

Design and implementation of an AI-controlled spraying drone for agricultural applications using advanced image preprocessing techniques

PurposeNowadays, food problems are likely to arise because of the increasing global population and decreasing arable land. Therefore, it is necessary to increase the yield of agricultural products. Pesticides can be used to improve agricultural land products. This study aims to make the spraying of cherry trees more effective and efficient with the designed artificial intelligence (AI)-based agricultural unmanned aerial vehicle (UAV).Design/methodology/approachTwo approaches have been adopted for the AI-based detection of cherry trees: In approach 1, YOLOv5, YOLOv7 and YOLOv8 models are trained with 70, 100 and 150 epochs. In Approach 2, a new method is proposed to improve the performance metrics obtained in Approach 1. Gaussian, wavelet transform (WT) and Histogram Equalization (HE) preprocessing techniques were applied to the generated data set in Approach 2. The best-performing models in Approach 1 and Approach 2 were used in the real-time test application with the developed agricultural UAV.FindingsIn Approach 1, the best F1 score was 98% in 100 epochs with the YOLOv5s model. In Approach 2, the best F1 score and mAP values were obtained as 98.6% and 98.9% in 150 epochs, with the YOLOv5m model with an improvement of 0.6% in the F1 score. In real-time tests, the AI-based spraying drone system detected and sprayed cherry trees with an accuracy of 66% in Approach 1 and 77% in Approach 2. It was revealed that the use of pesticides could be reduced by 53% and the energy consumption of the spraying system by 47%.Originality/valueAn original data set was created by designing an agricultural drone to detect and spray cherry trees using AI. YOLOv5, YOLOv7 and YOLOv8 models were used to detect and classify cherry trees. The results of the performance metrics of the models are compared. In Approach 2, a method including HE, Gaussian and WT is proposed, and the performance metrics are improved. The effect of the proposed method in a real-time experimental application is thoroughly analyzed.

Study the yield and quality of bitter gourd fruit (Momordica charantia) in inoculation with two species of mycorrhizal fungi and phosphorus fertilizer under different irrigation regimes

Drought is known to be the most important constraint to the growth and yield of agricultural products in the world, and plant symbiosis with arbuscular mycorrhizal fungi (AMF) can be a way to reduce drought stress negative impacts. A two-year experiment to investigate the factorial combination of mycorrhizal fungi (Glomus mosseae, Glomus intraradices, Control) and phosphorus fertilizer (application and non-application of phosphorus) on fruit yield and phenolic acids changes bitter gourd under different irrigation regimes as a split factorial based on a randomized complete block design. Three irrigation regimes, including irrigation after 20%, 50%, and 80% available soil water content depletion (ASWD), were considered in the main plots. The results showed that under water deficit stress, fruit yield and physiological (photosynthesis rate (Pn), transpiration rate (Tr), stomatal conductance (Gs), RWC, total chlorophyll, and root colonization) parameters decreased compared to 20% ASWD, and biochemical (proline, soluble sugar, MDA, CAT, SOD, phenol) parameters and fruit phenolic acids (caffeic acid, coumaric acid, ferulic acid) increased. However, the inoculation of AMF and phosphorus fertilizer in three irrigation regimes decreased MDA content, but physiological and biochemical parameters and fruit phenolic acids were increased. In this study, the factorial combination of AMF and sufficient phosphorus improved the resistance of bitter gourd to water deficit, and this not only improved fruit yield but also increased fruit phenolic acids under 80% ASWD, which can be an innovation in the management of water resources and the production industry of medicinal plants with high antioxidant properties in water deficit areas.

Soil Pollution Status, Sources and Control Methods in China

Soil is the solid foundation for our survival and development. With the progress of society, our country pays more and more attention to soil pollution. Soil pollution can cause the quality and yield of agricultural products to decline, pollute groundwater and surface water resources, affect air quality, lead to soil pollution in regional areas, and cause the destruction of ecological environment and the sharp decline of biodiversity. These problems will affect China's future sustainable development and social stability, so a detailed analysis of the sources of soil pollution is conducive to analyzing the nature of pollutants and pollution methods, which plays an important role in the prevention and control of soil pollution. And to introduce to the public some new technologies of soil pollution control, new ideas and the future direction of soil remediation.

The concept of a forest as a natural object and an object of legal protection

This scientific article provides a general description of forest resources as an object of administrative and legal research and an object of legal protection, as well as a study of the general concept of a forest and the concept of a forest as an object of legal protection. In modern conditions (aggression from Russia and crimes against ecocide), the issue of forest protection is very relevant. As you know, everything in nature is interconnected and interdependent, so by preserving the forest, we also preserve other natural objects. It has been established that forest resources are wood, technical, medicinal and other forest products that are used to meet the needs of the population and production and are reproduced in the process of forming forest natural complexes.
 An analysis of the concepts of the forest from the point of view of scientists and legal acts was carried out, and the role of the forest and its functions were determined. Our vision of the concept of forest is proposed, namely: a forest is a complex, multifunctional, reproducible object of nature (plant life) and an object of state ownership, which is part of the natural environment and performs two important functions: economic and ecological. The economic includes: the use of wood in economic activities, the use of gifts of the forest (berries, mushrooms, medicinal plants) - in vital activities and human health. Ecological can include the restoration and purification of air in the atmosphere, a positive effect on the regulation of the ecosystem in all territories of Ukraine, from an agricultural point of view, it has a positive effect on the protection of soils from erosion and increases the yield of agricultural products, etc. After all, the possibility of the existence of a forest as such is related to its inherent property, namely: restoration, i.e., a continuous, constantly operating process of vital activity of interconnected organisms that go through a complete life cycle from birth to death with subsequent reproduction.
 The issue of legal protection of forests of Ukraine was also considered, measures were proposed to improve the legal regulation of relations. Based on the analysis of scientists, the features regulating relations in the use and protection of forests have been established. After all, the shortcomings of the national forest legislation and insufficient protection of forests, illegal felling, problems in delimiting the powers of control bodies, ecocide against natural forest objects by the aggressor state lead to the destruction of forests and damage to the surrounding natural environment.

Innovative measurement, trade-off-synergy relationship and influencing factors for agricultural net carbon emissions and effective supply of agricultural products in China

Sensitive zone of global climate change has been formed in China, and it has become a hot topic how can agriculture ensure food security and the supply of important agricultural products while achieving the “Dual Carbon” goal in the country. Based on such background, this paper uses the IPCC carbon emission calculation method, environmental input-output model and economic-water-carbon coefficient method to measure agricultural net carbon emissions, adopts bivariate spatial auto-correlation analysis and SYS-GMM to explore separately the relationship between agricultural net carbon emissions and effective supply of agricultural products, as well as the carbon reduction effect, growth effect and reasonable range of green technology innovation. The results show that: (1) China's agricultural net carbon emissions reveal a spatial distribution of “higher in the east than in the west than in the center” and a temporal characteristic of increasing year by year; China's effective supply of agricultural products shows an increasing trend and a spatial distribution of “higher in the east than in the center than in the west” in 2006–2012 and “higher in the east than in the west than in the center” in 2013–2020. (2) In 2006, 2010, 2015 and 2020, the number of provinces that belong to low-low agglomeration trade-off zone, low-high agglomeration synergy zone, non-significant zone, high-low agglomeration non-trade-off-synergy zone and high-high agglomeration trade-off zone averagely accounted for 12.500 %, 30.000 %, 26.667 %, 9.167 % and 21.667 % of the totality, respectively. (3) The carbon reduction and production growth effects of green technology innovation both show an inverted “U-shape”, and green technology innovation is conducive to both reducing agricultural net carbon emissions and improving supply of agricultural products when it is within a reasonable range of greater than 0.930. (4) Green technology innovation not only has significant spatial and temporal heterogeneity impact, but also exhibits a differential effect on productive agricultural carbon emissions, agricultural trade carbon emissions, agricultural carbon sinks, total output of agricultural products and agricultural net imports in international trade. Therefore, it is proposed that China should establish and improve green technology innovation incubation platforms, guide all participants to ensure the investment and application of green technology products within a reasonable range, formulate and implement regional differential policies and plan in accordance with local conditions, drive ultimately coordinated promotion of agricultural carbon emission reduction and product supply guarantee and lay an important foundation for achieving high-quality economic development and efficient ecological protection.

Impacts of UV radiation on Bacillus biocontrol agents and their resistance mechanisms.

Bacillus biocontrol agent(s) BCA(s) such as Bacillus cereus, Bacillus thuringiensis and Bacillus subtilis have been widely applied to control insects' pests of plants and pathogenic microbes, improve plant growth, and facilitate their resistance to environmental stresses. In the last decade, researchers have shown that, the application of Bacillus biocontrol agent(s) BCA(s) optimized agricultural production yield, and reduced disease risks in some crops. However, these bacteria encountered various abiotic stresses, among which ultraviolet (UV) radiation severely decrease their efficiency. Researchers have identified several strategies by which Bacillus biocontrol agents resist the negative effects of UV radiation, including transcriptional response, UV mutagenesis, biochemical and artificial means (addition of protective agents). These strategies are governed by distinct pathways, triggered by UV radiation. Herein, the impact of UV radiation on Bacillus biocontrol agent(s) BCA(s) and their mechanisms of resistance were discussed.

On the First Kilometer Problem of Agricultural Products Cold Chain Logistics in Huaihua City Considering Green Development

Huaihua has vast land and is the largest prefecture level city in Hunan Province. With abundant natural resources, it is a city with a high output of agricultural products in Hunan Province. The loss and quality of agricultural products during transportation have always been a key concern for farmers and consumers.This article analyzes the problems that exist in the first one kilometer of agricultural product cold chain logistics in Huaihua City, proposes a mobile pre cooling mode to ensure the quality of agricultural products in the first one kilometer, and conducts a simple calculation of its operational return, aiming to solve the problems of high loss and difficult quality assurance in the transportation process of agricultural products.

Modelling Lithuanian family farms’ participation in agri-environmental subsidy schemes: a Neural Network Approach

Properly targeted agri-environmental subsidies (AES) can ensure the implementation of the European Green Deal goals. Hence, it is important to know what factors encourage family farms to participate in the AES schemes in order to select appropriate political tools and properly use the allocated subsidies. We propose a Multilayer Perceptron neural network to examine 34 Lithuanian crop family farms and identify the factors affecting their participation in the AES. The results indicate that the decision by the Lithuanian family farms regarding the participation mainly depends on a few factors, including the agricultural production output of the farm and farmers’ education, while other factors, such as farmer age and farm size, were less important.

Classification of diseases in tomato leaves with Deep Transfer Learning

Plant diseases are important factors because they significantly affect the quality, quantity, and yield of agricultural products. Therefore, it is important to detect and diagnose these diseases at an early stage. The overall objective of this study is to develop an acceptable deep learning model to correctly classify diseases on tomato leaves in RGB color images. To address this challenge, we use a new approach based on combining two deep learning models VGG16 and ResNet152v2 with transfer learning. The image dataset contains 55 000 images of tomato leaves in 5 different classes, 4 diseases and one healthy class. The results of our experiment are promising and encouraging, showing that the proposed model achieves 99,08 % accuracy in training, 97,66 % in validation, and 99,0234 % in testing

Impact of climate change on household development resilience: Evidence from rural China

The adverse impacts of climate change have emerged as one of the primary threats to human development; however, the climate change economic impact evaluation literature has mostly estimated effects under an assumption of full certainty. This article aims to estimate the long-run impact of climate change on household development resilience. To do so, we use the China Health and Nutrition Survey, which contains detailed longitudinal socioeconomic information from 1989 to 2015, and matches them to weather data. We find that climate change negatively impacts household development resilience. This negative impact stems from both a decline in upward mobility of households and an increase in downside risks. In other words, the rise in temperatures not only reduces the pace of household economic growth but also amplifies economic volatility, elevating consumption risk for households. The effects are driven by the decrease in agricultural production output and the risk of off-farm income. However, longer-run adaptation has offset 22.09% of the short-run effects of climate change on household development resilience. In contrast, climate adaptation mitigates household welfare loss to a greater extent through agricultural input adjustments and income diversification in the long run.

Greenhouse environmental control target constrained by discrete surface curvature and multi-objective optimization algorithm

Protected greenhouse provides a suitable micro-environment for crop growth, which enhances the quality and yield of agricultural products. How to optimize the greenhouse environment is currently a pressing issue in protected agriculture. In this research, we propose a new environmental control method from the perspectives of crop photosynthesis rate and environmental control cost. The photosynthesis rate data are collected from a nested experiment of tomatoes, and these data are combined with the support vector regression algorithm to establish a photosynthesis rate model. The cost function of greenhouse environmental control is established by theoretical calculation. The non-dominated solution sets of photosynthesis rate and control cost are obtained through an improved non-dominated sorting genetic algorithm Ⅱ. To determine the optimal solution among the numerous Pareto solutions, the geometric characteristics of the photosynthetic response surface are considered to constrain the control target. Then, a discrete surface curvature method is proposed to computer the constraint region. The intersection between the Pareto solutions and the constraint region is regarded as the environmental control target. The validation experiment demonstrates that the proposed method can effectively enhance photosynthetic accumulation and reduce control cost. Compared to the traditional saturation method, this method achieves a 53% cost reduction while still obtaining 89% photosynthetic accumulation. Compared to the U-chord method, this method increases photosynthetic accumulation by 12% with only a 3% increase in cost. The research has important implications for protected agriculture production.

Dynamics of the Agricultural Water Footprint and the Decoupling Associations with Agricultural Economic Growth in Hangzhou, China

Understanding the relationship between the agricultural water footprint (AWF) and agricultural economic growth (AEG) is of great significance for promoting sustainable agriculture and regional economic development. In this study, we used agricultural statistics data from Hangzhou from 2010 to 2021 to calculate the AWF, predicted the decoupling relationship between the AWF and AEG, and explored the influencing factors of the decoupling relationship between the AWF and AEG. The results showed the following: (1) The AWF in Hangzhou exhibited a decreasing trend, with a reduction from 58.88 × 108 m3 in 2010 to 37.80 × 108 m3 in 2021; this was mainly related to the decline in the water footprints of grain, pork, and egg production. (2) The strong decoupling accounted for 63.64% of the decoupling between the AWF and AEG in Hangzhou during the study period. It was found that an agricultural structure adjustment was the main factor for achieving decoupling between the AWF and AEG. Under the guidance of policy, the decoupling between them could be changed by regulating the output of agricultural products with different water footprint contents per unit. (3) From 2022 to 2026, the AWF in Hangzhou is expected to decrease to 28.21 × 108 m3, while the agricultural economy is projected to increase to CNY 40.008 billion. There will continue to be a strong decoupling status between the AWF and AEG in Hangzhou.

Based on the multi-scale information sharing network of fine-grained attention for agricultural pest detection.

It is of great significance to identify the pest species accurately and control it effectively to reduce the loss of agricultural products. The research results of this project will provide theoretical basis for preventing and controlling the spread of pests and reducing the loss of agricultural products, and have important practical significance for improving the quality of agricultural products and increasing the output of agricultural products. At the same time, it provides a kind of effective prevention and control measures for farmers, so as to ensure the safety and health of crops. Because of the slow speed and high cost of manual identification, it is necessary to establish a set of automatic pest identification system. The traditional image-based insect classifier is mainly realized by machine vision technology, but because of its high complexity, the classification efficiency is low and it is difficult to meet the needs of applications. Therefore, it is necessary to develop a new automatic insect recognition system to improve the accuracy of insect classification. There are many species and forms of insects, and the field living environment is complex. The morphological similarity between species is high, which brings difficulties to the classification of insects. In recent years, with the rapid development of deep learning technology, using artificial neural network to classify pests is an important method to establish a fast and accurate classification model. In this work, we propose a novel convolutional neural network-based model (MSSN), which includes attention mechanism, feature pyramid, and fine-grained model. The model has good scalability, can better capture the semantic information in the image, and achieve more accurate classification. We evaluated our approach on a common data set: large-scale pest data set, PlantVillage benchmark data set, and evaluated model performance using a variety of evaluation indicators, namely, macro mean accuracy (MPre), macro mean recall rate (MRec), macro mean F1-score (MF1), Accuracy (Acc) and geometric mean (GM). Experimental results show that the proposed algorithm has better performance and universality ability than the existing algorithm. For example, on the data set, the maximum accuracy we obtained was 86.35%, which exceeded the corresponding technical level. The ablation experiment was conducted on the experiment itself, and the comprehensive evaluation of the complete MSSN(scale 1+2+3) was the best in various performance indexes, demonstrating the feasibility of the innovative method in this paper.