Due to the various shapes, textures, and colors of fires, forest fire detection is a challenging task. The traditional image processing method relies heavily on manmade features, which not universally applicable all scenarios. In order solve this problem, deep learning technology applied learn extract features fires adaptively. However, limited perception ability individual learners sufficient make them perform well in complex tasks. Furthermore, tend focus too much local information, namely...

Diseases and insect pests of tea leaves cause huge economic losses to the industry every year, so accurate identification them is significant. Convolutional neural networks (CNNs) can automatically extract features from images suffering disease infestation. However, photographs tree taken in a natural environment have problems such as leaf shading, illumination, small-sized objects. Affected by these problems, traditional CNNs cannot satisfactory recognition performance. To address this...

Tea diseases have a significant impact on the yield and quality of tea during growth trees. The shape scale are variable, disease targets usually small, with intelligent detection processes also easily disturbed by complex background growing region. In addition, some concentrated in entire area leaves, needing to be inferred from global information. Common target models difficult solve these problems. Therefore, we proposed an improved model called TSBA-YOLO. We use dataset collected at...

The frequent occurrence of forest fires causes irreparable damage to the environment and economy. Therefore, accurate detection is particularly important. Due various shapes textures flames large variation in target scales, traditional fire methods have high false alarm rates poor adaptability, which results severe limitations. To address problem low accuracy caused by multi-scale characteristics changeable morphology fires, this paper proposes YOLOv5s-CCAB, an improved model based on...

Forest fires are destructive and rapidly spreading, causing great harm to forest ecosystems humans. Deep learning techniques can adaptively learn extract features of smoke. However, the complex backgrounds different fire smoke in captured images make detection difficult. Facing background smoke, it is difficult for traditional machine methods design a general feature extraction module extraction. effective many fields, so this paper improves on You Only Look Once v5 (YOLOv5s) model, improved...

The frequent occurrence of forest fires in recent years has not only seriously damaged the forests’ ecological environments but also threatened safety public life and property. Smoke, as main manifestation flame before it is produced, advantage a wide diffusion range that easily obscured. Therefore, timely detection fire smoke with better real-time for early warnings wins valuable time firefighting great significance applications development systems. However, existing methods still have...

Forest fires have the characteristics of strong unpredictability and extreme destruction. Hence, it is difficult to carry out effective prevention control. Once fire spreads, devastating damage will be caused natural resources ecological environment. In order detect early forest in real-time provide firefighting assistance, we propose a vision-based detection spatial localization scheme develop system carried on unmanned aerial vehicle (UAV) with an OAK-D camera. During high incidence fires,...

Currently, the detection of tea pests and diseases remains a challenging task due to complex background diverse spot patterns leaves. Traditional methods pest mainly rely on experience farmers experts in specific fields, which is inefficient can easily lead misclassification omission diseases. single model often used for disease identification; however, its learning perception capabilities are insufficient complete target garden environments. To address problem that existing algorithms...

Video, as a key driver in the global explosion of digital information, can create tremendous benefits for human society. Governments and enterprises are deploying innumerable cameras variety applications, e.g., law enforcement, emergency management, traffic control, security surveillance, all facilitated by video analytics (VA). This trend is spurred rapid advancement deep learning (DL), which enables more precise models object classification, detection, tracking. Meanwhile, with...

Forest fires are major forestry disasters that cause loss of forest resources, ecosystem safety, and personal injury. It is often difficult for current fire detection models to achieve high accuracy on both large small targets at the same time. In addition, most existing single models, using only a model in complex environment has misclassification rate, rate needs be improved. Aiming above problems, this paper designs two (named WSB WSS) proposes an integrated learning-based WSB_WSS), which...

This study developed a new method to extract abalone visceral polysaccharide (AVP) and remove its heavy metals (except for mercury) by combining enzymatic hydrolysis, plate frame filtration, alcohol precipitation, high activated clay absorption with cation exchange resin anion resin. Alcohol dialysis, carbon adsorption were used mercury in AVP, respectively. Results indicated that these technologies are credible both AVP higher purity met the limit of standard. High could more efficiently...

Over the last decade, various deep neural network models have achieved great success in image recognition and classification tasks. The vast majority of high-performing a huge number parameters often require sacrificing performance accuracy when they are deployed on mobile devices with limited area power consumption. To address this problem, we present an SSD-MobileNet-v1 acceleration method based compression subgraph fusion for Field-Programmable Gate Arrays (FPGAs). Firstly, regularized...

The terrestrial laser scanner (TLS) has been widely used in forest inventories. However, with increasing precision of TLS, storing and transmitting tree point clouds become more challenging. In this paper, a novel compressed sensing (CS) scheme for broad-leaved is proposed by analyzing comparing different sparse bases, observation matrices, reconstruction algorithms. Our starts eliminating outliers simplifying statistical filtering voxel filtering. then applies Haar basis to thin the...

China is the largest turbot farming in world, where germplasm source security closely related to national security. To further promote sustainable development of aquaculture industry, Marine Fish Industry Technology System team has developed a new variety fast-growing (hereinafter referred as varieties) 2012. However, promotion effect varieties not reached expectation, requiring research on adoption behavior analysis varieties. Based 272 farmers, this study conducts theoretical framework...

Few-Shot classification aims at solving problems that only a few samples are available in the training process. Due to lack of samples, researchers generally employ set tasks from other domains assist target task, where distribution between assistant and task is usually different. To reduce gap, several lines methods have been proposed, such as data augmentation domain alignment. However, one common drawback these algorithms they ignore similarity selection before training. The fundamental...