Search

In a greenhouse environment where tomatoes are grown, pests in yellow sticky traps need to be detected in order to control the pest population. However, tomato pests typically found on yellow sticky traps are small in size and lack distinct visual features, making it difficult for convolutional networks to extract sufficient contextual information, thereby rendering the tasks of localization and classification exceptionally challenging. In this work, an improved approach based on the advanced object detection model You Only Look Once version 7-tiny (YOLOv7-tiny) is introduced, aiming to enhance the accuracy of detecting small tomato pests while maintaining computational complexity. Firstly, a context information extraction block (CIE) based on a Transformer encoder is proposed, and this block aims to capture global context, explore potential relationships between features, and emphasize important characteristics. Secondly, an Tiny-ELAN fusion network is introduced, which enhanced the feature fusion ability of the network. Thirdly, the feature fusion part takes the P2 feature layer into account and adds a P2 small target detection head. Finally, the SCYLLA-IoU (SIoU) loss function is introduced, and its components are redefined to incorporate direction information, which enhances the model’s learning ability and convergence performance. Experimental results show that our method can accurately detect three insects: whitefly (WF), macrolophus (MR), and nesidiocoris (NC) in the yellow sticky trap images of tomato crops. Compared with Faster R-CNN, SSD, YOLOv3-tiny, YOLOv5s, YOLOv7-tiny, YOLOv7, YOLOv7-x, YOLOv8n, YOLOv8s, YOLOv10n, and RT-DETR, the mean average precision of our method increased by 3.14%, 11.8%, 4.7%, 4.7%, 4.4%, 3.5%, 2.9%, 4.6%, 4.4%, 4.2%, and 4.2%, respectively.

This study provides a comprehensive analysis of the spatio-temporal dynamics of fish resources in the confluence waters of Poyang Lake and the Yangtze River, focusing on the initial phase of a 10-year fishing ban implemented in January 2020. Through hydroacoustic surveys conducted during both high-water (September 2020) and low-water (January 2021) periods, we identified significant variations in fish density and individual size across different sections. During the high water level period, fish concentrations were primarily observed in the confluence area between the Yangtze River and Poyang Lake, exhibiting higher densities compared to other regions. Conversely, fish congregated in the deep-water zones of the main river during the low water level period. The fish population was dominated by small to medium-sized individuals, with mean body lengths of 12.47 cm and 12.62 cm during the high and low water level periods, respectively. Notably, 42 and 33 fish species were recorded during the high-water and low-water surveys, respectively, emphasizing the region’s rich biodiversity. Importantly, the study demonstrates that the fishing ban has resulted in substantial increases in both fish density and mean body length, underscoring its effectiveness in fostering fish population recovery. These findings provide critical baseline data to inform scientific conservation and management strategies in this ecologically sensitive river–lake ecosystem.

In complex scenes, factors such as tree branches and leaves occlusion, dense distribution of tomato fruits, and similarity of fruit color to the background color make it difficult to correctly identify the ripeness of the tomato fruits when harvesting them. Therefore, in this study, an improved YOLOv8 algorithm is proposed to address the problem of tomato fruit ripeness detection in complex scenarios, which is difficult to carry out accurately. The algorithm employs several technical means to improve detection accuracy and efficiency. First, Swin Transformer is used to replace the third C2f in the backbone part. The modeling of global and local information is realized through the self-attention mechanism, which improves the generalization ability and feature extraction ability of the model, thereby bringing higher detection accuracy. Secondly, the C2f convolution in the neck section is replaced with Distribution Shifting Convolution, so that the model can better process spatial information and further improve the object detection accuracy. In addition, by replacing the original CIOU loss function with the Focal–EIOU loss function, the problem of sample imbalance is solved and the detection performance of the model in complex scenarios is improved. After improvement, the mAP of the model increased by 2.3%, and the Recall increased by 6.8% on the basis of YOLOv8s, and the final mAP and Recall reached 86.9% and 82.0%, respectively. The detection speed of the improved model reaches 190.34 FPS, which meets the demand of real-time detection. The results show that the improved YOLOv8 algorithm proposed in this study exhibits excellent performance in the task of tomato ripeness detection in complex scenarios, providing important experience and guidance for tomato ripeness detection.

The Qinghai–Tibet Plateau (QTP) is a unique ecological area that has faced issues like diminishing ecosystem stability and increasing pressures on resources and the environment. These issues have arisen as a result of the combined impact of global warming and human activities in recent times. The study of the growth and distribution patterns of schizothoracinae fishes can support guiding policy decisions about the conservation of aquatic species and ecological habitats in the QTP. The investigation on fish resources was carried out in the QTP section of the Nujiang River during the spring and autumn seasons of 2017, 2018, and 2019. A total of seven sampling sites were established based on variations in elevation. According to length–weight relationship (LWR) analysis, Schizothorax nukiangensis mainly displayed a negative allometric growth while Ptychobarbus kaznakovi and Schizopygopsis thermalis mainly showed near isometric growth or positive allometric growth in the QTP section of the Nujiang River. Due to temperature and food abundance, the three schizothoracinae fishes showed better growth performance in autumn than spring. Spatial heterogeneity exhibited a greater influence on the LWR of S. nukiangensis and P. kaznakovi than seasonal variation. In contrast, seasonal variation on S. thermalis showed greater influence than spatial heterogeneity. According to the linear mixed effect model (LMM), both spatial factors and seasons had influence on fish growth in the QTP. Schizothorax nukiangensis was identified as the predominant species from CWL to BS, spanning an altitude range of 1800 to 2700 m. Ptychobarbus kaznakovi was identified as the main species at LL, BB, and BR, occupying an altitude range of 2700 to 3800 m. Schizopygopsis thermalis is primarily distributed at altitudes beyond 4000 m and along the tributary river Yuqu. Principal coordinates analysis (PCOA) and nonmetric multidimensional scaling (NMDS) divided schizothoracinae fish populations into three clusters by spatial differences. Redundancy analysis (RDA) and Monte Carlo Permutation analysis revealed that habitat elevation and water temperature had a significant impact on schizothoracinae fish distribution. This article enhances our understanding of the distribution and environmental adaptation of indigenous fish in the Qinghai–Tibet Plateau.

Photothermal techniques are infrared remote sensing techniques that have been used for biomedical applications, as well as industrial non-destructive testing (NDT). Machine learning is a branch of artificial intelligence, which includes a set of algorithms for learning from past data and analyzing new data, without being explicitly programmed to do so. In this paper, we first review the latest development of machine learning and its applications in photothermal techniques. Next, we present our latest work on machine learning for data analysis in opto-thermal transient emission radiometry (OTTER), which is a type of photothermal technique that has been extensively used in skin hydration, skin hydration depth profiles, skin pigments, as well as topically applied substances and skin penetration measurements. We have investigated different algorithms, such as random forest regression, gradient boosting regression, support vector machine (SVM) regression, and partial least squares regression, as well as deep learning neural network regression. We first introduce the theoretical background, then illustrate its applications with experimental results.

A multi-mesh trammel net has been developed and used for sampling freshwater fishes. However, little is known about the catch efficiency of the net. This research investigated the catch efficiency of a multi-mesh trammel net (nominal mesh size; 10.0, 30.0, 50.0, and 70.0 mm) for fish sampling in the Yangtze River of China. Catch composition and factors affecting catch per unit effort based on fish number (CPUEN), weight (CPUEW), and species number (CPUES) were analyzed. The results showed that the net was capable of capturing a wide range of fish sizes (total length, 5.5 to 121.0 cm) and species (n = 50). Increasing soak time from 9.4 to 24.0 h resulted in a decrease in CPUEN and CPUEW while a longer soak time increased CPUES. CPUEN, CPUEW, and CPUES varied significantly with fishing locations. The net provides a potential complement to the current fish sampling techniques used in freshwater ecosystems (e.g., large rivers, lakes, and reservoirs). The findings of this research help to improve our understanding of the catch efficiency of the multi-mesh trammel net and provides insight into better designs for gears and methods for sampling diverse fish sizes and species.

The aim of this study is to develop new analysis techniques for skin capacitive image stitching and occlusion measurements. Through image stitching, small skin capacitive images can be stitched into large skin capacitive images and, therefore, provide more skin image information. Through occlusion, e.g., keeping the measurement device on skin for a period of time, the skin health status can be studied through time-dependent response curves. Results show that time-dependent skin capacitive imaging curves can tell us the information about transdermal water loss (TEWL) as well as skin surface profiles. By using the structural similarity index measure (SSIM), the TEWL map can be constructed, which shows the water loss map on the skin surface. We first present the theoretical background and then the experimental results.

Stratum corneum (SC) is the outermost skin layer. SC hydration is important for its cosmetic properties and barrier function. SC trans-epidermal water loss (TEWL) measurements and skin water content measurements are two key indexes used for SC characterisation. The instrument stability and accuracy are vitally important when measuring small changes. In this paper, we present our latest study on the effect of sun tan lotion on skin by using skin TEWL and skin water content measurements. We developed techniques to improve the measurement stability and to visualise small changes, as well as developed machine learning algorithms for processing the skin capacitive images. The overall results show that TEWL and skin water content measurements are capable of measuring the subtle changes of skin conditions due to the application of sun tan lotions. The results show that the TEWL values decreased after the sun tan lotion application. The sun tan lotion with SPF 20 had the lowest decrease, whilst the sun tan lotion with SPF 50+ had the highest decrease. The results also show that the skin water content increased after the sun tan lotion application, with SPF 20 having the highest increase, whilst SPF 50+ had the lowest increase.

As an alternative approach, viseme-based lipreading systems have demonstrated promising performance results in decoding videos of people uttering entire sentences. However, the overall performance of such systems has been significantly affected by the efficiency of the conversion of visemes to words during the lipreading process. As shown in the literature, the issue has become a bottleneck of such systems where the system’s performance can decrease dramatically from a high classification accuracy of visemes (e.g., over 90%) to a comparatively very low classification accuracy of words (e.g., only just over 60%). The underlying cause of this phenomenon is that roughly half of the words in the English language are homophemes, i.e., a set of visemes can map to multiple words, e.g., “time” and “some”. In this paper, aiming to tackle this issue, a deep learning network model with an Attention based Gated Recurrent Unit is proposed for efficient viseme-to-word conversion and compared against three other approaches. The proposed approach features strong robustness, high efficiency, and short execution time. The approach has been verified with analysis and practical experiments of predicting sentences from benchmark LRS2 and LRS3 datasets. The main contributions of the paper are as follows: (1) A model is developed, which is effective in converting visemes to words, discriminating between homopheme words, and is robust to incorrectly classified visemes; (2) the model proposed uses a few parameters and, therefore, little overhead and time are required to train and execute; and (3) an improved performance in predicting spoken sentences from the LRS2 dataset with an attained word accuracy rate of 79.6%—an improvement of 15.0% compared with the state-of-the-art approaches.

We present our latest research work on the development of a skin image analysis tool by using machine-learning algorithms. Skin imaging is very import in skin research. Over the years, we have used and developed different types of skin imaging techniques. As the number of skin images and the type of skin images increase, there is a need of a dedicated skin image analysis tool. In this paper, we report the development of such software tool by using the latest MATLAB App Designer. It is simple, user friendly and yet powerful. We intend to make it available on GitHub, so that others can benefit from the software. This is an ongoing project; we are reporting here what we have achieved so far, and more functions will be added to the software in the future.

The time resource management of phased-array radars is the key to fulfilling their performance, such as how phased-array radar can efficiently and reasonably schedule tasks under limited resources. Therefore, this paper proposes a task scheduling algorithm for phased-array radar based on dynamic three-way decision. The algorithm introduces three-way decision into the scheduling algorithm and divides the target into three threat areas according to the threat degree (i.e., threat area, nonthreat area, and potential threat area). Different threat domains are assigned different weights and combine the working mode and the task deadline to carry out comprehensive priority planning, so that the radar can reasonably allocate time according to the difference of the target threat level and the threat area in the tracking stage. In addition, an improved adaptive threshold algorithm is proposed to obtain a dynamic three-way decision to achieve the adaptation of the algorithm. A set of performance indicators have been defined to evaluate the algorithm. The relevant experiments have demonstrated that the proposed algorithm can effectively improve the processing capability of phased-array radars when dealing with high-threat targets.