Detection Precision Research Articles

Point of Interest Recognition and Tracking in Aerial Video during Live Cycling Broadcasts

Road cycling races, such as the Tour de France, captivate millions of viewers globally, combining competitive sportsmanship with the promotion of regional landmarks. Traditionally, points of interest (POIs) are highlighted during broadcasts using manually created static overlays, a process that is both outdated and labor-intensive. This paper presents a novel, fully automated methodology for detecting and tracking POIs in live helicopter video streams, aiming to streamline the visualization workflow and enhance viewer engagement. Our approach integrates a saliency and Segment Anything-based technique to propose potential POI regions, which are then recognized using a keypoint matching method that requires only a few reference images. This system supports both automatic and semi-automatic operations, allowing video editors to intervene when necessary, thereby balancing automation with manual control. The proposed pipeline demonstrated high effectiveness, achieving over 75% precision and recall in POI detection, and offers two tracking solutions: a traditional MedianFlow tracker and an advanced SAM 2 tracker. While the former provides speed and simplicity, the latter delivers superior segmentation tracking, albeit with higher computational demands. Our findings suggest that this methodology significantly reduces manual workload and opens new possibilities for interactive visualizations, enhancing the live viewing experience of cycling races.

High Sensitivity Bi2O3/Ti3C2Tx Ammonia Sensor Based on Improved Synthetic MXene Method at Room Temperature.

The MXene Ti3C2Tx was synthesized using hydrofluoric acid and an improved multilayer method in this study. Subsequently, a Bi2O3/Ti3C2Tx composite material was produced through hydrothermal synthesis. This composite boasts a unique layered structure, offering a large surface area that provides numerous contact and reaction sites, facilitating the adsorption of ammonia on its surface. The prepared Bi2O3/Ti3C2Tx-based sensor exhibits excellent sensing performance for ammonia gas, including high responsiveness, good repeatability, and rapid response-recovery time. The sensor's response to 100 ppm ammonia gas is 61%, which is 11.3 times and 1.6 times the response values of the Ti3C2Tx gas sensor and Bi2O3 gas sensor, with response/recovery times of 61 s/164 s at room temperature, respectively. Additionally, the gas sensitivity mechanism of the Bi2O3/Ti3C2Tx-based sensor was analyzed, and the gas sensing response mechanism was proposed. This study shows that the sensor can effectively enhance the accuracy and precision of ammonia detection at room temperature and has a wide range of application scenarios.

An Improved YOLOv8n Algorithm for Steel Surface Defect Detection

Due to the manufacturing process and other aspects, steel surface defects vary in shape and size, which is a difficult point for defect detection. For this reason, this paper proposes a steel surface defect detection algorithm based on YOLOv8n. The algorithm improves the C2f module of Backbone in this network by adding Context Guided Block (CG block), which can effectively extract the local features, surrounding context and global context, and integrate this information to improve the precision and accuracy of detection. Meanwhile, we also change the detection header of YOLOv8 to Detect_ASFF by introducing the Adaptive Spatial Feature Fusion (ASFF) technique. This improvement effectively filters out the conflicting information, which can allow the model to have a better adaptability at different scales. In this paper, Shape-IoU is adopted as the bounding box loss function to make up for the traditional IoU loss that ignores the intrinsic properties of the bounding box in the calculation process. After optimisation, mAP@0.5 are improved to 67.9% and mAP@0.5~0.95 are improved to 35.1%, which is 2.8% and 1.1% respectively. The improved algorithm in this paper has improved in detection accuracy, which proves the practicality and effectiveness of the algorithm in this paper.

Optimizing Catheter Verification

Purpose Accurate detection of central venous catheter (CVC) misplacement is crucial for patient safety and effective treatment. Existing artificial intelligence (AI) often grapple with the limitations of label inaccuracies and output interpretations that lack clinician-friendly comprehensibility. This study aims to introduce an approach that employs segmentation of support material and anatomy to enhance the precision and comprehensibility of CVC misplacement detection. Materials and Methods The study utilized 2 datasets: the publicly accessible RANZCR CLiP dataset and a bespoke in-house dataset of 1006 annotated supine chest x-rays. Three deep learning models were trained: a classification network, a segmentation network, and a combination of both. These models were evaluated using receiver operating characteristic analysis, area under the curve, DICE similarity coefficient, and Hausdorff distance. Results The combined model demonstrated superior performance with an area under the curve of 0.99 for correctly positioned CVCs and 0.95 for misplacements. The model maintained high efficacy even with reduced training data from the local dataset. Sensitivity and specificity rates were high, and the model effectively managed the segmentation and classification tasks, even in images with multiple CVCs and other support materials. Conclusions This study illustrates the potential of AI-based models in accurately and reliably determining CVC placement in chest x-rays. The proposed method shows high accuracy and offers improved interpretability, important for clinical decision-making. The findings also highlight the importance of dataset quality and diversity in training AI models for medical image analysis.

Rapid quality control for recycled coarse aggregates (RCA) streams: Multi-sensor integration for advanced contaminant detection

Recycling coarse aggregates from construction and demolition waste is essential for sustainable construction practices. However, the quality of recycled coarse aggregates (RCA) often fluctuates significantly, in contrast to the more stable quality of natural aggregates. Contaminants in RCA notably compromise its quality and usability. Therefore, automating the quality control of RCA is necessary for the recycling industry. This study introduces an industry-focused, innovative, and rapid quality control system that combines Laser-Induced Breakdown Spectroscopy (LIBS) with 3D scanning technologies to enhance the detection of contaminants in RCA streams. The system involves a synchronized application of LIBS for spectral analysis and 3D scanning for the physical characterization of different materials. Results reveal that the dependability of single-shot LIBS analysis has been enhanced, thus elevating the precision of contaminant detection. This improvement is achieved by accounting for the laser shot's angle of incidence and focal length adjustments. The introduced technology holds potential for application in the real-time examination of substantial volumes of RCA, facilitating a rapid and reliable quality control method. This rapid assessment technique delivers online data about the concentration of contaminants in RCA, including recycled fine aggregates, cement paste, bricks, foam, glass, gypsum, mineral fibers, plastics, and wood. This data is both essential and sufficient for choosing a cost-effective mortar recipe and guaranteeing the performance of the final concrete product in terms of strength and durability in construction projects. The system can monitor the quality of RCA flows at throughputs of 50 tons per hour per conveyor, characterizing approximately 4000 particles in every ton of RCA, in this way signaling the most critical contaminants at levels of less than 50 parts per million. With these characteristics, the system could also become relevant for other applications, such as characterizing mining waste or solid biofuels for power plants.

Enhancing image recognition robustness in early weed detection through optimal training data curation

Abstract Using convolutional neural networks (CNNs) for image recognition is effective for early weed detection. However, the impact of training data curation, specifically concerning morphological changes during the early growth phases of weeds, on recognition robustness remains unclear. We focused on four weed species (giant ragweed [Ambrosia trifida L.], red morningglory [Ipomoea coccinea L.], pitted morningglory [Ipomoea lacunosa L.], and burcucumber [Sicyos angulatus L.]) with varying cotyledon and true leaf shapes. Creating 16 models in total, we employed four dataset patterns with different growth stage combinations, two image recognition algorithms (object detection: You Look Only Once [YOLO] v5 and image classification: Visual Geometry Group [VGG] 19), and two conditions regarding the number of species treated (four and two species). We evaluated the effects of growth stage training on weed recognition success using two datasets. One evaluation revealed superior results with a single class/species training dataset, achieving >90% average precision for detection and classification accuracy under most conditions. The other dataset revealed that merging different growth stages with different shapes as a class effectively prevented misrecognition among different species when using YOLOv5. Both results suggest that integrating different shapes in a plant species as a single class is effective for maintaining robust recognition success amid temporal morphological changes during the early growth stage. This finding not only enhances early detection of weed seedlings but also bolsters the robustness of general plant species identification.

Cherry Tomato Detection for Harvesting Using Multimodal Perception and an Improved YOLOv7-Tiny Neural Network

Robotic fruit harvesting has great potential to revolutionize agriculture, but detecting cherry tomatoes in farming environments still faces challenges in accuracy and efficiency. To overcome the shortcomings of existing cherry tomato detection methods for harvesting, this study introduces a deep-learning-based cherry tomato detection scheme for robotic harvesting in greenhouses using multimodal RGB-D perception and an improved YOLOv7-tiny Cherry Tomato Detection (YOLOv7-tiny-CTD) network, which has been modified from the original YOLOv7-tiny by eliminating the “Objectness” output layer, introducing a new “Classness” method for the prediction box, and incorporating a new hybrid non-maximum suppression. Acquired RGB-D images undergo preprocessing such as color space transformation, point cloud normal vector angle computation, and multimodal regions of interest segmentation before being fed into the YOLOv7-tiny-CTD. The proposed method was tested using an AGV-based robot in a greenhouse cherry tomato farming facility. The results indicate that the multimodal perception and deep learning method improves detection precision and accuracy over existing methods while running in real time, and the robot achieved over 80% successful picking rates in two-trial mode in the greenhouse farm, showing promising potential for practical harvesting applications.

High-precision and lightweight small-target detection algorithm for low-cost edge intelligence

The proliferation of edge devices driven by advancements in Internet of Things (IoT) technology has intensified the challenge of achieving high-precision small target detection, as it demands extensive computational resources. This amplifies the conflict between the need for precise detection and the requirement for cost-efficiency across numerous edge devices. To solve this problem, this paper introduces an enhanced target detection algorithm, MSGD-YOLO, built upon YOLOv8. The Faster Implementation of CSP Bottleneck with 2 convolutions (C2f) module is enhanced through the integration of the Ghost module and dynamic convolution, resulting in a more lightweight architecture while enhancing feature generation. Additionally, Spatial Pyramid Pooling with Enhanced Local Attention Network (SPPELAN) replaces Spatial Pyramid Pooling Fast (SPPF) to expand the receptive field, optimizing multi-level feature aggregation for improved performance. Furthermore, a novel Multi-Scale Ghost Convolution (MSGConv) and Multi-Scale Generalized Feature Pyramid Network (MSGPFN) are introduced to enhance feature fusion and integrate multi-scale information. Finally, four optimized dynamic convolutional detection heads are employed to capture target features more accurately and improve small target detection precision. Evaluation on the VisDrone2019 dataset shows that compared with YOLOv8-n, MSGD-YOLO improves mAP@50 and mAP@50–95 by 14.1% and 11.2%, respectively. In addition, the model not only achieves a 16.1% reduction in parameters but also attains a processing speed of 24.6 Frames Per Second (FPS) on embedded devices, thereby fulfilling real-time detection requirements.

A Comparative Model for Blurred Text Detection in Wild Scene Using Independent Component Analysis (ICA) and Enhanced Genetic Algorithm (Using a Bird Approach) with Classifiers

The advent of the study of Scene Text Detection and Recognition has exposed some significant challenges text recognition faces, such as blurred text detection. This study proposes a comparative model for detecting blurred text in wild scenes using independent component analysis (ICA) and enhanced genetic algorithm (E-GA) with support vector machine (SVM) and k-nearest neighbors (KNN) as classifiers. The proposed model aims to improve the accuracy of blurred text detection in challenging environments with complex backgrounds, noise, and illumination variations. The proposed model consists of three main stages: preprocessing, feature extraction, and classification. In the preprocessing stage, the input image is first preprocessed to remove noise and enhance edges using a median filter and a Sobel filter, respectively. Then, the blurred text regions are extracted using the Laplacian of Gaussian (LoG) filter. In the feature extraction stage, ICA is used to extract independent components from the blurred text regions. The extracted components are then fed into an E-GA-based feature selection algorithm to select the most discriminative features. The E-GA simply fine tunes the selection functionalities of the traditional GA using a bird approach. The selected features are then normalized and fed into the SVM and KNN classifiers. Experimental results on a benchmarking dataset (ICDAR 2019 LSVT) shows that the model outperforms state-of-the-art methods in terms of detection accuracy, precision, recall, and F1-score. The proposed model achieves an overall accuracy of 95.13% for SVM and 88.69% for KNN, which is significantly higher than the already existing methods which for SVM is 93%. In conclusion, the proposed model provides a promising approach for detecting blurred text in wild scenes. The combination of ICA, E-GA, and SVM/KNN classifiers enhances the robustness and accuracy of the detection system, which can be beneficial for a wide range of applications, such as text recognition, document analysis, and security systems.

Enhancing IoT intrusion detection through machine learning with AN-SFS: a novel approach to high performing adaptive feature selection

The vast volume of redundant and irrelevant network traffic data poses significant hurdles for intrusion detection. Effective feature selection is crucial for eliminating irrelevant information. Presently, most filtering and embedded methods rely on fixed thresholds or ratios, necessitating prior knowledge. Conversely, wrapper methods are computationally intensive, and individual feature selection methods may introduce biases in evaluation. To address these challenges, this study introduces Adaptive Neighborhood based Feature Selection (AN-SFS), a dynamic feature selection approach that adapts to local statistical properties of the data. Unlike traditional methods, AN-SFS adjusts its threshold based on the characteristics of the current feature subset and incorporates statistical measures of neighboring features, capturing subtle relationships and dependencies. This adaptability enables AN-SFS to achieve robust and effective feature selection outcomes. Using NSL-KDD and UNSW-NB15 datasets, our model demonstrates superiority over conventional ML classifiers in detection rate, precision, and recall, achieving outstanding accuracy rates of 99.3% on NSL-KDD and 97.5% on UNSW-NB15, significantly outperforming contemporary methods. To further demonstrate the effectiveness of our feature selection approach, we conducted a series of comparisons with other statistical methods, including ANOVA, Pearson correlation, and the chi-square test. In each comparison, our approach consistently outperformed these alternatives, underscoring its superior efficacy.

An Analytical Review on Predictive Fault Tolerance for Industrial IoT Using Advanced Learning Techniques

This article presents a thorough examination of hybrid machine learning methodologies employed to improve fault tolerance in Industrial Internet of Things (IIoT) systems. These hybrid models enhance fault detection precision and resilience by integrating various machine learning techniques, surpassing conventional methods. The research commences with a comprehensive literature review of contemporary fault tolerance techniques, subsequently presenting a comparison of various datasets and a proposal for hybrid predictive algorithms. The efficacy of these algorithms is subsequently assessed in comparison to cutting-edge techniques. Principal findings underscore the enhanced precision of hybrid models in real-time applications, their diminished computing complexity, and their scalability in dynamic industrial settings. Nonetheless, issues pertaining to integration, data reliance, and real-time implementation are also addressed. The article concluded by outlining prospective research methods, encompassing sophisticated integration techniques, automated model optimisation, adaptive self-learning frameworks, and methodologies to safeguard data privacy. This paper outlines a framework for future advancements in fault tolerance, with considerable implications for both academia and industry.

Small Ruminant Parturition Detection Based on Inertial Sensors—A Review

The birth process in animals, much like in humans, can encounter complications that pose significant risks to both offspring and mothers. Monitoring these events can provide essential nursing support, but human monitoring is expensive. Although there are commercial monitoring systems for large ruminants, there are no effective solutions for small ruminants, despite various attempts documented in the literature. Inertial sensors are very convenient given their low cost, low impact on animal life, and their flexibility for monitoring animal behavior. This study offers a systematic review of the literature on detecting parturition in small ruminants using inertial sensors. The review analyzed the specifics of published research, including data management and monitoring processes, behaviors indicative of parturition, processing techniques, detection algorithms, and the main results achieved in each study. The results indicated that some methods for detecting birth concentrate on classifying unique animal behaviors, employing diverse processing techniques, and developing detection algorithms. Furthermore, this study emphasized that employing techniques that include analyzing animal activity peaks, specifically recurrent lying down and getting up occurrences, could result in improved detection precision. Although none of the studies provided a completely valid detection algorithm, most results were promising, showing significant behavioral changes in the hours preceding delivery.

Real-Time Freeway Crash Detection Framework using Connected Vehicle Waypoint Data

Accurate and timely detection of traffic accidents is crucial for transportation agencies seeking swift responses and effective traffic management, particularly on freeways where crash severity and traffic flow disruptions are amplified. While several methods have been developed using cameras or infrastructure-mounted sensors, these approaches may face challenges in geographical scalability. To address this issue, a novel approach has been developed that leverages real-time connected vehicle trajectories at a market penetration rate of 4%–9%. The method involves extracting five key features from individual journeys that sense the speed and acceleration fluctuations. The results demonstrate promising outcomes, achieving an overall accuracy of 63.5% with 4.1% false detections and 33.4% non-detections. This accuracy surpasses the speed-alone model using the same data by 24%. Furthermore, higher traffic volumes lead to even greater accuracy, exceeding 80% at level of service D and 90% at level of service E. The developed algorithm also exhibits precision in crash detection, with a mean latency of only 2.5 min after the actual incidents. The entire processing time may additionally add up to 1 min. The robustness of the model is tested for a single day on one of the interstates, providing promising results. This study highlights the potential of connected vehicle data in enhancing crash prediction methods and underlines its value for the safety and efficiency of future transportation systems.

Falling Detection of Toddlers Based on Improved YOLOv8 Models.

If toddlers are not promptly checked and rescued after falling from relatively high locations at homes, they are at risk of severe health complications. We present a toddler target extraction method and real-time falling alarm. The procedure is executed in two stages: In stage I, a GELAN-integrated YOLOv8 model is used to extract the body features. Based on this, a head capture technique is developed to obtain the head features. In stage II, the "safe zone" is calculated through Generalized Hough Transform (GHT). The spatial location is compared to the preceding stage's two centers of mass points, K for the toddler's body and H for the head. Position status detection is performed on the extracted data. We gathered 230 RGB-captured daily videos of toddlers aged 13 to 30 months playing and experiencing upside-down falls. We split 500 video clips (×30 FPS) from 200 videos into 8:2 training and validation sets. A test set of 100 clips (×30 FPS) was cut from another 30 videos. The experimental results suggested that the framework has higher precision and recall in detection, as well as improved mean average precision and F1 scores compared to YOLOv3, v5, v6, and v8. It meets the standard FPS requirement for surveillance cameras and has an accuracy of 96.33 percent.

A Hybrid Method Based on Haris Corner and Maximally Stable Extremal Regions for Vehicle Plate Number Detection

An intelligent transportation system (ITS) is a crucial element of a smart city, and it includes the capability to identify vehicle license plates. Utilizing digital image processing is a cost-effective method for identification. The tiny size of the number plate is just one of the many unfortunate difficulties with this approach. Hence, this research is crucial, particularly in enhancing the precision of detection. The Harris Corner approach is one way to locate the motor vehicle number plate location. However, the Harris corner method could be more optimal for analyzing moving vehicle videos as input. Since frame-by-frame variations in the video input's lighting, accuracy cannot be maximized. Furthermore, the vehicle and license plate appear significantly smaller due to the camera's distant positioning. Consequently, the authors suggest a hybrid approach using the Maximally Stable Extremal Regions (MSER) method. The Harris Corner and MSER methods will concurrently identify the initial position of the vehicle number plate. Moreover, the initial detection outcomes of the two techniques are compared and adjusted to achieve a more precise determination of the placement of the number plate. The results show that integrating the MSER method into the Harris Corner method yields a hybrid approach that enhances accuracy by 13%. Furthermore, it visually represents the selected number plate with greater accuracy.

Improved YOLOv8n for Lightweight Ship Detection

Automatic ship detection is a crucial task within the domain of maritime transportation management. With the progressive success of convolutional neural networks (CNNs), a number of advanced CNN models have been presented in order to detect ships. Although these detection models have achieved marked performance, several undesired results may occur under complex maritime conditions, such as missed detections, false positives, and low detection accuracy. Moreover, the existing detection models endure large number of parameters and heavy computation cost. To deal with these problems, we suggest a lightweight ship model of detection called DSSM–LightNet based upon the improved YOLOv8n. First, we introduce a lightweight Dual Convolutional (DualConv) into the model to lower both the number of parameters and the computational complexity. The principle is that DualConv combines two types of convolution kernels, 3x3 and 1x1, and utilizes group convolution techniques to effectively reduce computational costs while processing the same input feature map channels. Second, we propose a Slim-neck structure in the neck network, which introduces GSConv and VoVGSCSP modules to construct an efficient feature-fusion layer. This fusion strategy helps the model better capture the features of targets of different sizes. Meanwhile, a spatially enhanced attention module (SEAM) is leveraged to integrate with a Feature Pyramid Network (FPN) and the Slim-neck to achieve simple yet effective feature extraction, minimizing information loss during feature fusion. CIoU may not accurately reflect the relative positional relationship between bounding boxes in some complex scenarios. In contrast, MPDIoU can provide more accurate positional information in bounding-box regression by directly minimizing point distance and considering comprehensive loss. Therefore, we utilize the minimum point distance IoU (MPDIoU) rather than the Complete Intersection over Union (CIoU) Loss to further enhance the detection precision of the suggested model. Comprehensive tests carried out on the publicly accessible SeaShips dataset have demonstrated that our model greatly exceeds other algorithms in relation to their detection accuracy and efficiency, while reserving its lightweight nature.

Enhancing detection accuracy of highly overlapping targets in agricultural imagery using IoA-SoftNMS algorithm across diverse image sizes

Object detection is crucial for agricultural monitoring and agriculture automation, yet conventional non-maximum suppression (NMS) algorithms often struggle in agricultural scenes with complex backgrounds and highly overlapping targets, especially in intersection over union (IoU) threshold setting and abnormal non-maximum removal. This study proposed the soft non-maximum suppression based on intersection over area (IoA-SoftNMS) algorithm, an advanced NMS algorithm designed to enhance detection accuracy in such challenging agricultural environments. IoA-SoftNMS employs intersection over area (IoA) to quantify overlap and integrate SoftNMS to boost overall detection precision and effectiveness. Validation experiments demonstrated that IoA-SoftNMS significantly increased detection accuracy, improving mean average precision (mAP) by 0.7 % to 3.5 % across various object detection models, without necessitating model retraining. Importantly, these enhancements barely affect the detection time. IoA-SoftNMS exhibited robust adaptability in detecting large images, especially when integrated with the slicing-aided hyper inference (SAHI) technique, where it notably improved detection accuracy during the slicing process. Application of IoA-SoftNMS elevates detection accuracy in medium (1000 × 3300 pixels) and large (2500 × 3400 pixels) images from 93 % to 96.9 % and 90.8 % to 95.3 %, respectively. These findings underscored the effectiveness of the proposed IoA-SoftNMS algorithm in improving object detection accuracy in agricultural settings and its potential to alleviate challenges posed by large image sizes, offering avenues for enhancing agricultural production precision and intelligence.

ECBoA-OFS: An Ensemble Classification Model for Botnet Attacks based on Optimal Feature Selection using CPR in IoT

The rapid growth of the Internet of Things (IoT) has indeed introduced new security challenges, and the proliferation of compromised IoT devices has become a significant concern. Botnet attacks, where multiple corrupted devices are managed by a particular object, have become a widespread threat in IoT environments. These are used for a variety of malicious activities, including distributed DDoS attacks, data breaches, and malware distribution. However, detecting IoT botnets poses several challenges due to the resource constraints inherent in many IoT devices. The limitations in computation, storage, and communication capabilities make it challenging to deploy complex ML and deep learning models directly on these devices. This paper proposes an ensemble classification model ECBoA-OFS (Ensemble Classification for Botnet Attack Prediction using Optimal Feature Selection). It focuses on enhancing the accuracy of botnet attack prediction through the integration of ensemble methods and optimal feature selection. It describes a method for optimal feature selection in the context of analyzing the behavior of BoA and malicious traffic flow features in a network using Central Pivot Ranges (CPR). Feature selection is an important step in machine learning and data analysis because it supports to identification of the most important features for a given problem, thereby improving model performance and interpretation. The extracted features are used for model training and ensemble classification for prediction. To evaluate ECBoA-OFS, the N-BaIoT-2021 dataset consisting of regular IoT network traffic and BoA traffic records of corrupted IoT devices is utilized, considering detection precision, sensitivity, specificity, accuracy, and F1-score. Although all ensemble classifier models achieved better detection accuracy through optimal feature selection, the proposed ECBA-OFS shows better results compared to other ensemble classifier results.

Development of ExG, ExR, ExGR, HSV, CIELAB Images from RGB Images Using Image Segmentation Algorithm in Computer Vision Based Herbicide Spraying Applications

Weed management in agriculture is critical for preventing crop yield losses, with traditional methods often leading to environmental harm and increased production costs. This study explores the development of color indices and models for weed detection in computer vision-based herbicide spraying applications. Among the all sensors, RGB colour cameras offer several advantages, including low cost and wide availability of image processing libraries tailored for RGB image analysis. In present study a Logitech C270 webcam was used for acquiring the RGB images and a specially python algorithm was developed for image segmentation based on color. Excess Green (ExG), Excess Red (ExR), Excess Green minus Red (ExGR), HSV, and CIELAB color model images are developed by using algorithm. The research demonstrates that while ExG and ExGR indices are effective under specific conditions, the CIELAB model offers superior segmentation results across varying lighting environments. Among all these color indices and models, particularly CIELAB, can enhance the precision of weed detection in automated herbicide applications, thereby reducing environmental impact and improving agricultural efficiency.

Improved Small Object Detection Algorithm CRL-YOLOv5.

Detecting small objects in images poses significant challenges due to their limited pixel representation and the difficulty in extracting sufficient features, often leading to missed or false detections. To address these challenges and enhance detection accuracy, this paper presents an improved small object detection algorithm, CRL-YOLOv5. The proposed approach integrates the Convolutional Block Attention Module (CBAM) attention mechanism into the C3 module of the backbone network, which enhances the localization accuracy of small objects. Additionally, the Receptive Field Block (RFB) module is introduced to expand the model's receptive field, thereby fully leveraging contextual information. Furthermore, the network architecture is restructured to include an additional detection layer specifically for small objects, allowing for deeper feature extraction from shallow layers. When tested on the VisDrone2019 small object dataset, CRL-YOLOv5 achieved an mAP50 of 39.2%, representing a 5.4% improvement over the original YOLOv5, effectively boosting the detection precision for small objects in images.