Faster R-CNN Algorithm Research Articles

Sugarcane stem node detection and localization for cutting using deep learning.

In order to promote sugarcane pre-cut seed good seed and good method planting technology, we combine the development of sugarcane pre-cut seed intelligent 0p99oposeed cutting machine to realize the accurate and fast identification and cutting of sugarcane stem nodes. In this paper, we proposed an algorithm to improve YOLOv4-Tiny for sugarcane stem node recognition. Based on the original YOLOv4-Tiny network, the three maximum pooling layers of the original YOLOv4-tiny network were replaced with SPP (Spatial Pyramid Pooling) modules, which fuse the local and global features of the images and enhance the accurate localization ability of the network. And a 1×1 convolution module was added to each feature layer to reduce the parameters of the network and improve the prediction speed of the network. On the sugarcane dataset, compared with the Faster-RCNN algorithm and YOLOv4 algorithm, the improved algorithm yielded an mean accuracy precision (MAP) of 99.11%, a detection accuracy of 97.07%, and a transmission frame per second (fps) of 30, which can quickly and accurately detect and identify sugarcane stem nodes. In this paper, the improved algorithm is deployed in the sugarcane stem node fast identification and dynamic cutting system to achieve accurate and fast sugarcane stem node identification and cutting in real time. It improves the seed cutting quality and cutting efficiency and reduces the labor intensity.

Research on Surface Defect Detection Technology of Flywheel Disc Semi-Finished Products Based on Deep Learning

In order to avoid the disadvantages of traditional methods such as low detection efficiency and poor detection quality, this paper designs and develops an intelligent detection system based on Faster R-CNN algorithm for the problem of detecting surface defects on the flywheel disc semi-finished products. A new network based on the ResNet50 network can better catch and describe the surface defects. Based on the k-means++ clustering analysis algorithm, the anchor generation rules are optimized. The experimental results show that the average accuracy of the optimized algorithm is improved. And the detection accuracy of small defects has been greatly improved. The feasibility and effectiveness of the optimized algorithm are verified.

An improved faster RCNN-based weld ultrasonic atlas defect detection method

In view of the complex multi-scale target detection environment of ultrasonic atlas of weld defect and the poor detection performance of existing algorithms for the multiple small target defects, the Faster RCNN convolution neural network is applied to weld defect detection, and a Fast RCNN deep learning network is proposed in combination with an improved ResNet 50. Based on the coexistence of multiple small targets and multi-scale target detection, this paper proposes to combine deformable network, FPN network and ResNet50 to improve the detection performance of the algorithm for multi-scale targets, especially small targets. Based on the efficiency and accuracy of candidate frame selection, K-means clustering algorithm and ROI Align algorithm are proposed, and the anchors points and candidate frames suitable for weld defect data sets are customized for accurate positioning. Through the self-made ultrasonic atlas data set of weld defects and experimental verification of the improved algorithm in this paper, the overall mean average precision has reaches 93.72%, and the average precision of small target defects such as “stoma” and “crack” has reaches 92.5% and 88.9% respectively, which is 4.8% higher than the original Faster RCNN algorithm. At the same time, through the ablation experiments and comparison experiments with other mainstream target detection algorithms, it is proved that the improved method proposed in this paper improves the detection performance and is superior to other algorithms. The actual industrial detection scene proves that it basically meets the requirements of weld defect detection, and can provide a reference for the intelligent detection method of weld defects.

Individual Tree Detection in Coal Mine Afforestation Area Based on Improved Faster RCNN in UAV RGB Images

Forests are the most important part of terrestrial ecosystems. In the context of China’s industrialization and urbanization, mining activities have caused huge damage to the forest ecology. In the Ulan Mulun River Basin (Ordos, China), afforestation is standard method for reclamation of coal mine degraded land. In order to understand, manage and utilize forests, it is necessary to collect local mining area’s tree information. This paper proposed an improved Faster R-CNN model to identify individual trees. There were three major improved parts in this model. First, the model applied supervised multi-policy data augmentation (DA) to address the unmanned aerial vehicle (UAV) sample label size imbalance phenomenon. Second, we proposed Dense Enhance Feature Pyramid Network (DE-FPN) to improve the detection accuracy of small sample. Third, we modified the state-of-the-art Alpha Intersection over Union (Alpha-IoU) loss function. In the regression stage, this part effectively improved the bounding box accuracy. Compared with the original model, the improved model had the faster effect and higher accuracy. The result shows that the data augmentation strategy increased AP by 1.26%, DE-FPN increased AP by 2.82%, and the improved Alpha-IoU increased AP by 2.60%. Compared with popular target detection algorithms, our improved Faster R-CNN algorithm had the highest accuracy for tree detection in mining areas. AP was 89.89%. It also had a good generalization, and it can accurately identify trees in a complex background. Our algorithm detected correct trees accounted for 91.61%. In the surrounding area of coal mines, the higher the stand density is, the smaller the remote sensing index value is. Remote sensing indices included Green Leaf Index (GLI), Red Green Blue Vegetation Index (RGBVI), Visible Atmospheric Resistance Index (VARI), and Normalized Green Red Difference Index (NGRDI). In the drone zone, the western area of Bulianta Coal Mine (Area A) had the highest stand density, which was 203.95 trees ha−1. GLI mean value was 0.09, RGBVI mean value was 0.17, VARI mean value was 0.04, and NGRDI mean value was 0.04. The southern area of Bulianta Coal Mine (Area D) was 105.09 trees ha−1 of stand density. Four remote sensing indices were all the highest. GLI mean value was 0.15, RGBVI mean value was 0.43, VARI mean value was 0.12, and NGRDI mean value was 0.09. This study provided a sustainable development theoretical guidance for the Ulan Mulun River Basin. It is crucial information for local ecological environment and economic development.

Application of artificial intelligence image recognition technology based on Faster-R CNN in automatic inspection of power transmission and transformation

In order to effectively improve the work efficiency and real-time performance of an automatic inspection of power transmission and transformation, this paper studies the automatic inspection method of power transmission and transformation based on artificial intelligence image recognition technology. First, we use cameras and unmanned aerial vehicles (UAVs) to obtain image data for power transmission and transformation inspection. Second, through the artificial sample calibration and reconstruction technology, the power transmission, and transformation inspection sample database is constructed. Then, the model is trained using the Faster-R CNN algorithm based on regional relationship features. Finally, the model is further optimized through model release and evaluation to realize intelligent diagnosis in power transmission and transformation detection. This research raises the recognition accuracy of power transmission and transformation inspection, improves the intelligence and efficiency of inspection, and provides good technical support for the construction of intelligent operation inspection.

A Real-Time Detection Algorithm for Sweet Cherry Fruit Maturity Based on YOLOX in the Natural Environment

Fast, accurate, and non-destructive large-scale detection of sweet cherry ripeness is the key to determining the optimal harvesting period and accurate grading by ripeness. Due to the complexity and variability of the orchard environment and the multi-scale, obscured, and even overlapping fruit, there are still problems of low detection accuracy even using the mainstream algorithm YOLOX in the absence of a large amount of tagging data. In this paper, we proposed an improved YOLOX target detection algorithm to quickly and accurately detect sweet cherry ripeness categories in complex environments. Firstly, we took a total of 2400 high-resolution images of immature, semi-ripe, and ripe sweet cherries in an orchard in Hanyuan County, Sichuan Province, including complex environments such as sunny days, cloudy days, branch and leaf shading, fruit overlapping, distant views, and similar colors of green fruits and leaves, and formed a dataset dedicated to sweet cherry ripeness detection by manually labeling 36068 samples, named SweetCherry. On this basis, an improved YOLOX target detection algorithm YOLOX-EIoU-CBAM was proposed, which embedded the Convolutional Block Attention Module (CBAM) between the backbone and neck of the YOLOX model to improve the model’s attention to different channels, spaces capability, and replaced the original bounding box loss function of the YOLOX model with Efficient IoU (EIoU) loss to make the regression of the prediction box more accurate. Finally, we validated the feasibility and reliability of the YOLOX-EIoU-CBAM network on the SweetCherry dataset. The experimental results showed that the method in this paper significantly outperforms the traditional Faster R-CNN and SSD300 algorithms in terms of mean Average Precision (mAP), recall, model size, and single-image inference time. Compared with the YOLOX model, the mAP of this method is improved by 4.12%, recall is improved by 4.6%, F-score is improved by 2.34%, while model size and single-image inference time remain basically comparable. The method in this paper can cope well with complex backgrounds such as fruit overlap, branch and leaf occlusion, and can provide a data base and technical reference for other similar target detection problems.

Armaments Detection Using Artificial Intelligence and Deep Learning for Security Application

Abstract: Security is usually a main concern in every domain, thanks to a rise in crime rate in a crowded event or suspicious lonely areas. Abnormal detection and monitoring have major applications of computer vision to tackle various problems. Thanks to growing demand in the protection of safety, security and private properties, needs and deployment of video surveillance systems can recognize and interpret the scene and anomaly events play an important role in intelligence monitoring. This project implements automatic gun (or) weapon detection employing a Convolution Neural Network (CNN) based Single Shot Multibox Detector and Faster R-CNN algorithms. Proposed implementation uses two sorts of datasets. One dataset, which had pre-labeled images and therefore the other one, may be a set of images, which were labeled manually. Results are tabulated, both algorithms achieve good accuracy, but their application in real situations is often based on the trade-off between speed and accuracy. The system is entailed with automatic detection of the handgun/ knife or the other crime objects without manual intervention. This technique operates directly on the raw inputs i.e. security footage of camera in live; thus is proven to exhibit increased accuracy thanks to usage of datasets and instant identification.

WEAPON DETECTION USING ARTIFICIAL INTELLEGENCE AND DEEP LEARNING FOR SECURITY APPLICATIONS

Security is always a main concern in every domain, due to a rise in crime rate in a crowded event or suspicious lonely areas. In This project we implement automatic gun (or) weapon detection using a Convolution Neural Network (CNN) based SSD and Faster RCNN algorithms. Proposed implementation uses two types of datasets. One dataset, which had pre-labelled images and the other one is a set of images, which were labelled manually

An OpenCL-Based FPGA Accelerator for Faster R-CNN

In recent years, convolutional neural network (CNN)-based object detection algorithms have made breakthroughs, and much of the research corresponds to hardware accelerator designs. Although many previous works have proposed efficient FPGA designs for one-stage detectors such as Yolo, there are still few accelerator designs for faster regions with CNN features (Faster R-CNN) algorithms. Moreover, CNN’s inherently high computational complexity and high memory complexity bring challenges to the design of efficient accelerators. This paper proposes a software-hardware co-design scheme based on OpenCL to implement a Faster R-CNN object detection algorithm on FPGA. First, we design an efficient, deep pipelined FPGA hardware accelerator that can implement Faster R-CNN algorithms for different backbone networks. Then, an optimized hardware-aware software algorithm was proposed, including fixed-point quantization, layer fusion, and a multi-batch Regions of interest (RoIs) detector. Finally, we present an end-to-end design space exploration scheme to comprehensively evaluate the performance and resource utilization of the proposed accelerator. Experimental results show that the proposed design achieves a peak throughput of 846.9 GOP/s at the working frequency of 172 MHz. Compared with the state-of-the-art Faster R-CNN accelerator and the one-stage YOLO accelerator, our method achieves and inference throughput improvements, respectively.

Artificial intelligence to detect the femoral intertrochanteric fracture: The arrival of the intelligent-medicine era.

Objective: To explore a new artificial intelligence (AI)-aided method to assist the clinical diagnosis of femoral intertrochanteric fracture (FIF), and further compare the performance with human level to confirm the effect and feasibility of the AI algorithm. Methods: 700 X-rays of FIF were collected and labeled by two senior orthopedic physicians to set up the database, 643 for the training database and 57 for the test database. A Faster-RCNN algorithm was applied to be trained and detect the FIF on X-rays. The performance of the AI algorithm such as accuracy, sensitivity, miss diagnosis rate, specificity, misdiagnosis rate, and time consumption was calculated and compared with that of orthopedic attending physicians. Results: Compared with orthopedic attending physicians, the Faster-RCNN algorithm performed better in accuracy (0.88 vs. 0.84 ± 0.04), specificity (0.87 vs. 0.71 ± 0.08), misdiagnosis rate (0.13 vs. 0.29 ± 0.08), and time consumption (5 min vs. 18.20 ± 1.92 min). As for the sensitivity and missed diagnosis rate, there was no statistical difference between the AI and orthopedic attending physicians (0.89 vs. 0.87 ± 0.03 and 0.11 vs. 0.13 ± 0.03). Conclusion: The AI diagnostic algorithm is an available and effective method for the clinical diagnosis of FIF. It could serve as a satisfying clinical assistant for orthopedic physicians.

Semi-finished flywheel disk based on deep learning research on surface defect detection technology

The surface defects of flywheel disc semi-finished products have complex and changeable morphological characteristics and random distribution. At present, relevant enterprises can only detect them through manual visual inspection. However, the low efficiency of manual inspection and the unstable inspection quality can easily lead to false inspections and missed inspections, which cannot meet the growing demand for production capacity. In order to achieve intelligent and efficient detection of defects, this paper proposes a surface defect detection algorithm for flywheel disc semi-finished products based on improved faster region-based convolutional neural networks (Faster R-CNN). First of all, based on multi-scale feature fusion, residual feature recalibration and deformable convolution, this paper designs a feature extraction network that can better capture and characterize defect morphology. Secondly, optimize the design of Faster R-CNN algorithm, use k-means++ cluster analysis to optimize the anchor generation rules in the network, so as to adapt to the defects of large aspect ratio, the region of interest (ROI) pooling calculation method incorporating global feature information is redesigned to prevent the position deviation of candidate areas when they are mapped back to the original image. Aiming at the problem that adjacent overlapping positive samples are deleted by mistake, the soft non-maximum suppression (Soft-NMS) algorithm is used to optimize the non-maximum suppression process and increase the number of positive samples output by the region proposal network(RPN). Then, the surface defect images are collected to build a data set, aiming at the problem that the data set is small and the distribution of the number of defects in each category is unbalanced, the classical data enhancement methods are used to augment the data set and equalize the defect categories. Finally, the surface defect detection and application experiment research of flywheel disc semi-finished products is carried out. The detection accuracy of the algorithm in this paper on the surface defect test set reaches 92.7%, which is 9.6% higher than the original Faster R-CNN detection accuracy, and 18.5% higher for the detection accuracy of small minor defects, and the improvement effect is more obvious.

Spark plug defects detection based on improved Faster-RCNN algorithm

The objective of this study is to apply an improved Faster-RCNN model in order to solve the problems of low detection accuracy and slow detection speed in spark plug defect detection. In detail, an attention module based symmetrical convolutional network (ASCN) is designed as the backbone to extract multi-scale features. Then, a multi-scale region generation network (MRPN), in which InceptionV2 is used to achieve sliding windows of different scales instead of a single sliding window, is proposed and tested. Additionally, a dataset of X-ray spark plug images is established, which contains 1,402 images. These images are divided into two subsets with a ratio of 4:1 for training and testing the improved Faster-RCNN model, respectively. The proposed model is transferred and learned on the pre-training model of MS COCO dataset. In the test experiments, the proposed method achieves an average accuracy of 89% and a recall of 97%. Compared with other Faster-RCNN models, YOLOv3, SSD and RetinaNet, our proposed new method improves the average accuracy by more than 6% and the recall by more than 2%. Furthermore, the new method can detect at 20fps when the input image size is 1024×1024×3 and can also be used for real-time automatic detection of spark plug defects.

Development of image-based wheat spike counter through a Faster R-CNN algorithm and application for genetic studies

Spike number (SN) per unit area is one of the major determinants of grain yield in wheat. Development of high-throughput techniques to count SN from large populations enables rapid and cost-effective selection and facilitates genetic studies. In the present study, we used a deep-learning algorithm, i.e., Faster Region-based Convolutional Neural Networks (Faster R-CNN) on Red-Green-Blue (RGB) images to explore the possibility of image-based detection of SN and its application to identify the loci underlying SN. A doubled haploid population of 101 lines derived from the Yangmai 16/Zhongmai 895 cross was grown at two sites for SN phenotyping and genotyped using the high-density wheat 660K SNP array. Analysis of manual spike number (MSN) in the field, image-based spike number (ISN), and verification of spike number (VSN) by Faster R-CNN revealed significant variation (P < 0.001) among genotypes, with high heritability ranged from 0.71 to 0.96. The coefficients of determination (R2) between ISN and VSN was 0.83, which was higher than that between ISN and MSN (R2 = 0.51), and between VSN and MSN (R2 = 0.50). Results showed that VSN data can effectively predict wheat spikes with an average accuracy of 86.7% when validated using MSN data. Three QTL Qsnyz.caas-4DS, Qsnyz.caas-7DS, and QSnyz.caas-7DL were identified based on MSN, ISN and VSN data, while QSnyz.caas-7DS was detected in all the three data sets. These results indicate that using Faster R-CNN model for image-based identification of SN per unit area is a precise and rapid phenotyping method, which can be used for genetic studies of SN in wheat.

Detection of Large Herbivores in UAV Images: A New Method for Small Target Recognition in Large-Scale Images

Algorithm design and implementation for the detection of large herbivores from low-altitude (200 m–350 m) UAV remote sensing images faces two key problems: (1) the size of a single image from the UAV is too large, and the mainstream algorithm cannot adapt to it, and (2) the number of animals in the image is very small and densely distributed, which makes the model prone to missed detection. This paper proposes the following solutions: For the problem of animal size, we optimized the Faster-RCNN algorithm in terms of three aspects: selecting a HRNet feature extraction network that is more suitable for small target detection, using K-means clustering to obtain the anchor frame size that matches the experimental object, and using NMS to eliminate detection frames that have sizes inconsistent with the size range of the detection target after the algorithm generates the target detection frames. For image size, bisection segmentation was used when training the model, and when using the model to detect the whole image, we propose the use of a new overlapping segmentation detection method. The experimental results obtained for detecting yaks, Tibetan sheep (Tibetana folia), and the Tibetan wild ass in remote sensing images of low-altitude UAV from Maduo County, the source region of the Yellow River, show that the mean average precision (mAP) and average recall (AR) of the optimized Faster-RCNN algorithm are 97.2% and 98.2%, respectively, which are 9.5% and 12.1% higher than the values obtained by the original Faster-RCNN. In addition, the results obtained from applying the new overlap segmentation method to the whole UAV image detection process also show that the new overlap segmentation method can effectively solve the problems of the detection frames not fitting the target, missing detection, and creating false alarms due to bisection segmentation.

Development of a core feature identification application based on the Faster R-CNN algorithm

As engineering rock mass quality assessment is an important part of the evaluation of deposit mining technical conditions, the identification and counting of core features are essential but time-consuming, and the complexity of core features leads to the invalidity of traditional image processing programs. In such a case, we developed an efficient automated core feature identification and counting application by adopting the Faster R-CNN algorithm together with a self-designed batch processing and counting program, which allows for the high-speed identification of target features among many pictures and can count and output formatted identification results. The evaluation results show that this application can significantly improve the identification accuracy and speed up the process with the help of a deep learning algorithm and our computer program.In the comparison and selection of the Faster R-CNN and YOLO algorithms, YOLO was eliminated due to poor performance. The main reason is that the multiscale self-similarity of core features has adverse effects on the multiscale segmentation method of the YOLO algorithm, making YOLO identify one feature repeatedly.The overall training evaluation F1-score of the Faster R-CNN-based AI model reaches 0.91, showing an ideal result. In the practical test, the overall AI identification F1-score is 0.93, and the application processing F1-score reaches 0.92. The time complexities of the AI model and application are both acceptable, with T(n) = O(n). In terms of identification speed, the application process is 48 times faster than that of manual identification.

Sea Cucumber Detection Algorithm Based on Deep Learning.

The traditional single-shot multiBox detector (SSD) for the recognition process in sea cucumbers has problems, such as an insufficient expression of features, heavy computation, and difficulty in application to embedded platforms. To solve these problems, we proposed an improved algorithm for sea cucumber detection based on the traditional SSD algorithm. MobileNetv1 is selected as the backbone of the SSD algorithm. We increase the feature receptive field by receptive field block (RFB) to increase feature details and location information of small targets. Combined with the attention mechanism, features at different depths are strengthened and irrelevant features are suppressed. The experimental results show that the improved algorithm has better performance than the traditional SSD algorithm. The average precision of the improved algorithm is increased by 5.1%. The improved algorithm is also more robust. Compared with YOLOv4 and the Faster R-CNN algorithm, the performance of this algorithm on the P-R curve is better, indicating that the performance of this algorithm is better. Thus, the improved algorithm can stably detect sea cucumbers in real time and provide reliable feedback information.

Surface Defect Detection Model for Aero-Engine Components Based on Improved YOLOv5

Aiming at the problems of low efficiency and poor accuracy in conventional surface defect detection methods for aero-engine components, a surface defect detection model based on an improved YOLOv5 object detection algorithm is proposed in this paper. First, a k-means clustering algorithm was used to recalculate the parameters of the preset anchors to make them match the samples better. Then, an ECA-Net attention mechanism was added at the end of the backbone network to make the model pay more attention to feature extraction from defect areas. Finally, the PANet structure of the neck network was improved through its replacement with BiFPN modules to fully integrate the features of all scales. The results showed that the mAP of the YOLOv5s-KEB model was 98.3%, which was 1.0% higher than the original YOLOv5s model, and the average inference time for a single image was 2.6 ms, which was 10.3% lower than the original model. Moreover, compared with the Faster R-CNN, YOLOv3, YOLOv4 and YOLOv4-tiny object detection algorithms, the YOLOv5s-KEB model has the highest accuracy and the smallest size, which make it very efficient and convenient for practical applications.

The Stress Detection and Segmentation Strategy in Tea Plant at Canopy Level.

As compared with the traditional visual discrimination methods, deep learning and image processing methods have the ability to detect plants efficiently and non-invasively. This is of great significance in the diagnosis and breeding of plant disease resistance phenotypes. Currently, the studies on plant diseases and pest stresses mainly focus on a leaf scale. There are only a few works regarding the stress detection at a complex canopy scale. In this work, three tea plant stresses with similar symptoms that cause a severe threat to the yield and quality of tea gardens, including the tea green leafhopper [Empoasca (Matsumurasca) onukii Matsuda], anthracnose (Gloeosporium theae-sinensis Miyake), and sunburn (disease-like stress), are evaluated. In this work, a stress detection and segmentation method by fusing deep learning and image processing techniques at a canopy scale is proposed. First, a specified Faster RCNN algorithm is proposed for stress detection of tea plants at a canopy scale. After obtaining the stress detection boxes, a new feature, i.e., RGReLU, is proposed for the segmentation of tea plant stress scabs. Finally, the detection model at the canopy scale is transferred to a field scale by using unmanned aerial vehicle (UAV) images. The results show that the proposed method effectively achieves canopy-scale stress adaptive segmentation and outputs the scab type and corresponding damage ratio. The mean average precision (mAP) of the object detection reaches 76.07%, and the overall accuracy of the scab segmentation reaches 88.85%. In addition, the results also show that the proposed method has a strong generalization ability, and the model can be migrated and deployed to UAV scenarios. By fusing deep learning and image processing technology, the fine and quantitative results of canopy-scale stress monitoring can provide support for a wide range of scouting of tea garden.

Research on the Application of Object Detection in the Installation of Spacer Bars

The correct installation of spacer bars is an important part of ensuring people’s daily life. This paper studies and implements a spacer bar installation confirmation system based on a deep learning object detection algorithm, which makes up for the deficiencies in the installation. Among the object detection algorithms, Faster-RCNN has higher detection accuracy, is more accurate than one-stage and can detect multi-scale and small objects. Through historical use, it is found that Faster-RCNN is excellent in detecting multi-class image sets tasks. For personal data sets, only fine-tuning is needed to achieve better results, and it also has a fast detection rate. Therefore, this paper is based on the Faster-RCNN algorithm to realize the detection and confirmation of the plug and rubber during the installation of the spacer bar.

Security Monitoring Strategy of Distribution Community Operation Site Based on Intelligent Image Processing Method

With the continuous development of energy infrastructure, a large number of distributed new energy, distributed energy storage, and various power electronic equipment are connected to distribution communities. The access to a large amount of equipment not only increases the workload of power operators, but also leads to a complex field operation environment, which will threaten the security of field operators. Traditional monitoring strategy for distribution community operation site relies on manual operation, which is tedious, labor-intensive, and error-prone. To solve this problem, this paper proposes a security monitoring strategy for distribution community operation sites based on an intelligent image processing method. Firstly, a power image enhancement method based on multi-filtering algorithm is proposed. The filter operators are used to smooth and denoise the image, enhance the edge and detail information, and expand the security monitoring data set of operation sites by merging the filtered images. Secondly, an object detection method for personnel and security protection tools based on the improved Faster R-CNN algorithm is proposed, and a pyramid structure is constructed to improve the detection and localization accuracy of small objects such as safety helmets and safety belts. To simplify the object detection task and the complexity of the labeling system, it is disassembled into four single tasks: operators, non-operators, safety helmets, and safety belts. Finally, the detection frame labels are fused by calculating the overlap of the object detection boxes to describe the security of the operator. Experiments show that the method proposed in this paper can effectively locate the object and accurately describe the security of the operator.