Automatic Fire Detection and Notification System Based on Improved YOLOv4 for the Blind and Visually Impaired

Firefighting Water Jet Trajectory Detection from Unmanned Aerial Vehicle Imagery Using Learnable Prompt Vectors

This research presents an innovative methodology aimed at monitoring jet trajectory during the jetting process using imagery captured by unmanned aerial vehicles (UAVs). This approach seamlessly integrates UAV imagery with an offline learnable prompt vector module (OPVM) to enhance trajectory monitoring accuracy and stability. By leveraging a high-resolution camera mounted on a UAV, image enhancement is proposed to solve the problem of geometric and photometric distortion in jet trajectory images, and the Faster R-CNN network is deployed to detect objects within the images and precisely identify the jet trajectory within the video stream. Subsequently, the offline learnable prompt vector module is incorporated to further refine trajectory predictions, thereby improving monitoring accuracy and stability. In particular, the offline learnable prompt vector module not only learns the visual characteristics of jet trajectory but also incorporates their textual features, thus adopting a bimodal approach to trajectory analysis. Additionally, OPVM is trained offline, thereby minimizing additional memory and computational resource requirements. Experimental findings underscore the method's remarkable precision of 95.4% and efficiency in monitoring jet trajectory, thereby laying a solid foundation for advancements in trajectory detection and tracking. This methodology holds significant potential for application in firefighting systems and industrial processes, offering a robust framework to address dynamic trajectory monitoring challenges and augment computer vision capabilities in practical scenarios.

YOLOFM: an improved fire and smoke object detection algorithm based on YOLOv5n

To address the current difficulties in fire detection algorithms, including inadequate feature extraction, excessive computational complexity, limited deployment on devices with limited resources, missed detections, inaccurate detections, and low accuracy, we developed a highly accurate algorithm named YOLOFM. We utilized LabelImg software to manually label a dataset containing 18644 images, named FM-VOC Dataset18644. In addition, we constructed a FocalNext network, which utilized the FocalNextBlock module from the CFnet network. This improves the integration of multi-scale information and reduces model parameters. We also proposed QAHARep-FPN, an FPN network that integrates the structure of quantization awareness and hardware awareness. This design effectively reduces redundant calculations of the model. A brand-new compression decoupled head, named NADH, was also created to enhance the correlation between the decoupling head structure and the calculation logic of the loss function. Instead of using the CIoU loss for bounding box regression, we proposed a Focal-SIoU loss. This promotes the swift convergence of the network and enhances the precision of the regression. The experimental results showed that YOLOFM improved the baseline network's accuracy, recall, F1, mAP50, and mAP50-95 by 3.1%, 3.9%, 3.0%, 2.2%, and 7.9%, respectively. It achieves an equilibrium that combines performance and speed, resulting in a more dependable and accurate solution for detection jobs.

Early Fire Detection Using Long Short-Term Memory-Based Instance Segmentation and Internet of Things for Disaster Management

Fire outbreaks continue to cause damage despite the improvements in fire-detection tools and algorithms. As the human population and global warming continue to rise, fires have emerged as a significant worldwide issue. These factors may contribute to the greenhouse effect and climatic changes, among other detrimental consequences. It is still challenging to implement a well-performing and optimized approach, which is sufficiently accurate, and has tractable complexity and a low false alarm rate. A small fire and the identification of a fire from a long distance are also challenges in previously proposed techniques. In this study, we propose a novel hybrid model, called IS-CNN-LSTM, based on convolutional neural networks (CNN) to detect and analyze fire intensity. A total of 21 convolutional layers, 24 rectified linear unit (ReLU) layers, 6 pooling layers, 3 fully connected layers, 2 dropout layers, and a softmax layer are included in the proposed 57-layer CNN model. Our proposed model performs instance segmentation to distinguish between fire and non-fire events. To reduce the intricacy of the proposed model, we also propose a key-frame extraction algorithm. The proposed model uses Internet of Things (IoT) devices to alert the relevant person by calculating the severity of the fire. Our proposed model is tested on a publicly available dataset having fire and normal videos. The achievement of 95.25% classification accuracy, 0.09% false positive rate (FPR), 0.65% false negative rate (FNR), and a prediction time of 0.08 s validates the proposed system.

An Improved Wildfire Smoke Detection Based on YOLOv8 and UAV Images

Forest fires rank among the costliest and deadliest natural disasters globally. Identifying the smoke generated by forest fires is pivotal in facilitating the prompt suppression of developing fires. Nevertheless, succeeding techniques for detecting forest fire smoke encounter persistent issues, including a slow identification rate, suboptimal accuracy in detection, and challenges in distinguishing smoke originating from small sources. This study presents an enhanced YOLOv8 model customized to the context of unmanned aerial vehicle (UAV) images to address the challenges above and attain heightened precision in detection accuracy. Firstly, the research incorporates Wise-IoU (WIoU) v3 as a regression loss for bounding boxes, supplemented by a reasonable gradient allocation strategy that prioritizes samples of common quality. This strategic approach enhances the model's capacity for precise localization. Secondly, the conventional convolutional process within the intermediate neck layer is substituted with the Ghost Shuffle Convolution mechanism. This strategic substitution reduces model parameters and expedites the convergence rate. Thirdly, recognizing the challenge of inadequately capturing salient features of forest fire smoke within intricate wooded settings, this study introduces the BiFormer attention mechanism. This mechanism strategically directs the model's attention towards the feature intricacies of forest fire smoke, simultaneously suppressing the influence of irrelevant, non-target background information. The obtained experimental findings highlight the enhanced YOLOv8 model's effectiveness in smoke detection, proving an average precision (AP) of 79.4%, signifying a notable 3.3% enhancement over the baseline. The model's performance extends to average precision small (APS) and average precision large (APL), registering robust values of 71.3% and 92.6%, respectively.

Fire Detection and Notification Method in Ship Areas Using Deep Learning and Computer Vision Approaches

Fire incidents occurring onboard ships cause significant consequences that result in substantial effects. Fires on ships can have extensive and severe wide-ranging impacts on matters such as the safety of the crew, cargo, the environment, finances, reputation, etc. Therefore, timely detection of fires is essential for quick responses and powerful mitigation. The study in this research paper presents a fire detection technique based on YOLOv7 (You Only Look Once version 7), incorporating improved deep learning algorithms. The YOLOv7 architecture, with an improved E-ELAN (extended efficient layer aggregation network) as its backbone, serves as the basis of our fire detection system. Its enhanced feature fusion technique makes it superior to all its predecessors. To train the model, we collected 4622 images of various ship scenarios and performed data augmentation techniques such as rotation, horizontal and vertical flips, and scaling. Our model, through rigorous evaluation, showcases enhanced capabilities of fire recognition to improve maritime safety. The proposed strategy successfully achieves an accuracy of 93% in detecting fires to minimize catastrophic incidents. Objects having visual similarities to fire may lead to false prediction and detection by the model, but this can be controlled by expanding the dataset. However, our model can be utilized as a real-time fire detector in challenging environments and for small-object detection. Advancements in deep learning models hold the potential to enhance safety measures, and our proposed model in this paper exhibits this potential. Experimental results proved that the proposed method can be used successfully for the protection of ships and in monitoring fires in ship port areas. Finally, we compared the performance of our method with those of recently reported fire-detection approaches employing widely used performance matrices to test the fire classification results achieved.

Forest Fire Smoke Detection Based on Deep Learning Approaches and Unmanned Aerial Vehicle Images

Wildfire poses a significant threat and is considered a severe natural disaster, which endangers forest resources, wildlife, and human livelihoods. In recent times, there has been an increase in the number of wildfire incidents, and both human involvement with nature and the impacts of global warming play major roles in this. The rapid identification of fire starting from early smoke can be crucial in combating this issue, as it allows firefighters to respond quickly to the fire and prevent it from spreading. As a result, we proposed a refined version of the YOLOv7 model for detecting smoke from forest fires. To begin, we compiled a collection of 6500 UAV pictures of smoke from forest fires. To further enhance YOLOv7's feature extraction capabilities, we incorporated the CBAM attention mechanism. Then, we added an SPPF+ layer to the network's backbone to better concentrate smaller wildfire smoke regions. Finally, decoupled heads were introduced into the YOLOv7 model to extract useful information from an array of data. A BiFPN was used to accelerate multi-scale feature fusion and acquire more specific features. Learning weights were introduced in the BiFPN so that the network can prioritize the most significantly affecting characteristic mapping of the result characteristics. The testing findings on our forest fire smoke dataset revealed that the proposed approach successfully detected forest fire smoke with an AP50 of 86.4%, 3.9% higher than previous single- and multiple-stage object detectors.

A YOLOv6-Based Improved Fire Detection Approach for Smart City Environments

Authorities and policymakers in Korea have recently prioritized improving fire prevention and emergency response. Governments seek to enhance community safety for residents by constructing automated fire detection and identification systems. This study examined the efficacy of YOLOv6, a system for object identification running on an NVIDIA GPU platform, to identify fire-related items. Using metrics such as object identification speed, accuracy research, and time-sensitive real-world applications, we analyzed the influence of YOLOv6 on fire detection and identification efforts in Korea. We conducted trials using a fire dataset comprising 4000 photos collected through Google, YouTube, and other resources to evaluate the viability of YOLOv6 in fire recognition and detection tasks. According to the findings, YOLOv6's object identification performance was 0.98, with a typical recall of 0.96 and a precision of 0.83. The system achieved an MAE of 0.302%. These findings suggest that YOLOv6 is an effective technique for detecting and identifying fire-related items in photos in Korea. Multi-class object recognition using random forests, k-nearest neighbors, support vector, logistic regression, naive Bayes, and XGBoost was performed on the SFSC data to evaluate the system's capacity to identify fire-related objects. The results demonstrate that for fire-related objects, XGBoost achieved the highest object identification accuracy, with values of 0.717 and 0.767. This was followed by random forest, with values of 0.468 and 0.510. Finally, we tested YOLOv6 in a simulated fire evacuation scenario to gauge its practicality in emergencies. The results show that YOLOv6 can accurately identify fire-related items in real time within a response time of 0.66 s. Therefore, YOLOv6 is a viable option for fire detection and recognition in Korea. The XGBoost classifier provides the highest accuracy when attempting to identify objects, achieving remarkable results. Furthermore, the system accurately identifies fire-related objects while they are being detected in real-time. This makes YOLOv6 an effective tool to use in fire detection and identification initiatives.

Improved Cattle Disease Diagnosis Based on Fuzzy Logic Algorithms

The health and productivity of animals, as well as farmers' financial well-being, can be significantly impacted by cattle illnesses. Accurate and timely diagnosis is therefore essential for effective disease management and control. In this study, we consider the development of models and algorithms for diagnosing diseases in cattle based on Sugeno's fuzzy inference. To achieve this goal, an analytical review of mathematical methods for diagnosing animal diseases and soft computing methods for solving classification problems was performed. Based on the clinical signs of diseases, an algorithm was proposed to build a knowledge base to diagnose diseases in cattle. This algorithm serves to increase the reliability of informative features. Based on the proposed algorithm, a program for diagnosing diseases in cattle was developed. Afterward, a computational experiment was performed. The results of the computational experiment are additional tools for decision-making on the diagnosis of a disease in cattle. Using the developed program, a Sugeno fuzzy logic model was built for diagnosing diseases in cattle. The analysis of the adequacy of the results obtained from the Sugeno fuzzy logic model was performed. The processes of solving several existing (model) classification and evaluation problems and comparing the results with several existing algorithms are considered. The results obtained enable it to be possible to promptly diagnose and perform certain therapeutic measures as well as reduce the time of data analysis and increase the efficiency of diagnosing cattle. The scientific novelty of this study is the creation of an algorithm for building a knowledge base and improving the algorithm for constructing the Sugeno fuzzy logic model for diagnosing diseases in cattle. The findings of this study can be widely used in veterinary medicine in solving the problems of diagnosing diseases in cattle and substantiating decision-making in intelligent systems.

An Improved Forest Fire Detection Method Based on the Detectron2 Model and a Deep Learning Approach

With an increase in both global warming and the human population, forest fires have become a major global concern. This can lead to climatic shifts and the greenhouse effect, among other adverse outcomes. Surprisingly, human activities have caused a disproportionate number of forest fires. Fast detection with high accuracy is the key to controlling this unexpected event. To address this, we proposed an improved forest fire detection method to classify fires based on a new version of the Detectron2 platform (a ground-up rewrite of the Detectron library) using deep learning approaches. Furthermore, a custom dataset was created and labeled for the training model, and it achieved higher precision than the other models. This robust result was achieved by improving the Detectron2 model in various experimental scenarios with a custom dataset and 5200 images. The proposed model can detect small fires over long distances during the day and night. The advantage of using the Detectron2 algorithm is its long-distance detection of the object of interest. The experimental results proved that the proposed forest fire detection method successfully detected fires with an improved precision of 99.3%.

Masked Face Emotion Recognition Based on Facial Landmarks and Deep Learning Approaches for Visually Impaired People

Current artificial intelligence systems for determining a person's emotions rely heavily on lip and mouth movement and other facial features such as eyebrows, eyes, and the forehead. Furthermore, low-light images are typically classified incorrectly because of the dark region around the eyes and eyebrows. In this work, we propose a facial emotion recognition method for masked facial images using low-light image enhancement and feature analysis of the upper features of the face with a convolutional neural network. The proposed approach employs the AffectNet image dataset, which includes eight types of facial expressions and 420,299 images. Initially, the facial input image's lower parts are covered behind a synthetic mask. Boundary and regional representation methods are used to indicate the head and upper features of the face. Secondly, we effectively adopt a facial landmark detection method-based feature extraction strategy using the partially covered masked face's features. Finally, the features, the coordinates of the landmarks that have been identified, and the histograms of the oriented gradients are then incorporated into the classification procedure using a convolutional neural network. An experimental evaluation shows that the proposed method surpasses others by achieving an accuracy of 69.3% on the AffectNet dataset.

Improved Face Detection Method via Learning Small Faces on Hard Images Based on a Deep Learning Approach

Most facial recognition and face analysis systems start with facial detection. Early techniques, such as Haar cascades and histograms of directed gradients, mainly rely on features that had been manually developed from particular images. However, these techniques are unable to correctly synthesize images taken in untamed situations. However, deep learning's quick development in computer vision has also sped up the development of a number of deep learning-based face detection frameworks, many of which have significantly improved accuracy in recent years. When detecting faces in face detection software, the difficulty of detecting small, scale, position, occlusion, blurring, and partially occluded faces in uncontrolled conditions is one of the problems of face identification that has been explored for many years but has not yet been entirely resolved. In this paper, we propose Retina net baseline, a single-stage face detector, to handle the challenging face detection problem. We made network improvements that boosted detection speed and accuracy. In Experiments, we used two popular datasets, such as WIDER FACE and FDDB. Specifically, on the WIDER FACE benchmark, our proposed method achieves AP of 41.0 at speed of 11.8 FPS with a single-scale inference strategy and AP of 44.2 with multi-scale inference strategy, which are results among one-stage detectors. Then, we trained our model during the implementation using the PyTorch framework, which provided an accuracy of 95.6% for the faces, which are successfully detected. Visible experimental results show that our proposed model outperforms seamless detection and recognition results achieved using performance evaluation matrices.

Improved Agricultural Field Segmentation in Satellite Imagery Using TL-ResUNet Architecture

Currently, there is a growing population around the world, and this is particularly true in developing countries, where food security is becoming a major problem. Therefore, agricultural land monitoring, land use classification and analysis, and achieving high yields through efficient land use are important research topics in precision agriculture. Deep learning-based algorithms for the classification of satellite images provide more reliable and accurate results than traditional classification algorithms. In this study, we propose a transfer learning based residual UNet architecture (TL-ResUNet) model, which is a semantic segmentation deep neural network model of land cover classification and segmentation using satellite images. The proposed model combines the strengths of residual network, transfer learning, and UNet architecture. We tested the model on public datasets such as DeepGlobe, and the results showed that our proposed model outperforms the classic models initiated with random weights and pre-trained ImageNet coefficients. The TL-ResUNet model outperforms other models on several metrics commonly used as accuracy and performance measures for semantic segmentation tasks. Particularly, we obtained an IoU score of 0.81 on the validation subset of the DeepGlobe dataset for the TL-ResUNet model.

An Algorithm with Iteration Uncertainty Eliminate Based on Geomagnetic Fingerprint under Mobile Edge Computing for Indoor Localization

Indoor localization problems are difficult due to that the information, such as WLAN and GPS, cannot achieve enough precision for indoor issues. This paper presents a novel indoor localization algorithm, GeoLoc, with uncertainty eliminate based on fusion of acceleration, angular rate, and magnetic field sensor data. The algorithm can be deployed in edge devices to overcome the problems of insufficient computing resources and long delay caused by high complexity of location calculation. Firstly, the magnetic map is built and magnetic values are matched. Secondly, orientation updating and position selection are iteratively executed using the fusion data, which gradually reduce uncertainty of orientation. Then, we filter the trajectory from a path set. By gradually reducing uncertainty, GeoLoc can bring a high positioning precision and a smooth trajectory. In addition, this method has an advantage in that it does not rely on any infrastructure such as base stations and beacons. It solves the common problems regarding the non-uniqueness of the geomagnetic fingerprint and the deviation of the sensor measurement. The experimental results show that our algorithm achieves an accuracy of less than 2.5 m in indoor environment, and the positioning results are relatively stable. It meets the basic requirements of indoor location-based services (LBSs).

Development of Real-Time Landmark-Based Emotion Recognition CNN for Masked Faces

Owing to the availability of a wide range of emotion recognition applications in our lives, such as for mental status calculation, the demand for high-performance emotion recognition approaches remains uncertain. Nevertheless, the wearing of facial masks has been indispensable during the COVID-19 pandemic. In this study, we propose a graph-based emotion recognition method that adopts landmarks on the upper part of the face. Based on the proposed approach, several pre-processing steps were applied. After pre-processing, facial expression features need to be extracted from facial key points. The main steps of emotion recognition on masked faces include face detection by using Haar-Cascade, landmark implementation through a media-pipe face mesh model, and model training on seven emotional classes. The FER-2013 dataset was used for model training. An emotion detection model was developed for non-masked faces. Thereafter, landmarks were applied to the upper part of the face. After the detection of faces and landmark locations were extracted, we captured coordinates of emotional class landmarks and exported to a comma-separated values (csv) file. After that, model weights were transferred to the emotional classes. Finally, a landmark-based emotion recognition model for the upper facial parts was tested both on images and in real time using a web camera application. The results showed that the proposed model achieved an overall accuracy of 91.2% for seven emotional classes in the case of an image application. Image based emotion detection of the proposed model accuracy showed relatively higher results than the real-time emotion detection.

Improved Classification Approach for Fruits and Vegetables Freshness Based on Deep Learning

Classification of fruit and vegetable freshness plays an essential role in the food industry. Freshness is a fundamental measure of fruit and vegetable quality that directly affects the physical health and purchasing motivation of consumers. In addition, it is a significant determinant of market price; thus, it is imperative to study the freshness of fruits and vegetables. Owing to similarities in color, texture, and external environmental changes, such as shadows, lighting, and complex backgrounds, the automatic recognition and classification of fruits and vegetables using machine vision is challenging. This study presents a deep-learning system for multiclass fruit and vegetable categorization based on an improved YOLOv4 model that first recognizes the object type in an image before classifying it into one of two categories: fresh or rotten. The proposed system involves the development of an optimized YOLOv4 model, creating an image dataset of fruits and vegetables, data argumentation, and performance evaluation. Furthermore, the backbone of the proposed model was enhanced using the Mish activation function for more precise and rapid detection. Compared with the previous YOLO series, a complete experimental evaluation of the proposed method can obtain a higher average precision than the original YOLOv4 and YOLOv3 with 50.4%, 49.3%, and 41.7%, respectively. The proposed system has outstanding prospects for the construction of an autonomous and real-time fruit and vegetable classification system for the food industry and marketplaces and can also help visually impaired people to choose fresh food and avoid food poisoning.

Improved Feature Parameter Extraction from Speech Signals Using Machine Learning Algorithm

Speech recognition refers to the capability of software or hardware to receive a speech signal, identify the speaker's features in the speech signal, and recognize the speaker thereafter. In general, the speech recognition process involves three main steps: acoustic processing, feature extraction, and classification/recognition. The purpose of feature extraction is to illustrate a speech signal using a predetermined number of signal components. This is because all information in the acoustic signal is excessively cumbersome to handle, and some information is irrelevant in the identification task. This study proposes a machine learning-based approach that performs feature parameter extraction from speech signals to improve the performance of speech recognition applications in real-time smart city environments. Moreover, the principle of mapping a block of main memory to the cache is used efficiently to reduce computing time. The block size of cache memory is a parameter that strongly affects the cache performance. In particular, the implementation of such processes in real-time systems requires a high computation speed. Processing speed plays an important role in speech recognition in real-time systems. It requires the use of modern technologies and fast algorithms that increase the acceleration in extracting the feature parameters from speech signals. Problems with overclocking during the digital processing of speech signals have yet to be completely resolved. The experimental results demonstrate that the proposed method successfully extracts the signal features and achieves seamless classification performance compared to other conventional speech recognition algorithms.

Improved Real-Time Fire Warning System Based on Advanced Technologies for Visually Impaired People

Early fire detection and notification techniques provide fire prevention and safety information to blind and visually impaired (BVI) people within a short period of time in emergency situations when fires occur in indoor environments. Given its direct impact on human safety and the environment, fire detection is a difficult but crucial problem. To prevent injuries and property damage, advanced technology requires appropriate methods for detecting fires as quickly as possible. In this study, to reduce the loss of human lives and property damage, we introduce the development of the vision-based early flame recognition and notification approach using artificial intelligence for assisting BVI people. The proposed fire alarm control system for indoor buildings can provide accurate information on fire scenes. In our proposed method, all the processes performed manually were automated, and the performance efficiency and quality of fire classification were improved. To perform real-time monitoring and enhance the detection accuracy of indoor fire disasters, the proposed system uses the YOLOv5m model, which is an updated version of the traditional YOLOv5. The experimental results show that the proposed system successfully detected and notified the occurrence of catastrophic fires with high speed and accuracy at any time of day or night, regardless of the shape or size of the fire. Finally, we compared the competitiveness level of our method with that of other conventional fire-detection methods to confirm the seamless classification results achieved using performance evaluation matrices.

Attention 3D U-Net with Multiple Skip Connections for Segmentation of Brain Tumor Images

Among researchers using traditional and new machine learning and deep learning techniques, 2D medical image segmentation models are popular. Additionally, 3D volumetric data recently became more accessible, as a result of the high number of studies conducted in recent years regarding the creation of 3D volumes. Using these 3D data, researchers have begun conducting research on creating 3D segmentation models, such as brain tumor segmentation and classification. Since a higher number of crucial features can be extracted using 3D data than 2D data, 3D brain tumor detection models have increased in popularity among researchers. Until now, various significant research works have focused on the 3D version of the U-Net and other popular models, such as 3D U-Net and V-Net, while doing superior research works. In this study, we used 3D brain image data and created a new architecture based on a 3D U-Net model that uses multiple skip connections with cost-efficient pretrained 3D MobileNetV2 blocks and attention modules. These pretrained MobileNetV2 blocks assist our architecture by providing smaller parameters to maintain operable model size in terms of our computational capability and help the model to converge faster. We added additional skip connections between the encoder and decoder blocks to ease the exchange of extracted features between the two blocks, which resulted in the maximum use of the features. We also used attention modules to filter out irrelevant features coming through the skip connections and, thus, preserved more computational power while achieving improved accuracy.

Augmentation Method for High Intra-Class Variation Data in Apple Detection

Deep learning is widely used in modern orchard production for various inspection missions, which helps improve the efficiency of orchard operations. In the mission of visual detection during fruit picking, most current lightweight detection models are not yet effective enough to detect multi-type occlusion targets, severely affecting automated fruit-picking efficiency. This study addresses this problem by proposing the pioneering design of a multi-type occlusion apple dataset and an augmentation method of data balance. We divided apple occlusion into eight types and used the proposed method to balance the number of annotation boxes for multi-type occlusion apple targets. Finally, a validation experiment was carried out using five popular lightweight object detection models: yolox-s, yolov5-s, yolov4-s, yolov3-tiny, and efficidentdet-d0. The results show that, using the proposed augmentation method, the average detection precision of the five popular lightweight object detection models improved significantly. Specifically, the precision increased from 0.894 to 0.974, recall increased from 0.845 to 0.972, and mAP0.5 increased from 0.982 to 0.919 for yolox-s. This implies that the proposed augmentation method shows great potential for different fruit detection missions in future orchard applications.

A Sample Balance-Based Regression Module for Object Detection in Construction Sites

Object detection plays an important role in safety monitoring, quality control, and productivity management at construction sites. Currently, the dominant method for detection is deep neural networks (DNNs), and the state-of-the-art object detectors rely on a bounding box regression (BBR) module to localize objects. However, the detection results suffer from a bounding box redundancy problem, which is caused by inaccurate BBR. In this paper, we propose an improvement of the object detection regression module for the bounding box redundancy problem. The inaccuracy of BBR in the detection results is caused by the imbalance between the hard and easy samples in the BBR process, i.e., the number of easy samples with small regression errors is much smaller than the hard samples. Therefore, the strategy of balancing hard and easy samples is introduced into the EIOU (Efficient Intersection over Union) loss and FocalL1 regression loss function, respectively, and the two are combined as the new regression loss function, namely EFocalL1-SEIOU (Efficient FocalL1-Segmented Efficient Intersection over Union) loss. Finally, the proposed EFocalL1-SEIOU loss is evaluated on four different DNN-based detectors based on the MOCS (Moving Objects in Construction Sites) dataset in construction sites. The experimental results show that the EFocalL1- SEIOU loss improves the detection ability of objects on different detectors at construction sites.