Single Shot Multibox Detector Model Research Articles

Novel intelligent defects detection of boiler water walls in thermal power plants based on OFM_SSD

The boiler is a critical component of conventional thermal power systems, where surface flaws in boiler water walls can significantly compromise safety and availability, potentially leading to substantial loss of life and property. Traditional detection methods, whether manual or based on machine learning, often prove costly, inefficient and time-consuming, failing to meet the stringent requirements for water wall inspection. Therefore, a novel surface defect detection model integrating an improved single shot multibox detector (SSD) with the optical flow method (OFM) (OFM_SSD) is proposed. The OFM enhances data sample diversity by augmenting the dataset derived from thermal power plants, while the incorporation of deconvolution techniques improves the model receptive field, thereby enhancing its ability to detect and classify small defects. Comprehensive experiments demonstrate that the OFM_SSD outperforms several existing models including the SSD model based on traditional expanded datasets (T_SSD), you only look once (YOLO), ordinary SSD, Regions with the CNN(R_CNN), and Deconvolution-only SSD (DSSD) in terms of accuracy in defect localization and classification. This advancement of the OFM_SSD not only reduces operational costs but also enhances detection capabilities, ultimately contributing to safer and more efficient operations within thermal power plants.

AlertTrap: A Study on Object Detection in Remote Insect Trap Monitoring System Using on the Edge Deep Learning Platform

Fruit flies are one of the most harmful insect species to fruit yields. In AlertTrap, implementation of Single-Shot Multibox Detector (SSD) architecture with different state-of-the-art backbone feature extractors such as MobileNetV1 and MobileNetV2 appears to be potential solutions for the real-time detection problem. SSD-MobileNetV1 and SSD-MobileNetV2 perform well and result in AP at 0.5 of 0.957 and 1.0, respectively. You Only Look Once (YOLO) v4-tiny outperforms the SSD family with 1.0 in AP at 0.5; however, its throughput velocity is considerably slower, which shows SSD models are better candidates for real-time implementation. We also tested the models with synthetic test sets simulating expected environmental disturbances. The YOLOv4-tiny had better tolerance to these disturbances than the SSD models. The Raspberry Pi system successfully gathered environmental data and pest counts, sending them via email over 4 G. However, running the full YOLO version in real time on Raspberry Pi is not feasible, indicating the need for a lighter object detection algorithm for future research. Among model candidates, YOLOv4-tiny generally performs best, with SSD MobileNetV2 also comparable and sometimes better, especially in scenarios with synthetic disturbances. SSD models excel in processing time, enabling real-time, high-accuracy detection. TFLITE versions of SSD models also process faster than their inference graph on TPU hardware, suggesting real-time implementation on edge devices like the Google Coral Dev Board. The results demonstrate the feasibility of real-time implementation of the fruit fly detection models on edge devices with high performance. In addition, YOLOv4-tiny is shown to be the most probable candidate because YOLOv4-tiny demonstrates a robust testing performance toward citrus fruit fly detection. Nevertheless, SSD-MobileNetV2 will be the better model, considering the inference time.

Dual attention-enhanced SSD: A novel deep learning model for object detection

Object detection is a fundamental task in computer vision with significant implications across various applications, including autonomous driving, surveillance, and image understanding. The accurate and efficient detection of objects within images is crucial for enabling machines to interpret visual information and make informed decisions. In this paper, we present an enhanced version of the Single Shot MultiBox Detector (SSD) for object detection, leveraging the concept of dual attention mechanisms. Our proposed approach, named SSD-Dual Attention, integrates dual attention layers into the SSD framework. These dual attention layers are strategically positioned between feature maps and prediction convolutions, enhancing the model's ability to capture informative feature representations across a wide range of object scales and backgrounds. Experimental results on the PASCAL VOC 2007 and 2012 datasets validate the effectiveness of our approach. Notably, SSD-Dual Attention achieves an impressive mean Average Precision (mAP) of 78.1%, surpassing the performance of SSD models enhanced with attention mechanisms such as SSD-ECA, SSD-CBAM, SSD-Non-local attention and SSD-SE attention, as well as the original SSD. These results underscore the enhanced accuracy and precision of our object detection system, marking a substantial advancement in the field of computer vision. Code is available at https://github.com/AlexHunterLeo/Dual-attention-Enhanced-SSD-A-Novel-Deep-Learning-Model-for-Object-Detection

Lightweight Small-Tailed Han Sheep Facial Recognition Based on Improved SSD Algorithm

We propose a lightweight detection algorithm based on the Single Shot MultiBox Detector (SSD) algorithm in order to facilitate sheep management and to realize sheep facial identification, and we take the self-constructed dataset as the research object. First, the SSD replaces the VGG16 backbone network with MobileNetv3, a lightweight neural network, to create a hybrid model that is much smaller. Second, the ECA attention mechanism is incorporated into the backend of the 72 × 160 bottleneck layer. Finally, the SmoothL1 loss function is substituted with the BalancedL1 loss function. The optimized model’s size decreases significantly from the original SSD’s 132 MB to just 22.4 MB. It achieves a mean average precision of 83.47% and maintains an average frame rate of 68.53 frames per second. Compared to the basic SSD model, the mean average precision has increased by 3.25 percentage points, the model size has decreased by 109.6 MB, and the detection speed has improved by 9.55 frames per second. In comparative experiments using the same dataset with different object detection models, the proposed model outperforms the SSD, Faster R-CNN, Retinanet, and CenterNet in terms of mean average precision, with improvements of 3.25 percentage points, 4.71 percentage points, 2.38 percentage points, and 8.13 percentage points, respectively. The detection speed has shown significant improvements, increasing by 9.55, 58.55, 53.1, and 12.37 frames per second, respectively. The improved model presented in this paper significantly reduces the model’s size and computational requirements while maintaining an excellent performance. This provides a valuable reference for the digitalization of animal husbandry and livestock farming.

Automated Recognition and Tracking of Stationary and Moving Cars in Images and Videos: YOLOv5 and SSD Analysis

The detection and tracking of cars are a significant and useful aspect of traffic surveillance systems, which is essential for the efficient management of traffic and the security of drivers and passengers. The primary objectives of this investigation are vehicle detection and tracking. The automatic detection of cars in both still photos and video recordings are the main topic of this investigation. One of the many uses for Deep Learning, which combines fuzzy logic, neural networks, and evolutionary algorithms, is the detection and tracking of moving objects. In this work, the key object detection techniques YOLOv5 and SSD were used to analyse vehicle recognition and tracking with deep learning. The Single Shot Multi-Box Detector model architecture is then utilized as the main foundation for the detection of vehicles. The vehicle recognition model is then trained using the YOLOv5 and SSD algorithms, each of which contributes to the illustration of the detection effect. Comparing the detection rates obtained by both models on a range of cars is necessary to locate it. The objective of this work is to create an automated system for locating and tracking both moving and stationary cars in still images and moving movies. According to the research, using this method has improved the success rate of recognizing automobiles to 97.65%.

Improved Deep Learning Model for Workpieces of Rectangular Pipeline Surface Defect Detection

This study introduces a novel approach to address challenges in workpiece surface defect identification. It presents an enhanced Single Shot MultiBox Detector model, incorporating attention mechanisms and multi-feature fusion. The research methodology involves carefully curating a dataset from authentic on-site factory production, enabling the training of a model with robust real-world generalization. Leveraging the Single Shot MultiBox Detector model lead to improvements integrating channel and spatial attention mechanisms in the feature extraction network. Diverse feature extraction methods enhance the network’s focus on crucial information, improving its defect detection efficacy. The proposed model achieves a significant Mean Average Precision (mAP) improvement, reaching 99.98% precision, a substantial 3% advancement over existing methodologies. Notably, the proposed model exhibits a tendency for the values of the P-R curves in object detection for each category to approach 1, which allows a better balance between the requirements of real-time detection and precision. Within the threshold range of 0.2 to 1, the model maintains a stable level of precision, consistently remaining between 0.99 and 1. In addition, the average running speed is 2 fps lower compared to other models, and the reduction in detection speed after the model improvement is kept within 1%. The experimental results indicate that the model excels in pixel-level defect identification, which is crucial for precise defect localization. Empirical experiments validate the algorithm’s superior performance. This research represents a pivotal advancement in workpiece surface defect identification, combining technological innovation with practical efficacy.

Electrolytic capacitor surface defect detection based on deep convolution neural network

The existing methods for detecting surface defects in electrolytic capacitors are typically based on conventional machine vision, with limited feature extraction capabilities, poor versatility, slow detection speed, and the inability to achieve accurate and real-time defect detection. In this study, a real-time object detection algorithm based on an improved single shot multibox detector (SSD) is proposed to achieve omnidirectional surface defect detection of electrolytic capacitors. First, an electrolytic capacitor surface image acquisition device was established to capture omnidirectional surface images of the capacitors, and an electrolytic capacitor surface defect dataset was created. Next, the visual geometry group (VGG)-16 network structure was replaced with the MobileNetv2 network structure, effectively reducing the model’s parameter count and improving inference speed. Moreover, the Multibox Loss function was replaced with the Focal Loss function to increase the model’s attention to difficult-to-classify samples and improve model accuracy. Additionally, a transfer learning network model was designed to apply the model to electrolytic capacitors of different colors using small sample learning. Finally, the performance of the improved network model was tested on a dataset of electrolytic capacitor surface defects. The experimental results demonstrate that the parameters quantity of improved model is 3.50 M, the mAP value reaches 92.67 %, which is improved by 2.54 %, and the Macro-F1 value reaches 92.15 %, which is 11.32 % higher than that before improvement. Thus, the proposed improved SSD model provides a theoretical basis and technical prerequisites for automated and intelligent surface defect detection in electrolytic capacitors.

A high-precision jujube disease spot detection based on SSD during the sorting process.

The development of automated grading equipment requires achieving high throughput and precise detection of disease spots on jujubes. However, the current algorithms are inadequate in accomplishing these objectives due to their high density, varying sizes and shapes, and limited location information regarding disease spots on jujubes. This paper proposes a method called JujubeSSD, to boost the precision of identifying disease spots in jujubes based on a single shot multi-box detector (SSD) network. In this study, a diverse dataset comprising disease spots of varied sizes and shapes, varying densities, and multiple location details on jujubes was created through artificial collection and data augmentation. The parameter information obtained from transfer learning into the backbone feature extraction network of the SSD model, which reduced the time of spot detection to 0.14 s. To enhance the learning of target detail features and improve the recognition of weak information, the traditional convolution layer was replaced with deformable convolutional networks (DCNs). Furthermore, to address the challenge of varying sizes and shapes of disease spot regions on jujubes, the path aggregation feature pyramid network (PAFPN) and balanced feature pyramid (BFP) were integrated into the SSD network. Experimental results demonstrate that the mean average precision at the IoU (intersection over union) threshold of 0.5 (mAP@0.5) of JujubeSSD reached 97.1%, representing an improvement of approximately 6.35% compared to the original algorithm. When compared to existing algorithms, such as YOLOv5 and Faster R-CNN, the improvements in mAP@0.5 were 16.84% and 8.61%, respectively. Therefore, the proposed method for detecting jujube disease spot achieves superior performance in jujube surface disease detection and meets the requirements for practical application in agricultural production.

Enhancing artistic analysis through deep learning: a graphic art element recognition model based on SSD and FPT.

For the analysis of art works, accurate identification of various elements of works through deep learning methods is helpful for artists to appreciate and learn works. In this study, we leverage deep learning methodologies to precisely identify the diverse elements within graphic art designs, aiding artists in their appreciation and learning process. Our approach involves integrating the attention mechanism into an enhanced Single Shot MultiBox Detector (SSD) model to refine the recognition of artistic design elements. Additionally, we improve the feature fusion structure of the SSD model by incorporating long-range attention mechanism information, thus enhancing target detection accuracy. Moreover, we refine the Feature Pyramid Transformer (FPT) attention mechanism model to ensure the output feature map aligns effectively with the requirements of object detection. Our empirical findings demonstrate that our refined approach outperforms the original SSD algorithm across all four evaluation metrics, exhibiting improvements of 1.52%, 1.89%, 3.09%, and 2.57%, respectively. Qualitative tests further illustrate the accuracy, robustness, and universality of our proposed method, particularly in scenarios characterized by dense artistic elements and challenging-to-distinguish categories within art compositions.

Classification and recognition of the Nantong blue calico pattern based on deep learning

The Nantong blue calico pattern is a significant and indispensable part of China’s intangible cultural heritage, representing an artistic form of weaving and dyeing. However, existing research on blue calico patterns is not extensive, and few studies have focused on the construction of a database categorizing them or on recognizing the Nantong blue calico pattern. Obtaining good efficiency and accuracy through manual recognition has been the primary challenge in recognizing the Nantong blue calico pattern. In light of these challenges, this study proposes the use of deep learning network model to intelligently classify and recognize blue calico patterns.First, the patterns are classified to establish a Nantong blue calico pattern database, and the corresponding category labels are then manually assigned to each image. Second, based on the database and a backbone feature extraction network, the abilities of SSD (Single Shot Multibox Detector), Faster RCNN (Region-CNN), and You Only Look Once (to recognize the Nantong blue calico pattern were compared. The results show that the SSD model based on a VGG (Visual Geometry Group) backbone network has the best recognition accuracy of these three algorithms, with an average accuracy of 79.42%. On this basis, we selected the SSD model for parameter optimization and adjustment, and we replaced the backbone with mobilenetv2, a lighter backbone extraction network, to recognize the Nantong blue calico pattern. The results show that compared with the original SSD model, the optimized SSD model can improve the pattern recognition rate of Nantong blue calico pattern. Furthermore, this paper makes use of the characteristics of the VGG deep network, the backbone network of the SSD model, to efficiently extract the features of blue calico patterns, which provides a basis for designers to design innovative blue calico patterns.

Enhanced SSD framework for detecting defects in cigarette appearance using variational Bayesian inference under limited sample conditions.

In high-speed cigarette manufacturing industries, occasional minor cosmetic cigarette defects and a scarcity of samples significantly hinder the rapid and accurate detection of defects. To tackle this challenge, we propose an enhanced single-shot multibox detector (SSD) model that uses variational Bayesian inference for improved detection of tiny defects given sporadic occurrences and limited samples. The enhanced SSD model incorporates a bounded intersection over union (BIoU) loss function to reduce sensitivity to minor deviations and uses exponential linear unit (ELU) and leaky rectified linear unit (ReLU) activation functions to mitigate vanishing gradients and neuron death in deep neural networks. Empirical results show that the enhanced SSD300 and SSD512 models increase the model's detection accuracy mean average precision (mAP) by up to 1.2% for small defects. Ablation studies further reveal that the model's mAP increases by 1.5%, which reduces the computational requirements by 5.92 GFLOPs. The model also shows improved inference in scenarios with limited samples, thus highlighting its effectiveness and applicability in high-speed, precision-oriented cigarette manufacturing industries.

Neural network models of apple fruit identification in tree crowns: comparative analysis

The article presents the results of an analysis conducted from 2022 to 2023 to assess the quality of modern neural network models of apple fruit identification in tree crowns shown in images. In order to conduct the studies on identifying the best detector, the following neural networks were used: SSD (Single Shot MultiBox Detector), YOLOv4 (You Only Look Once, Version 4), YOLOv5, YOLOv7, and YOLOv8. The performance of the considered models of apple fruit identification was assessed using such binary classification metrics as precision, recall, accuracy, F-score, and AUC-ROCTotal (area under the curve). To assess the accuracy in predicting apple fruit identification, the mean absolute percentage error (MAPE) of the analyzed neural network models was calculated. The neural network performance analysis used 300 photographs taken at an apple garden. The conducted studies revealed that the SSD model provides lower speed and accuracy, as well as having high requirements for computing resources, which may limit its use in lower performance devices. The YOLOv4 model surpasses the YOLOv5 model in terms of accuracy by 10.2 %, yet the processing speed of the YOLOv5 model is over twice that of the YOLOv4 model. This fact makes the YOLOv5 model preferable for tasks related to real-time big data processing. The YOLOv8 model is superior to the YOLOv7 model in terms of speed (by 37.3 %); however, the accuracy of the YOLOv7 model is 9.4 % higher. The highest area under the Precision-Recall curve amounts to 0.94 when using the YOLOv7 model. This fact suggests a high probability that the classifier can accurately distinguish between the positive and negative values of the apple fruit class. MAPE calculation for the analyzed neural network models showed that the lowest error in apple fruit identification amounted to 5.64 % for the YOLOv7 model as compared to the true value determined using the visual method. The performance analysis of modern neural network models shows that the YOLO family of neural networks provides high speed and accuracy of object detection, which allows them to operate in real time. The use of transfer learning (tuning of only the last layers to solve highly specialized problems) to adjust the performance of models for different apple fruit varieties can further improve the accuracy of apple fruit identification.

Improving resource recycling based on deep learning

The manual sorting of recyclable garbage has caused several issues such as the wastage of human resources and low resource utilization. To solve this problem, an improved Single Shot Multibox Detector (SSD) deep learning approach has been developed for recyclable garbage detection. To reduce the number of parameters and make the model easier to deploy and apply, a lightweight network called RepVGG has been chosen to replace the VGG16 network in the SSD. Additionally, the auxiliary convolutional layer structure of the SSD has been modified to further reduce the number of parameters. Additionally, the SK module has been integrated to adaptively adjust the size of the receptive field and enhance the detection accuracy. Experimental results of Waste Classification data set from Kaggle website have demonstrated that the improved SSD model has better detection accuracy and real-time performance, with an accuracy of 95.23%, which is 4.33 percentage points higher than the original SSD, and a detection speed of up to 64 FPS. This algorithm can be better applied in industry.

Research on the Efficiency of Bridge Crack Detection by Coupling Deep Learning Frameworks with Convolutional Neural Networks.

Bridge crack detection based on deep learning is a research area of great interest and difficulty in the field of bridge health detection. This study aimed to investigate the effectiveness of coupling a deep learning framework (DLF) with a convolutional neural network (CNN) for bridge crack detection. A dataset consisting of 2068 bridge crack images was randomly split into training, verification, and testing sets with a ratio of 8:1:1, respectively. Several CNN models, including Faster R-CNN, Single Shot MultiBox Detector (SSD), You Only Look Once (YOLO)-v5(x), U-Net, and Pyramid Scene Parsing Network (PSPNet), were used to conduct experiments using the PyTorch, TensorFlow2, and Keras frameworks. The experimental results show that the Harmonic Mean (F1) values of the detection results of the Faster R-CNN and SSD models under the Keras framework are relatively large (0.76 and 0.67, respectively, in the object detection model). The YOLO-v5(x) model of the TensorFlow2 framework achieved the highest F1 value of 0.67. In semantic segmentation models, the U-Net model achieved the highest detection result accuracy (AC) value of 98.37% under the PyTorch framework. The PSPNet model achieved the highest AC value of 97.86% under the TensorFlow2 framework. These experimental results provide optimal coupling efficiency parameters of a DLF and CNN for bridge crack detection. A more accurate and efficient DLF and CNN model for bridge crack detection has been obtained, which has significant practical application value.

Transfer Learning-Based Lightweight SSD Model for Detection of Pests in Citrus

In citrus cultivation, it is a difficult task for farmers to classify different pests correctly and make proper decisions to prevent citrus damage. This work proposes an efficient modified lightweight transfer learning model which combines the effectiveness and accuracy of citrus pest characterization with mobile terminal counting. Firstly, we utilized typical transfer learning feature extraction networks such as ResNet50, InceptionV3, VGG16, and MobileNetV3, and pre-trained the single-shot multibox detector (SSD) network to compare and analyze the classification accuracy and efficiency of each model. Then, to further reduce the amount of calculations needed, we miniaturized the prediction convolution kernel at the end of the model and added a residual block of a 1 × 1 convolution kernel to predict category scores and frame offsets. Finally, we transplanted the preferred lightweight SSD model into the mobile terminals developed by us to verify its usability. Compared to other transfer learning models, the modified MobileNetV3+RPBM can enable the SSD network to achieve accurate detection of Panonychus Citri Mcgregor and Aphids, with a mean average precision (mAP) up to 86.10% and the counting accuracy reaching 91.0% and 89.0%, respectively. In terms of speed, the mean latency of MobileNetV3+RPBM is as low as 185 ms. It was concluded that this novel and efficient modified MobileNetV3+RPBM+SSD model is effective at classifying citrus pests, and can be integrated into devices that are embedded for mobile rapid detection as well as for counting pests in citrus orchards. The work presented herein can help encourage farm managers to judge the degree of pest damage and make correct decisions regarding pesticide application in orchard management.

Diagnostic accuracy of a deep learning model using YOLOv5 for detecting developmental dysplasia of the hip on radiography images

Developmental dysplasia of the hip (DDH) is a cluster of hip development disorders and one of the most common hip diseases in infants. Hip radiography is a convenient diagnostic tool for DDH, but its diagnostic accuracy is dependent on the interpreter’s level of experience. The aim of this study was to develop a deep learning model for detecting DDH. Patients younger than 12 months who underwent hip radiography between June 2009 and November 2021 were selected. Using their radiography images, transfer learning was performed to develop a deep learning model using the “You Only Look Once” v5 (YOLOv5) and single shot multi-box detector (SSD). A total of 305 anteroposterior hip radiography images (205 normal and 100 DDH hip images) were collected. Of these, 30 normal and 17 DDH hip images were used as the test dataset. The sensitivity and the specificity of our best YOLOv5 model (YOLOv5l) were 0.94 (95% confidence interval [CI] 0.73–1.00) and 0.96 (95% CI 0.89–0.99), respectively. This model also outperformed the SSD model. This is the first study to establish a model for detecting DDH using YOLOv5. Our deep learning model provides good diagnostic performance for DDH. We believe our model is a useful diagnostic assistant tool.

Hazard source detection of longitudinal tearing of conveyor belt based on deep learning.

Belt tearing is the main safety accident of belt conveyor. The main cause of tearing is the doped bolt and steel in the conveying belt. In this paper, the bolt and steel are identified as the Hazard source of tear. In this paper, bolt and steel are defined as the risk sources of tearing. Effective detection of the source of danger can effectively prevent the occurrence of conveyor belt tearing accidents. Here we use deep learning to detect the hazard source image. We improved on the SSD(Single Shot MultiBox Detector) model. Replace the original backbone network with an improved Shufflenet_V2, and replace the original position loss function with the CIoU loss function. Additionally, it compares this new approach to previous methods. The proposed model has surpassed other state-of-art methods with more than 94% accuracy. In addition, when deployed without GPU acceleration, the detection speed can reach 20fps. It can meet the requirements of real-time detection. The experimental results show that the proposed model can realize the online detection of hazard sources, so as to prevent longitudinal tearing of conveyor belt.

Dual-View Single-Shot Multibox Detector at Urban Intersections: Settings and Performance Evaluation

The explosion of artificial intelligence methods has paved the way for more sophisticated smart mobility solutions. In this work, we present a multi-camera video content analysis (VCA) system that exploits a single-shot multibox detector (SSD) network to detect vehicles, riders, and pedestrians and triggers alerts to drivers of public transportation vehicles approaching the surveilled area. The evaluation of the VCA system will address both detection and alert generation performance by combining visual and quantitative approaches. Starting from a SSD model trained for a single camera, we added a second one, under a different field of view (FOV) to improve the accuracy and reliability of the system. Due to real-time constraints, the complexity of the VCA system must be limited, thus calling for a simple multi-view fusion method. According to the experimental test-bed, the use of two cameras achieves a better balance between precision (68%) and recall (84%) with respect to the use of a single camera (i.e., 62% precision and 86% recall). In addition, a system evaluation in temporal terms is provided, showing that missed alerts (false negatives) and wrong alerts (false positives) are typically transitory events. Therefore, adding spatial and temporal redundancy increases the overall reliability of the VCA system.

A Multiscale Polyp Detection Approach for GI Tract Images Based on Improved DenseNet and Single-Shot Multibox Detector

Small bowel polyps exhibit variations related to color, shape, morphology, texture, and size, as well as to the presence of artifacts, irregular polyp borders, and the low illumination condition inside the gastrointestinal GI tract. Recently, researchers developed many highly accurate polyp detection models based on one-stage or two-stage object detector algorithms for wireless capsule endoscopy (WCE) and colonoscopy images. However, their implementation requires a high computational power and memory resources, thus sacrificing speed for an improvement in precision. Although the single-shot multibox detector (SSD) proves its effectiveness in many medical imaging applications, its weak detection ability for small polyp regions persists due to the lack of information complementary between features of low- and high-level layers. The aim is to consecutively reuse feature maps between layers of the original SSD network. In this paper, we propose an innovative SSD model based on a redesigned version of a dense convolutional network (DenseNet) which emphasizes multiscale pyramidal feature maps interdependence called DC-SSDNet (densely connected single-shot multibox detector). The original backbone network VGG-16 of the SSD is replaced with a modified version of DenseNet. The DenseNet-46 front stem is improved to extract highly typical characteristics and contextual information, which improves the model's feature extraction ability. The DC-SSDNet architecture compresses unnecessary convolution layers of each dense block to reduce the CNN model complexity. Experimental results showed a remarkable improvement in the proposed DC-SSDNet to detect small polyp regions achieving an mAP of 93.96%, F1-score of 90.7%, and requiring less computational time.

Computer Vision and Deep Learning as Tools for Leveraging Dynamic Phenological Classification in Vegetable Crops

The efficiency of agricultural practices depends on the timing of their execution. Environmental conditions, such as rainfall, and crop-related traits, such as plant phenology, determine the success of practices such as irrigation. Moreover, plant phenology, the seasonal timing of biological events (e.g., cotyledon emergence), is strongly influenced by genetic, environmental, and management conditions. Therefore, assessing the timing the of crops’ phenological events and their spatiotemporal variability can improve decision making, allowing the thorough planning and timely execution of agricultural operations. Conventional techniques for crop phenology monitoring, such as field observations, can be prone to error, labour-intensive, and inefficient, particularly for crops with rapid growth and not very defined phenophases, such as vegetable crops. Thus, developing an accurate phenology monitoring system for vegetable crops is an important step towards sustainable practices. This paper evaluates the ability of computer vision (CV) techniques coupled with deep learning (DL) (CV_DL) as tools for the dynamic phenological classification of multiple vegetable crops at the subfield level, i.e., within the plot. Three DL models from the Single Shot Multibox Detector (SSD) architecture (SSD Inception v2, SSD MobileNet v2, and SSD ResNet 50) and one from You Only Look Once (YOLO) architecture (YOLO v4) were benchmarked through a custom dataset containing images of eight vegetable crops between emergence and harvest. The proposed benchmark includes the individual pairing of each model with the images of each crop. On average, YOLO v4 performed better than the SSD models, reaching an F1-Score of 85.5%, a mean average precision of 79.9%, and a balanced accuracy of 87.0%. In addition, YOLO v4 was tested with all available data approaching a real mixed cropping system. Hence, the same model can classify multiple vegetable crops across the growing season, allowing the accurate mapping of phenological dynamics. This study is the first to evaluate the potential of CV_DL for vegetable crops’ phenological research, a pivotal step towards automating decision support systems for precision horticulture.