Complex Background Images Research Articles

Threshold Segmentation and Length Measurement Algorithms for Irregular Curves in Complex Backgrounds.

It is an urgent problem to know how to quickly and accurately measure the length of irregular curves in complex background images. To solve the problem, we first proposed a quasi-bimodal threshold segmentation (QBTS) algorithm, which transforms the multimodal histogram into a quasi-bimodal histogram to achieve a faster and more accurate segmentation of the target curve. Then, we proposed a single-pixel skeleton length measurement (SPSLM) algorithm based on the 8-neighborhood model, which used the 8-neighborhood feature to measure the length for the first time, and achieved a more accurate measurement of the curve length. Finally, the two algorithms were tested and analyzed in terms of accuracy and speed on the two original datasets of this paper. The experimental results show that the algorithms proposed in this paper can quickly and accurately segment the target curve from the neon design rendering with complex background interference and measure its length.

TRAFFIC SIGN DESCRIPTION SYSTEM

Traffic Sign Recognition is used to give an audio description to a image sign traffic signs and thus warn a driver, and command or prohibit certain actions. A fast real-time and robust automatic traffic sign detection and recognition can support and disburden the driver and significantly increase driving safety and comfort. Automatic recognition of traffic signs is also important for automated intelligent driving vehicle or driver assistance systems. This paper presents a study recognize traffic sign patterns using Neural Networks technique. The images are pre-processed with several image processing techniques, such as, threshold techniques, Gaussian filter, Canny edge detection, Contour and Fit Ellipse. Then, the Neural Networks stages are performed to recognize the traffic sign patterns. The system is trained and validated to find the best network architecture. The experimental results show the highly accurate classifications of traffic sign patterns with complex background images and the computational cost of the proposed method.mental results

Feature fusion capsule network for cow face recognition

There are many difficulties in cow face recognition, including variable individual poses, challenges in collecting individual data, complex image backgrounds, etc. An improved capsule network (CapsNet) was used to solve these problems. First, we combined convolutional and local binary pattern (LBP) texture features with a feature extractor named C-LBP. Then, by utilizing the self-attention module, the feature extraction capability was enhanced. An intermediate capsule layer was added to improve capsules utilization. We tested our model on datasets of cow faces. According to the experimental data, the proposed model improved the structure of the original CapsNet. In the middle of the training, positional data was added, giving the model a higher performance and greater resilience in cow face recognition.

Optical Remote Sensing Image Object Detection Model Based On Improved YOLOv5s

Aiming at the problem of error and omission caused by the complex image background and the high similarity between the remote sensing objectt and the background in remote sensing image detection, an improved model of remote sensing image detection based on YOLOv5s is proposed. The dual attention mechanism of channel and space is added to the convolutional layer of the trunk feature extraction network and the feature fusion network, which enhances the features with high correlation of the objectt to be detected and suppresses the features with low correlation, and improves the ability of the model to extract the features of remote sensing objectts in a complex background, so as to improve the detection accuracy of the algorithm. In this study, a comparative experiment was conducted on the NWPU VHR-10 dataset, and the results showed that the average accuracy of the proposed model was 94% when the dataset was turned over and the ratio was 0.5, which was 2.3% higher than that of the original YOLov5s, which verified the effectiveness of the proposed method.

Research on Pavement Crack Detection Algorithm based on Deep Residual Unet Neural Network

Pavement crack is one of the main factors threatening highway safety. Rapid and accurate identification of pavement crack is very important to protect highway. Traditional image processing methods have been difficult to meet the requirements of practical engineering. So, this paper proposes an improved pavement crack detection method based on deep convolution neural network model, to overcome the inaccurate segmentation caused by complex background and fuzzy crack edge in asphalt pavement image, which is an encoder-decoder network structure, and integrates the idea of semantic segmentation into image detection. The encoder of the proposed method adopts the ResNet’s residual structure to strengthen image feature extraction, the decoder uses transpose convolution from the original UNet structure for up sampling, and integrates the encoder features with the up sampled features after cutting, Strengthen the learning of network features. Experimental results show that the F1 score, Precision and Recall are significantly improved on the asphalt crack dataset. Compared with the traditional algorithm, the proposed algorithm can segment the cracks better and improve the accuracy of pavement crack detection.

Research and Application of Fine-Grained Image Classification Based on Small Collar Dataset.

With the rapid development of apparel e-commerce, the variety of apparel is increasing, and it becomes more and more important to classify the apparel according to its collar design. Traditional image processing methods have been difficult to cope with the increasingly complex image backgrounds. To solve this problem, an EMRes-50 classification algorithm is proposed to solve the problem of garment collar image classification, which is designed based on the ECA-ResNet50 model combined with the MC-Loss loss function method. Applying the improved algorithm to the Coller-6 dataset, and the classification accuracy obtained was 73.6%. To further verify the effectiveness of the algorithm, it was applied to the DeepFashion-6 dataset, and the classification accuracy obtained was 86.09%. The experimental results show that the improved model has higher accuracy than the existing CNN model, and the model has better feature extraction ability, which is helpful to solve the problem of the difficulty of fine-grained collar classification and promote the further development of clothing product image classification.

ChickTrack – A quantitative tracking tool for measuring chicken activity

The automatic detection, counting and tracking of individual and flocked chickens in the poultry industry is of paramount to enhance farming productivity and animal welfare. Due to methodological difficulties, such as the complex background of images, varying lighting conditions, and occlusions from e.g., feeding stations, water nipple stations and barriers in the chicken rearing production floor, it is a challenging task to automatically recognize and track birds using computer software. Here, a deep learning model based on You Only Look Once (Yolov5) is proposed for detecting domesticated chickens from videos with varying complex backgrounds. A multiscale feature is being adapted to the Yolov5 network for mapping modules in the counting and tracking of the trajectories of the chickens. The Yolov5 network was trained and tested on our dataset which resulted in an enhanced tracking precision accuracy. Using Kalman Filter, the proposed model was able to track multiple chickens simultaneously with the focus to associate individual chickens across the frames of the video for real time and online applications. By being able to detect the chickens amid diverse background interference and counting them precisely along with tracking the movement and measuring their travelled path and direction, the proposed model provides excellent performance for on-farm applications. Artificial intelligence enabled automatic measurements of chicken behavior on-farm using cameras offers continuous monitoring of the chicken's ability to perch, walk, interact with other birds and the farm environment, as well as the assessment of dustbathing, thigmotaxis, and foraging frequency, which are important indicators for their ability to express natural behaviors. This study highlights the potential of automated monitoring of poultry through the usage of ChickTrack model as a digital tool in enabling science-based animal husbandry practices and thereby promote positive welfare for chickens in animal farming.

Vehicle Detection Based on Semantic-Context Enhancement for High-Resolution SAR Images in Complex Background

Small-scale target detection (such as vehicles) in complex synthetic aperture radar (SAR) image scenes has always been a pain point for the advanced convolutional neural network (CNN)-based target detectors because of the downsampling operations and the local receptive field characteristics of CNNs. To tackle these limitations, a vehicle detector named SCEDet for the small-scale vehicles in SAR images is proposed to improve the detection performance in this letter. SCEDet mainly consists of two parts: subaperture semantic feature extraction and subaperture semantic-context enhancement (SCE) with SCE module. First, ResNet34 with subaperture decomposition is used to efficiently exploit the latent subaperture semantic features. Then, the SCE module is proposed to balance the multiscale semantic information as well as aggregate the global context information for vehicle detection with a small number of parameters and computation costs. The experimental results on the FARAD dataset (0.1 m <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\times0.1$ </tex-math></inline-formula> m, Ka-band) demonstrate that both the detection performance and the speed are much better than other detection methods under the same hardware conditions.

CGC-NET: Aircraft Detection in Remote Sensing Images Based on Lightweight Convolutional Neural Network

In the past few years, aircraft detection in remote sensing (RS) images has been an important research hotspot, and it is very crucial in plenty of military applications. Based on the high computational cost of the model and numerous parameters, deep convolution neural networks-based algorithms have excellent performance in the aircraft detection task. However, it is still difficult to detect aircraft due to the complex background of RS images, various types of aircraft, and so on. In addition, it is difficult and costly to make labels for satellite-based optical RS images. Consequently, we propose an end-to-end lightweight aircraft detection framework called CGC-NET (a network based on circle grayscale characteristics), which can accurately detect aircraft with a few training samples. There are only a small number of trainable parameters in CGC-NET, which greatly reduces the need for large datasets. Extensive evaluations indicate the excellent performance of CGC-NET, in which the <i>F</i>-score can reach 91.06&#x0025; and the model size is only 0.88 M. Therefore, CGC-NET can be used to accurately detect aircraft targets simply and effectively.

A concatenation of deep and texture features for medicinal trash image classification using EnSegNet-DNN-based transfer learning

The foremost vital process in clinical trash classification is to classify the medicinal wastes into various categories like contagious, poisonous and normal wastes. For this purpose, several deep learning systems using different structures including ResNext, GoogleNet, etc., are designed. Among those systems, an Enhanced Segmentation Network (EnSegNet) with DNN-TC (EnSegNet-DNN-TC) system has achieved a higher efficiency by segmenting and classifying the trash input images. Although it segments and extracts features effectively, there are very subtle differences between many images because of their highly complex background. This leads to misjudgments of the deep learner system. Therefore in this article, an EnSegNet with Combined Feature Extraction (CFE) and DNN-TC called EnSegNet-CFE-DNN-TC system is proposed to solve misjudgments problem in deep learner classifiers due to high complex background images. First, the input images are segmented by the EnSegNet model. After that, various texture characteristics, namely Gray-Level Co-occurrence Matrix (GLCM), Multi-level Local Binary Pattern (MLBP), Local Derivative Pattern (LDP) and Local Ternary Pattern (LTP) are extracted including the deep features from the segmented images. Then, a new layer called the combination layer is introduced after the Fully Connected (FC) layers to fuse the extracted features and construct a new hybrid feature vector. This hybrid feature vector has a stronger discriminant ability compared to the single feature vector. Further, the softmax is performed to classify the medicinal wastes. Finally, the investigation outcomes reveal that the EnSegNet-CFE-DNN-TC system attains a 93.7% of classification accuracy for 100 trash images compared to the EnSegNet-DNN-TC and DNN-TC.

Fine-Grained Ship Recognition for Complex Background Based on Global to Local and Progressive Learning

The existing deep learning fine-grained recognition models have low recognition accuracy (Acc) in complex natural environments. We propose a fine-grained ship recognition model [global to local progressive learning module (GLPM)] based on global to local and progressive learning for complex backgrounds to address this problem. GLPM improves fine-grained ship targets’ recognition Acc in natural complex backgrounds by progressively guiding high-level global features to low-level local features for learning and enhancing local fine-grained features expression ability. First, the global features are obtained by adding the feature pyramid attention (FPA) module to the deepest level of the backbone network. Second, the decoding and reconstruction of low-level local feature information at different levels are guided, and the global features are used to weight the key pixel values of local features of the target. Finally, the output of the backbone network is combined with the output vector of the global to local progressive module for classification, and the whole model can be trained end-to-end. We experiment on naturally collected complex background images containing 59 ship classes, and the public maritime (MAR)-ships contain pictures of 23 types of ships. The results show that the proposed method outperforms the methods compared in this letter while providing valuable ideas for the research of fine-grained ship recognition models in complex natural environments.

Complex image background segmentation for cable force estimation of urban bridges with drone‐captured video and deep learning

Drone-assisted structural health monitoring has aroused extensive attention due to its high mobility and low cost. However, drone motion and complex image backgrounds severely impede its application in the cable force measurement of urban bridges. To fill this research gap, this paper proposed a deep learning-based complex background segmentation approach for cable force estimation of urban bridges from the drone-captured video. The main contribution of this article includes two aspects: (1) A pre-trained fully convolutional network (FCN) model was first adopted to identify bridge cables from drone-captured video and further to extract sub-pixel-level displacement using line segment detection (LSD) algorithm, and (2) an empirical mode decomposition algorithm was employed for extracting the vibration signal of bridge cables by eliminating the effect of drone motion on measured dynamic displacement. Finally, natural frequencies of bridge cables were obtained by performing Fourier analysis on extracted cable vibration and further adopted for cable force estimation. The effectiveness and robustness of the proposed method have been successfully verified by field testing of an urban cable-stayed footbridge. Estimated cable forces using the proposed method are consistent with the traditional contact-type measurements and design values, demonstrating the potential of this method for applying into rapid cable force estimation of numerous urban bridges.

Research on the Region-Growing and Segmentation Technology of Micro-Particle Microscopic Images Based on Color Features

Silkworm microparticle disease is a legal quarantine standard in the detection of silkworm disease all over the world. The current common detection method, the Pasteur manual microscopy method, has a low detection efficiency all over the world. The low efficiency of the current Pasteur manual microscopy detection method makes the application of machine vision technology to detect microparticle spores an important technology to advance silkworm disease research. For the problems of the low contrast, different illumination conditions and complex image background of microscopic images of the ellipsoidal symmetrical shape of silkworm microparticle spores collected in the detection solution, a region growth segmentation method based on microparticle color and grayscale information is proposed. In this method, the fuzzy contrast enhancement algorithm is used to enhance the color information of micro-particles and improve the discrimination between the micro-particles and background. In the HSV color space with stable color, the color information of micro-particles is extracted as seed points to eliminate the influence of light and reduce the interference of impurities to locate the distribution area of micro-particles accurately. Combined with the neighborhood gamma transformation, the highlight feature of the micro-particle target in the grayscale image is enhanced for region growing. Mea6nwhile, the accurate and complete micro-particle target is segmented from the complex background, which reduces the background impurity segmentation caused by a single feature in the complex background. In order to evaluate the segmentation performance, we calculate the IOU of the microparticle sample image segmented by this method with its corresponding true value image, and the experiments show that the combination of color and grayscale features using the region growth technique can accurately and completely segment the microparticle target in complex backgrounds with a segmentation accuracy IOU as high as 83.1%.

Retraction Note: Mountain bridge design based on remote sensing images

The bridge identification method in the high-resolution remote sensing image is realized by using the characteristics of the bridge. Now, due to the continuous leap forward in sensing technology, the data volume of remote sensing images has increased substantially, and the resolution of remote sensing images has also been continuously improved. People are paying more and more attention to the automatic recognition and positioning of bridges in remote sensing images. Automatic bridge recognition technology can automatically extract bridge signs from the complex background of earth observation images through high-speed real-time information processing based on specific regions and various input function template databases and detect, intercept, identify, and track targets. By using the characteristics of typical targets or image edges, grayscale, texture structure meaning, and regional shape, satellite remote sensing images have the advantage of being able to provide surface information at a high speed. However, in special cases such as low resolution (medium and low), the acquisition and measurement of GIS information cannot play a good role. In order to provide a reliable guarantee for mapping and map updates, it is necessary to develop high-resolution satellite remote sensing images. Use remote sensing images with high spatial resolution as the data source, and use a rule-based object-oriented method to correctly extract bridge targets from high-resolution images. First, through multi-scale scoring experiments, combined with the characteristics of the bridge, select the best segmentation ratio. Second, the rule set is established using the water index and threshold function method, and the vector files of the water area and the bridge are gradually obtained. Finally, the bridge target can be obtained normally through binarization, mathematical shape processing, overlay analysis, and other methods.

Improved U-Net based insulator image segmentation method based on attention mechanism

To realize the accurate identification and segmentation of the insulator string in the complex background image with diverse appearance and obscuration, this paper proposes an insulator segmentation method based on improved U-Net. The algorithm embeds the attention mechanism ECA-Net (Efficient Channel Attention Neural Networks) in the coding stage of U-Net to improve the model’s ability to extract semantic features, thereby improving the accuracy of insulator detection. Experimental results show that the average overlap IOU of the proposed method is 96.8%, which can more accurately segment different types of insulators in complex backgrounds.

Insulator defect detection: A detection method of target search and cascade recognition

Aiming at the difficulties of few real samples for defective insulators and complex background of aerial images, this paper proposes a detection method based on target search and cascade recognition. Using the SINet framework, we apply fine-grained texture enhancement to different sizes of receptive fields. Through nearest-neighbor decoding and grouping reverse attention, the more recognizable features are guided to aggregate and generate a refined location area map by performing cascading purification operations. Additionally, we integrate the classification network to complete the solution. Experimental results show that the AUC value is up to 99.82%, which demonstrates the effectiveness and superiority of the proposed method on insulator defect detection.

Detection of Abnormal Vibration Dampers on Transmission Lines in UAV Remote Sensing Images with PMA-YOLO

The accurate detection and timely replacement of abnormal vibration dampers on transmission lines are critical for the safe and stable operation of power systems. Recently, unmanned aerial vehicles (UAVs) have become widely used to inspect transmission lines. In this paper, we constructed a data set of abnormal vibration dampers (DAVDs) on transmission lines in images obtained by UAVs. There are four types of vibration dampers in this data set, and each vibration damper may be rusty, defective, or normal. The challenges in the detection of abnormal vibration dampers on transmission lines in the images captured by UAVs were as following: the images had a high resolution as well as the objects of vibration dampers were relatively small and sparsely distributed, and the backgrounds of cross stage partial networks of the images were complex due to the fact that the transmission lines were erected in a variety of outdoor environments. Existing methods of ground-based object detection significantly reduced the accuracy when dealing with complex backgrounds and small objects of abnormal vibration dampers detection. To address these issues, we proposed an end-to-end parallel mixed attention You Only Look Once (PMA-YOLO) network to improve the detection performance for abnormal vibration dampers. The parallel mixed attention (PMA) module was introduced and integrated into the YOLOv4 network. This module combines a channel attention block and a spatial attention block, and the convolution results of the input feature maps in parallel, allowing the network to pay more attention to critical regions of abnormal vibration dampers in complex background images. Meanwhile, in view of the problem that abnormal vibration dampers are prone to missing detections, we analyzed the scale and ratio of the ground truth boxes and used the K-means algorithm to re-cluster new anchors for abnormal vibration dampers in images. In addition, we introduced a multi-stage transfer learning strategy to improve the efficiency of the original training method and prevent overfitting by the network. The experimental results showed that the mAP@0.5 for PMA-YOLO in the detection of abnormal vibration dampers reached 93.8% on the test set of DAVD, 3.5% higher than that of YOLOv4. When the multi-stage transfer learning strategy was used, the mAP@0.5 was improved by a further 0.2%.

A robust end-to-end deep learning framework for detecting Martian landforms with arbitrary orientations

With increasingly massive amounts of high-resolution images of Mars, automated detection of geological landforms on Mars has received widespread interest. It is significant for acquiring knowledge of distant planetary surfaces and processes, or manifold onboard applications such as spacecraft motion estimation and obstacle avoidance. This is a challenging task, not only because of the multiple sizes of targets and complex image backgrounds, but also the various orientations of some bar-shaped landforms in satellite images captured from the top view. The existing methods for directed landform detection require several pre or post-processing operations to extract possible regions of interest and final detection results with orientation, which are very time consuming. In this paper, a new end-to-end deep learning framework is developed for detecting arbitrarily-directed landforms. This framework, named Rotated-SSD (Single Shot MultiBox Detector, SSD), can locate and identify different landforms on Mars in one pass, by using rotatable anchor-box based mechanism. To enhance its robustness against angle variation of the targets and complex backgrounds, a new efficient match strategy is proposed for anchoring default boxes to ground truth boxes in the model training process and an autoencoder-based unsupervised pre-training operation is introduced to improve both the model training and inference performance. The proposed framework is tested for detection of bar-shaped buttes and dark slope streaks on satellite images. The detection results show that our framework can significantly contribute to onboard motion estimation systems. The comparative results demonstrate that the proposed match strategy outperforms other state-of-the-art match strategies with regard to model training efficiency and prediction accuracy. The pre-training strategy can facilitate the training of deep architectures in case of limited available training data.

High-resolution representations and multistage region-based network for ship detection and segmentation from optical remote sensing images

Ship detection and segmentation in optical remote sensing (ORS) images is a current concern actively addressed by the academic community owing to its vast applications. Most present methods only detect the location of the ship but do not segment it at the pixel level. Considering the complex background of ORS images, identifying ships from interferences, such as clouds, waves, and some land architectures similar to ships, proves to be difficult. To address this issue, we propose a high-resolution representation and multistage region-based network (HR-MSRN) for ship detection and segmentation from ORS images. HR-MSRN mainly consists of three parts: the high-resolution feature pyramid network (HRFPN), region proposal network (RPN), and multistage detection and segmentation network (MSDSN). First, HRFPN is built as a backbone network to extract and fuse multilevel image feature maps. Second, ship candidate boxes are generated by defining numerous anchors through RPN. Third, using the idea of a cascade mask R-CNN as the reference method, the MSDSN is proposed to obtain the ship localization and mask shape. We utilize the proposed framework to evaluate an Airbus-ship dataset, and the experiments indicate that (1) HRFPN provides better feature representation ability than the ResNet-FPN when maintaining the same detection framework, especially for small ships; (2) the direct flow between mask branches refines the mask information, and the semantic segmentation branch enhances context information, which indicates that MSDSN is effective and promotes further improvements in ship detection and segmentation from ORS images; (3) in comparison to other region-based methods, HR-MSRN obtains superior performance of ship detection and segmentation in the ORS imagery.

MEAN-SSD: A novel real-time detector for apple leaf diseases using improved light-weight convolutional neural networks

Alternaria blotch, Brown spot, Mosaic, Grey spot, and Rust are 5 common apple leaf diseases that severely impact apple production and quality. At present, although many CNN methods have been proposed for apple leaf diseases, there are still lack of apple leaf disease detection models that can be applied on mobile devices, which limits their application in practical production. This paper proposes a light-weight CNN model that can be deployed on mobile devices to detect apple leaf diseases in real time. First, a dataset of apple leaf diseases composed of simple background images and complex background images, which is called AppleDisease5, is constructed via data augmentation technology and data annotation technology. Then a basic module called MEAN block(Mobile End AppleNet block) is proposed to increase the detection speed and reduce model’s size by reconstructing the common 3×3 convolution. Meanwhile, the Apple-Inception module is built by introducing GoogLeNet’s Inception module and replacing all 3×3 convolution kernels with MEAN block in Inception. Finally, a new apple leaf disease detection model, MEAN-SSD(Mobile End AppleNet based SSD algorithm), is built by using the MEAN block and Apple-Inception module. The experiment results show that MEAN-SSD can achieve the detection performance of 83.12% mAP and a speed of 12.53 FPS, which illustrates that the novel MEAN-SSD model can efficiently and accurately detect 5 common apple leaf diseases on mobile devices.