Image Transformation Research Articles

Optimizing Object Detection in Bio-Waste Management with Improved Hyperbolic Tangent YOLOv5 and Binarization-Adapted K-Means Algorithm

<p style="margin-bottom:11px;text-align:justify;"><span style="font-family:Calibri,sans-serif;font-size:11pt;line-height:107%;">With the increasing importance of green development, bio-waste classification has become an important element of green development. However, in garbage stacking scenarios, the task of segmenting highly overlapping objects is difficult because the bottom garbage is in an occluded state and it is usually difficult to distinguish its contours from occluded boundaries. This paper presents a rotationally invariant system for solid bio-waste container detection and bio-waste level classification. This paper proposes a rotationally invariant system for detecting solid bio-waste containers and classifying their fill levels. It employs Hough line detection to identify container orientations and positions, complemented by cross-correlation for distinguishing similar objects. Features extracted from within and outside the container area are utilized, enabling precise determination of discard levels. The system accommodates varying bio-waste amounts through transformation and rotation of container images. Object detection utilizes an Improved Hyperbolic Tangent-based YOLOv5 model with a Binarization-Adapted K-Means Algorithm (BKA) to enhance sorting accuracy. Additionally, the method efficiently calculates object motion through segmented BKA-detected objects. Experimental validation with a bio-waste image dataset demonstrates superior performance in accuracy, specificity, sensitivity, and overall precision compared to existing methodologies.</span></p>

Off-angle to frontal iris image transformation using Pix2Pix generative adversarial networks

Off-angle to frontal iris image transformation using Pix2Pix generative adversarial networks

Improved Faster R-CNN Based on Multi-Scale Module for Detecting Weak Ground Targets in Remote Sensing Images

In order to optimize the detection capability of small and weak ground targets in remote sensing images, a remote sensing image ground weak target detection method based on multi-scale module improved Faster R-CNN is proposed. By adjusting the Region Proposal Network (RPN) and Region of Interest (RoI) pooling layers, the model is better suited for feature extraction and classification of small targets. Generate feature maps at multiple scales by introducing additional convolutional layers to process inputs at different resolutions, with each layer focusing on targets within a specific size range. These multi-scale features are integrated and input into RPN to provide richer target proposal characteristics. By using image transformation techniques such as scaling, cropping, and rotation, the training dataset is expanded to simulate different small target scenes, enhancing the model’s adaptability to changes in target size, direction, and environment. The experimental results show that the proposed method not only outperforms other models in accuracy, recall, and F1 score, but also maintains high consistency and stability across different datasets. Its significant advantage in mean accuracy (MAP) further proves its reliability and effectiveness in remote sensing applications.

New Method for Tomato Disease Detection Based on Image Segmentation and Cycle-GAN Enhancement

A major concern in data-driven deep learning (DL) is how to maximize the capability of a model for limited datasets. The lack of high-performance datasets limits intelligent agriculture development. Recent studies have shown that image enhancement techniques can alleviate the limitations of datasets on model performance. Existing image enhancement algorithms mainly perform in the same category and generate highly correlated samples. Directly using authentic images to expand the dataset, the environmental noise in the image will seriously affect the model’s accuracy. Hence, this paper designs an automatic leaf segmentation algorithm (AISG) based on the EISeg segmentation method, separating the leaf information with disease spot characteristics from the background noise in the picture. This algorithm enhances the network model’s ability to extract disease features. In addition, the Cycle-GAN network is used for minor sample data enhancement to realize cross-category image transformation. Then, MobileNet was trained by transfer learning on an enhanced dataset. The experimental results reveal that the proposed method achieves a classification accuracy of 98.61% for the ten types of tomato diseases, surpassing the performance of other existing methods. Our method is beneficial in solving the problems of low accuracy and insufficient training data in tomato disease detection. This method can also provide a reference for the detection of other types of plant diseases.

Pre‐trained SAM as data augmentation for image segmentation

AbstractData augmentation plays an important role in training deep neural model by expanding the size and diversity of the dataset. Initially, data augmentation mainly involved some simple transformations of images. Later, in order to increase the diversity and complexity of data, more advanced methods appeared and evolved to sophisticated generative models. However, these methods required a mass of computation of training or searching. In this paper, a novel training‐free method that utilises the Pre‐Trained Segment Anything Model (SAM) model as a data augmentation tool (PTSAM‐DA) is proposed to generate the augmented annotations for images. Without the need for training, it obtains prompt boxes from the original annotations and then feeds the boxes to the pre‐trained SAM to generate diverse and improved annotations. In this way, annotations are augmented more ingenious than simple manipulations without incurring huge computation for training a data augmentation model. Multiple comparative experiments on three datasets are conducted, including an in‐house dataset, ADE20K and COCO2017. On this in‐house dataset, namely Agricultural Plot Segmentation Dataset, maximum improvements of 3.77% and 8.92% are gained in two mainstream metrics, mIoU and mAcc, respectively. Consequently, large vision models like SAM are proven to be promising not only in image segmentation but also in data augmentation.

Accuracy and efficiency of 2D/3D single-vertebra spine navigation registration method based on dual-view feature fusion

Objective: To investigate the accuracy and efficiency of spine 2D/3D preoperative CT and intraoperative X-ray registration through a framework for spine 2D/3D single-vertebra navigation registration based on the fusion of dual-position image features. Methods: The preoperative CT and intraoperative anteroposterior (AP) and lateral (LAT) X-ray images of 140 lumbar spine patients who visited Huashan Hospital Affiliated to Fudan University from January 2020 to December 2023 were selected. In order to achieve rapid and high-precision single vertebra registration in clinical orthopedic surgery, a designed transformation parameter feature extraction module combined with a lightweight module of channel and spatial attention (CBAM) was used to accurately extract the local single vertebra image transformation information. Subsequently, the fusion regression module was used to complement the features of the anterior posterior (AP) and lateral (LAT) images to improve the accuracy of the registration parameter regression. Two 1×1 convolutions were used to reduce the parameter calculation amount, improve computational efficiency, and accelerate intraoperative registration time. Finally, the regression module outputed the final transformation parameters. Comparative experiments were conducted using traditional iterative methods (Opt-MI, Opt-NCC, Opt-C2F) and existing deep learning methods convolutional neural network (CNN) as control group. The registration accuracy (mRPD), registration time, and registration success rate were compared among the iterative methods. Results: Through experiments on real CT data, the image-guided registration accuracy of the proposed method was verified. The method achieved a registration accuracy of (0.81±0.41) mm in the mRPD metric, a rotational angle error of 0.57°±0.24°, and a translation error of (0.41±0.21) mm. Through experimental comparisons on mainstream models, the selected DenseNet alignment accuracy was significantly better than ResNet as well as VGG (both P<0.05). Compared to existing deep learning methods [mRPD: (2.97±0.99) mm, rotational angle error: 2.64°±0.54°, translation error: (2.15±0.41) mm, registration time: (0.03±0.05) seconds], the proposed method significantly improved registration accuracy (all P<0.05). The registration success rate reached 97%, with an average single registration time of only (0.04±0.02) seconds. Compared to traditional iterative methods [mRPD: (0.78±0.26) mm, rotational angle error: 0.84°±0.57°, translation error: (1.05±0.28) mm, registration time: (35.5±10.5) seconds], registration efficiency of the proposed method was significantly improved (all P<0.05). The dual-position study also compensated for the limitations in the single-view perspective, and significantly outperforms both the front and side single-view perspectives in terms of positional transformation parameter errors (both P<0.05). Conclusion: Compared to existing methods, the proposed CT and X-ray registration method significantly reduces registration time while maintaining high registration accuracy, achieving efficient and precise single vertebra registration.

Live feature transformation scale extraction of face images based on deep learning algorithm

In view of the low extraction accuracy and efficiency of the current feature extraction methods of face images, a method for extracting the scale of live feature transformation of face images based on deep learning algorithm is proposed. First, the deep learning method is used to complete the denoising of face images; then the Gabor wave filter is used to decompose the face signal, and it is input into the deep subspace model to complete the extraction of feature transformation scale; finally, the live feature transformation scale extraction of face images is completed based on (particle PSO swarm optimization algorithm). Experimental results show that the proposed method for extracting face image features has higher accuracy, faster recognition speed and better overall application effect.

Hyperspectral image classification with token fusion on GPU

Hyperspectral images capture material nuances with spectral data, vital for remote sensing. Transformer has become a mainstream approach for tackling the challenges posed by high-dimensional hyperspectral data with complex structures. However, a major challenge they face when processing hyperspectral images is the presence of a large number of redundant tokens, which leads to a significant increase in computational load, adding to the model’s computational burden and affecting inference speed. Therefore, we propose a token fusion algorithm tailored to the operational characteristics of the hyperspectral image and pure transformer network, aimed at enhancing the final accuracy and throughput of the model. The token fusion algorithm introduces a token merging step between the attention mechanism and the multi-layer perceptron module in each Transformer layer. Experiments on four hyperspectral image datasets demonstrate that our token fusion algorithm can significantly improve inference speed without any training, while only causing a slight decrease in the pure transformer network’s classification accuracy.

Fantasy composition as a means of forming the artistic culture of architecture students

The article deals with the problem of the formation of the artistic culture of architectural students in the process of performing a fantasy composition. The authors consider the possibilities of developing these qualities when mastering the main professional disciplines of Drawing and Painting and coloristics in accordance with the standard of higher education. During the analysis of the literature, the authors consider the concepts of fantasy and composition, clarify the concept of fantasy composition, identify effective ways to form the artistic culture of architecture students using effective methods such as associative exercises, exercises for the transformation of forms and images, stylization and creation of a decorative image; creative tasks and projects to create fantasy compositions on various topics.

Transformation but not generation of motor images is disrupted following stimulation over the left inferior parietal lobe

Motor imagery (MI) involves the generation, maintenance, and transformation of motor images; yet, the neural underpinnings of each stage are not well understood. Here, we investigated the role of the left inferior parietal lobe (IPL) in the stages of MI. Healthy participants (N=20) engaged in a MI task (making judgments about hands presented on a screen; hand laterality judgment task) over two days. Past literature demonstrates the mental rotation of hands in this task involves implicit MI (i.e., where MI occurs spontaneously in the absence of explicit instructions). During the task, active (Day A; 120% resting motor threshold) or sham (Day B; placebo) neuronavigated transcranial magnetic stimulation (TMS) was applied to the left IPL (location determined from past neuroimaging work) on 50% of trials at 250, 500, or 750ms post-stimulus onset, corresponding to different stages of MI. A/B days were randomized across participants. Linear mixed effects (LME) modelling conducted on reaction time and accuracy revealed that longer reaction times were observed when TMS was delivered at 750ms after trial onset, and more greatly for active vs. sham stimulation. This effect was exacerbated for palm- vs. back-view stimuli and for left vs. right hands. Accuracy overall was decreased for active vs. sham stimulation, and to a greater extent for palm- vs. back-view stimuli. Findings suggest that the left IPL is involved in image transformation. Overall this work informs on the neural underpinnings of the stages of MI.

Batik Pattern Classification Using Machine Learning Approaches

Batik is one of Malaysia traditional textile art form that is well known. By using Artificial Neural Network (ANN), k-Nearest Neighbors (k-NN) and Decision Tree methods, this study aims to classify Kelantan batik designs according to flora, fauna and geometry motifs. 133 images of the three categories were collected from social media and underwent pre-processing techniques. Image augmentation was done to enhance the diversity and quality of available training data for machine learning models. Digital transformation on the images based on colour features was developed to classify the three types of batik motifs. Image embedding was employed which returned a vector representation of each image in a data table in Orange software. The data divided into training and testing dataset, with a ratio of 8:2. The performance of each machine learning techniques were compared. The findings showed that ANN produced an excellent performance as indicated by classification accuracy (CA) of 87.9% for original images and 82.2% using Luminance Gradient image transformation technique. ANN also returns the highest AUC score indicating it is the best in distinguishing the images motif. Results from confusion matrix showed that ANN have the least misclassification for batik classification. Based on the result, it signifies that ANN performs the best classification among the three approaches.

Deep Learning Classification of Ischemic Stroke Territory on Diffusion-Weighted MRI: Added Value of Augmenting the Input with Image Transformations.

Our primary aim with this study was to build a patient-level classifier for stroke territory in DWI using AI to facilitate fast triage of stroke to a dedicated stroke center. A retrospective collection of DWI images of 271 and 122 consecutive acute ischemic stroke patients from two centers was carried out. Pretrained MobileNetV2 and EfficientNetB0 architectures were used to classify territorial subtypes as middle cerebral artery, posterior circulation, or watershed infarcts along with normal slices. Various input combinations using edge maps, thresholding, and hard attention versions were explored. The effect of augmenting the three-channel inputs of pre-trained models on classification performance was analyzed. ROC analyses and confusion matrix-derived performance metrics of the models were reported. Of the 271 patients included in this study, 151 (55.7%) were male and 120 (44.3%) were female. One hundred twenty-nine patients had MCA (47.6%), 65 patients had posterior circulation (24%), and 77 patients had watershed (28.0%) infarcts for center 1. Of the 122 patients from center 2, 78 (64%) were male and 44 (34%) were female. Fifty-two patients (43%) had MCA, 51 patients had posterior circulation (42%), and 19 (15%) patients had watershed infarcts. The Mobile-Crop model had the best performance with 0.95 accuracy and a 0.91 mean f1 score for slice-wise classification and 0.88 accuracy on external test sets, along with a 0.92 mean AUC. In conclusion, modified pre-trained models may be augmented with the transformation of images to provide a more accurate classification of affected territory by stroke in DWI.

Texture Analysis and Classification using Local Binary Patterns and Statistical Features

Texture analysis and classification are crucial in many applications such as medical applications, satellite imagery analysis, and so on. This paper presents a new technique that incorporates the LBP with the HOG to improve the texture analysis and classification. LBP is famous for finding out the local texture descriptors by comparing the pixel intensities it deals with and can challenge light variations and image transformations. On the other hand, there is HOG which emphasizes edge and gradient information to detect the orientation and shape of a texture in a given image. While LBP describes the local micro-patterns HOG gives the perspective of the global gradient orientation. The combined use of local and global texture information is helpful in the presented problem when prioritizing the recognition of complex texture patterns that cannot be deciphered by a single algorithm. To demonstrate the efficiency of the proposed hybrid method, several experiments are performed on several texture databases. According to the findings, the proposed LBP-HOG strategy proves more beneficial than basic LBP and HOG techniques concerning the classification rate and computing time. The combination method shows improved performance in terms of agreement with the reference of fine and coarse level of texture descriptions along with providing large-scale description of the texture which results in the improved discriminatory power of the texture classes.

Multi-Temporal Snow-Covered Remote Sensing Image Matching via Image Transformation and Multi-Level Feature Extraction

To address the challenge of image matching posed by significant modal differences in remote sensing images influenced by snow cover, this paper proposes an innovative image transformation-based matching method. Initially, the Pix2Pix-GAN conversion network is employed to transform remote sensing images with snow cover into images without snow cover, reducing the feature disparity between the images. This conversion facilitates the extraction of more discernible features for matching by transforming the problem from snow-covered to snow-free images. Subsequently, a multi-level feature extraction network is utilized to extract multi-level feature descriptors from the transformed images. Keypoints are derived from these descriptors, enabling effective feature matching. Finally, the matching results are mapped back onto the original snow-covered remote sensing images. The proposed method was compared to well-established techniques such as SIFT, RIFT2, R2D2, and ReDFeat and demonstrated outstanding performance. In terms of NCM, MP, Rep, Recall, and F1-measure, our method outperformed the state of the art by 177, 0.29, 0.22, 0.21, and 0.25, respectively. In addition, the algorithm shows robustness over a range of image rotation angles from −40° to 40°. This innovative approach offers a new perspective on the task of matching multi-temporal snow-covered remote sensing images.

Bilateral fusion low‐light image enhancement with implicit information constraints

AbstractResearch on low‐light image enhancement focuses on improving image quality in dim conditions. Recently, deep learning has driven significant advancements, with many studies using neural networks to enhance low‐light images. However, most focus on complex network designs to increase nonlinearity, often neglecting the implicit information from local image transformations. This paper introduces an improved U‐net‐based method for low‐light enhancement, retaining the original encoding network and adding branch links in the decoding network. The method uses attention feature fusion to handle image noise and gradients separately, adjusting brightness through a gradient‐adaptive transform. This approach optimizes performance using loss functions like peak signal‐to‐noise ratio and colour consistency. Unlike previous methods, the approach emphasizes extracting implicit information from image gradients, achieving enhancement that aligns with the original brightness distribution. The result is enhanced images with high detail similarity to the original, achieved through end‐to‐end inference in experiments.

Multi-spectral image transformer descriptor classification combined with molecular tools for early detection of tomato grey mould

Accurate early detection of plant diseases is a critical milestone in the development of self-navigating robots for precision agriculture and still remains one of the most significant challenges. Plant pathogens, like Botrytis cinerea causing grey mould disease, pose significant threats to agriculture and food safety. This study focuses on the early detection of B. cinerea in tomato crops using Artificial Intelligence. Specifically, the Deep Learning (DL) YOLOv8 architecture was employed to apply single class segmentation on plant captures to extract the tomato leaves locational information. Each plant capture includes five images from hyperspectral wavelengths (at 460, 540, 640, 775 and 875 nm) and one RGB images. The leaf segmentation achieved 81.7 % Mean Average Precision (mAP) at Intersection over Union (IoU) threshold 0.5 (mAP50). Then the for each leaf segment, and array of descriptors is extracted through various Transformer models. Finally, the descriptors are classified through a KNN or an LSTM solution and the results of all the descriptors for each leaf from each of the Transformer models, for each of the six images of the capture, are ensembled to yield the class of each leaf. The best ensemble results were produced by only accumulating results from some of the Transformer and specifically the MaxViT-L, the ViT-B (P:16 × 16 – C), the VOLO (D:5) and the XCIT-L (L:24 – P:16 × 16), by only using the results provided by the images at the hyperspectral wavelengths at 540, 640 nm and the RGB image, and by applying the LSTM solution. The process achieved a 79.41 % classification accuracy and 71.12 % F1-Score for all classes. Additionally, early, rapid and accurate detection of grey mould at the early stages of the infection or at latent infections when symptoms are not visible was assessed by comparative qPCR-based methods (qRT-PCR) for fungal biomass estimation in plant tissues and the rest of this study's methodology. The findings of this study represent a significant advancement in crop disease detection and management, and highlight the potential for integrating these methods into on-site digital systems (robots, mobile apps, etc.) under real-world settings. The results, demonstrate the effectiveness of combining segmentation and transformer-based classification models for early and accurate detection of grey mould disease.

On image transformation for partial discharge source identification in vehicle cable terminals of high‐speed trains

AbstractPartial discharge (PD) detection of cable terminals is crucial for the safe operation of the traction power system in trains. However, similar PD signals in complex train‐operating environments cause difficulty to recognise the insulation defects. Therefore, a PD signal image transformation recognition method is proposed for PD detection of cable terminal defects to identify defects in cable terminals with similar PD characteristics accurately. In the proposed method, the raw PD signals are firstly transformed to images via the Gramian angular field (GAF) representation. This can reveal the discriminative characteristics embedded in the original PD signals and subsequently facilitate differentiating the PD sources, which exhibit similar characteristic in the time domain. The obtained GAF representation of PD signals (named as PD GAF images) is extracted from local and global features to train an efficient MobileVIT model, which is then utilised to identify similar types of PD sources in cable terminals. The results show that the proposed method achieves 97.5% recognition accuracy in the field experiment, which is superior to other methods.

Lightweight vision image transformer (LViT) model for skin cancer disease classification

Lightweight vision image transformer (LViT) model for skin cancer disease classification

Learning to reconstruct accelerated MRI through K-space cold diffusion without noise

Deep learning-based MRI reconstruction models have achieved superior performance these days. Most recently, diffusion models have shown remarkable performance in image generation, in-painting, super-resolution, image editing and more. As a generalized diffusion model, cold diffusion further broadens the scope and considers models built around arbitrary image transformations such as blurring, down-sampling, etc. In this paper, we propose a k-space cold diffusion model that performs image degradation and restoration in k-space without the need for Gaussian noise. We provide comparisons with multiple deep learning-based MRI reconstruction models and perform tests on a well-known large open-source MRI dataset. Our results show that this novel way of performing degradation can generate high-quality reconstruction images for accelerated MRI.

Identification of the damage location for the structural sealant based on deep learning

Structural sealants are essential to maintain the safety of panel units of hidden frame glass curtain wall. Damage localization of the concealed structural sealant is still difficult because of the unknown baseline model. In this paper, a convolutional neural network-based identification method is proposed to localize the damage of structural sealants without the baseline model. The method develops a novel input of the convolutional neural network (CNN), multi-symmetry-point images (MSPI), which is encoded by the response of four symmetrical points to impulse excitations. The CNN would identify the damage location by discerning differences among four images. Then, a dataset with 28803 samples, which considers the effect of the multiple damages and noise, was used to train the CNN. Three types of transformed images and four CNN models were compared to optimize the input signals and the configuration of the CNN. A series of numerical examples indicated that the Gram angle difference field is the optimal image transformation method, and improved-DenseNet121 is the optimal CNN for damage localization in structural sealants. Then, several laboratory experiments validate the effectiveness of the optimized image input and CNN with an accuracy rate of 92 % in the identification of damage location.