Image Descriptors Research Articles

SAR-MINF: A Novel SAR Image Descriptor and Matching Method for Large-Scale Multidegree Overlapping Tie Point Automatic Extraction

The automatic extraction of large-scale tie points (TPs) for Synthetic Aperture Radar (SAR) images is crucial for generating SAR Digital Orthophoto Maps (DOMs). This task involves matching SAR images under various conditions, such as different resolutions, incidence angles, and orbital directions, which is highly challenging. To address the feature extraction challenges of different SAR images, we propose a Gamma Modulated Phase Congruency (GMPC) model. This improved phase congruency model is defined by a Gamma Modulation Filter (GMF) and an adaptive noise model. Additionally, to reduce layover interference in SAR images, we introduce a GMPC-Harris feature point extraction method with layover perception. We also propose a matching method based on the SAR Modality Independent Neighborhood Fusion (SAR-MINF) descriptor, which fuses feature information from different neighborhoods. Furthermore, we present a graph-based overlap extraction algorithm and establish an automated workflow for large-scale TP extraction. Experiments show that the proposed SAR-MINF matching method increases the Correct Match Rate (CMR) by an average of 31.2% and the matching accuracy by an average of 57.8% compared with other prevalent SAR image matching algorithms. The proposed TP extraction algorithm can extract full-degree TPs with an accuracy of less than 0.5 pixels for more than 98% of 2-degree TPs and over 95% of multidegree TPs, meeting the requirements of DOM production.

Batik Image Representation using Multi Texton Co-occurrence Histogram

This paper introduces a novel approach to batik image representation using the texton-based and statistical Multi Texton Co-occurrence Histogram (MTCH). The MTCH framework is leveraged as a robust batik image descriptor, capable of encapsulating a comprehensive range of visual features, including the intricate interplay of color, texture, shape, and statistical attributes. The research extensively evaluates the effectiveness of MTCH through its application on two well-established public batik datasets, namely Batik 300 and Batik Nitik 960. These datasets serve as benchmarks for assessing the performance of MTCH in both classification and image retrieval tasks. In the classification domain, four distinct scenarios were explored, employing various classifiers: the K-Nearest Neighbors (K-NN), Support Vector Machine (SVM), Decision Tree (DT), and Naïve Bayes (NB). Each classifier was rigorously tested to determine its efficacy in correctly identifying batik patterns based on the MTCH descriptors. On the other hand, the image retrieval tasks were conducted using several distance metrics, including the Euclidean distance, City Block, Bray Curtis, and Canberra, to gauge the retrieval accuracy and the robustness of the MTCH framework in matching similar batik images. The empirical results derived from this study underscore the superior performance of the MTCH descriptor across all tested scenarios. The evaluation metrics, including accuracy, precision, and recall, indicate that MTCH not only achieves high classification performance but also excels in retrieving images with high similarity to the query. These findings suggest that MTCH is a highly effective tool for batik image analysis, offering significant potential for applications in cultural heritage preservation, textile pattern recognition, and automated batik classification systems.

ISFM-SLAM: dynamic visual SLAM with instance segmentation and feature matching.

Simultaneous Localization and Mapping (SLAM) is a technology used in intelligent systems such as robots and autonomous vehicles. Visual SLAM has become a more popular type of SLAM due to its acceptable cost and good scalability when applied in robot positioning, navigation and other functions. However, most of the visual SLAM algorithms assume a static environment, so when they are implemented in highly dynamic scenes, problems such as tracking failure and overlapped mapping are prone to occur. To deal with this issue, we propose ISFM-SLAM, a dynamic visual SLAM built upon the classic ORB-SLAM2, incorporating an improved instance segmentation network and enhanced feature matching. Based on YOLACT, the improved instance segmentation network applies the multi-scale residual network Res2Net as its backbone, and utilizes CIoU_Loss in the bounding box loss function, to enhance the detection accuracy of the segmentation network. To improve the matching rate and calculation efficiency of the internal feature points, we fuse ORB key points with an efficient image descriptor to replace traditional ORB feature matching of ORB-SLAM2. Moreover, the motion consistency detection algorithm based on external variance values is proposed and integrated into ISFM-SLAM, to assist the proposed SLAM systems in culling dynamic feature points more effectively. Simulation results on the TUM dataset show that the overall pose estimation accuracy of the ISFM-SLAM is 97% better than the ORB-SLAM2, and is superior to other mainstream and state-of-the-art dynamic SLAM systems. Further real-world experiments validate the feasibility of the proposed SLAM system in practical applications.

Pit slope unmanned aerial vehicle image matching based on hybrid scale space and boosted efficient binary local image descriptor

Pit slope unmanned aerial vehicle image matching based on hybrid scale space and boosted efficient binary local image descriptor

A novel binary hashing for agricultural scenery classification

In this research, we present PerceptHashing, a technique designed to categorize million-scale agricultural scenic images by incorporating human gaze shifting paths (GSPs) into a hashing framework. For each agricultural image, we identify visually and semantically significant object patches, such as fields, crops, and water bodies. These patches are linked to form a graphlet, establishing a network of spatially adjacent patches, and a GSP is then extracted using an active learning algorithm. The GSP reflects the distribution of human gaze across different regions of each agricultural scene, typically involving fewer than 12 regions, as validated by cross-validation. We then design a binary hashing framework that effectively leverages the semantics encoded in these GSPs. This framework integrates three key elements: (i) refinement of noisy labels, (ii) incorporation of deep image-level agricultural semantics, and (iii) updates to the adaptive data graph. The resulting hash codes from each GSP are converted into a kernelized visual descriptor for classification. To evaluate the influence of GSPs on agricultural image classification both qualitatively and quantitatively, we conducted an extensive user study comparing GSPs from typical observers with those from Alzheimer’s patients. The results demonstrate that: (1) the classification accuracy of 1.22 million agricultural aerial images using our approach is significantly higher than those processed by other methods, and (2) GSPs from 33 Alzheimer’s patients differ markedly from those of 37 normal observers, leading to notable differences in classification accuracy (accnormal=0.694\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$acc_{normal}=0.694$$\\end{document} versus accpatient=0.322\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$acc_{patient}=0.322$$\\end{document}).

Effect of micro-oxygenation on color of wines made with toasted vine-shoots.

Toasted vine-shoots (SEGs) are an enological tool to improve wines, to differentiate them, and to encourage sustainable wine production. Micro-oxygenation (MOX) is typically combined with the use of alternative oak products to simulate the oxygen transmission rate of traditional barrel aging, affecting wine color. Its use alongside SEGs has been studied. Tempranillo wines were treated with SEGs at two doses (12 and 24 g L-1) after malolactic fermentation at two fixed micro-oxygenation levels: (a) low, which received 6.24 ± 0.87 mg L-1·month-1 of oxygen; and, (b) high, which received 11.91 ± 0.71 mg L-1·month-1 of oxygen. The wines were bottled and stored for 6 months. At the end of the treatment, MOX affected the anthocyanins and color parameters, but not the enological characteristics. At this time, the anthocyanins content reduction presented a negatively significant correlation with oxygenfor wines treated with 12 g L-1. The factors that most influenced the development of color parameters during the time after bottling were the period for which the wine evolved in the bottle and SEG dose. The visual sensorial descriptors showed an evolution according to aged red wines, but without differences according to the SEG-MOX treatments. The SEG-MOX treatments caused significant changes in wine color. It would be advisable to adjust SEG-MOX techniques to increase their effectiveness. © 2024 The Author(s). Journal of the Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

EQID: Entangled quantum image descriptor an approach for early plant disease detection

In present day agriculture, early and accurate identification of plant diseases is essential for prompt response, which protects crop quality and output. This paper presents the Entangled Quantum-Inspired Deep learning model (EQID), a unique method that improves feature representation and classification in plant disease prediction by utilizing the concepts of quantum computing. Two different datasets with images of potatoes and tomatoes as leaves were used to test the EQID model, which performed better than traditional models. EQID obtained 98.96% accuracy, 98.98% precision, 98.96% recall, and 98.90% F1 score on images of potato leaves. For tomato leaves, comparable outcomes were noted, with accuracy, precision, recall, and F1 score all above 99.61%. The accuracy of disease prediction is greatly increased by the efficient and effective feature representation made possible by the EQID model's inclusion of quantum computing techniques. Additionally, the model outperformed other cutting-edge models such as DenseNet-121, VGGNet 16, and Xception Net, illustrating the potentially revolutionary effects of quantum-inspired models in agriculture. Future work will focus on applying the EQID model to a broader range of crops and plant diseases, as well as incorporating additional data sources to further enhance the model's predictive capabilities.

FilterGNN: Image feature matching with cascaded outlier filters and linear attention

The cross-view matching of local image features is a fundamental task in visual localization and 3D reconstruction. This study proposes FilterGNN, a transformer-based graph neural network (GNN), aiming to improve the matching efficiency and accuracy of visual descriptors. Based on high matching sparseness and coarse-to-fine covisible area detection, FilterGNN utilizes cascaded optimal graph-matching filter modules to dynamically reject outlier matches. Moreover, we successfully adapted linear attention in FilterGNN with post-instance normalization support, which significantly reduces the complexity of complete graph learning from O(N2) to O(N). Experiments show that FilterGNN requires only 6% of the time cost and 33.3% of the memory cost compared with SuperGlue under a large-scale input size and achieves a competitive performance in various tasks, such as pose estimation, visual localization, and sparse 3D reconstruction.

Feature-level Fusion vs. Score-level Fusion for Image Retrieval Based on Pre-trained Deep Neural Networks

Today’s complex multimedia content made retrieving images similar to the user’s query from the database a challenging task. The performance of a Content-Based Image Retrieval System (CBIR) system highly depends on the image representation in a form of low-level features and similarity measurement. The traditional visual descriptors that do not provide good prior domain knowledge could lead to poor performance retrieval results. On the other hand, Deep Convolutional Neural Networks (DCNNs) have recently achieved a remarkable success as methods for image classification in various domains. Recently, pre-trained deep convolution neural networks on thousands of classes have the ability to extract very accurate and representative features which, in addition to classification, can also be successfully used in image retrieval systems. ResNet152, GoogLeNet and InceptionV3 are some of the effective and successful examples of pre-trained DCNNs recently applied in a computer vision tasks such as object recognition, clustering, and classification. In this paper, two approaches for a CBIR system, namely early fusion and late fusion, have been presented and compared. The early fusion utilizes concatenation of the features extracted by each possible pair of DCNNs, that is ResNet152-GoogLeNet, ResNet152-InceptionV3, and GoogLeNet-InceptionV3, and the late fusion apply CombSum method with Z-Score standardization to combine the score results provided by each DCNN of the aforementioned pairs. In the experiments on a popular WANG dataset it has been shown that late fusion approach slightly outperforms early fusion approach. The best performance of our experiments in terms of Average Precision (AP) for the top 20 results reaches 96.82%.

A Novel Hybrid Attention-Based Dilated Network for Depression Classification Model from Multimodal Data Using Improved Heuristic Approach

Automatic depression classification from multimodal input data is a challenging task. Modern methods use paralinguistic information such as audio and video signals. Using linguistic information such as speech signals and text data for depression classification is a complicated task in deep learning models. Best audio and video features are built to produce a dependable depression classification system. Textual signals related to depression classification are analyzed using text-based content data. Moreover, to increase the achievements of the depression classification system, audio, visual, and text descriptors are used. So, a deep learning-based depression classification model is developed to detect the person with depression from multimodal data. The EEG signal, Speech signal, video, and text are gathered from standard databases. Four stages of feature extraction take place. In the first stage, the features from the decomposed EEG signals are attained by the empirical mode decomposition (EMD) method, and features are extracted by means of linear and nonlinear feature extraction. In the second stage, the spectral features of the speech signals from the Mel-frequency cepstral coefficients (MFCC) are extracted. In the third stage, the facial texture features from the input video are extracted. In the fourth stage of feature extraction, the input text data are pre-processed, and from the pre-processed data, the textual features are extracted by using the Transformer Net. All four sets of features are optimally selected and combined with the optimal weights to get the weighted fused features using the enhanced mountaineering team-based optimization algorithm (EMTOA). The optimal weighted fused features are finally given to the hybrid attention-based dilated network (HADN). The HDAN is developed by combining temporal convolutional network (TCN) with bidirectional long short-term memory (Bi-LSTM). The parameters in the HDAN are optimized with the assistance of the developed EMTOA algorithm. At last, the classified output of depression is obtained from the HDAN. The efficiency of the developed deep learning HDAN is validated by comparing it with various traditional classification models.

In search for representative elementary volume (REV) within heterogeneous materials: A survey of scalar and vector metrics using porous media as an example

The Representative Elementary Volume (REV) concept, a cornerstone in porous system heterogeneity assessment, was initially conceived to determine the minimal domain volume suitable for homogenization and upscaling. However, the definition of REV and usability in continuum-scale models is vague. In this study, we conduct comprehensive REV analyses on multiple samples, encompassing a range of scalar and vector metrics. Our investigation probes the representativity of crucial medium characteristics, including porosity, permeability, and Euler density, alongside descriptors rooted in pore-network statistics, correlation functions, and persistence diagrams. We explore both deterministic and statistical REV sizes (dREV and sREV), facilitating a robust comparative assessment. Crucially, we introduce an novel methodology tailored for harnessing vector metrics, known for their ability to reveal intricate structural insights. Our results underscore the superiority of the sREV approach, particularly for low-content metrics, addressing inherent limitations of dREV in characterizing homogeneities in such cases. Furthermore, the sREV approach incorporates stationarity analysis into REV evaluation, ensuring result consistency between sREV and dREV under stationarity conditions. Encouragingly, our findings suggest that high-information-content metrics, notably correlation functions combined with persistence diagrams, have the potential to establish a universal REV for steady-state physical properties. This proposition warrants further verification through a comprehensive assessment and comparison of REV values across major physical properties. REV analysis plays a pivotal role not only in assessing medium properties but also in scrutinizing different descriptors of 3D images – we note that REV analysis and image/field stationarity analysis are ultimately the same techniques under the hood. The discussion based on obtained results and recent finding by other researchers advances the understanding of REV within porous media, introduces a versatile methodology with broader applications, and is expected to be useful in numerous fields including materials science, cosmology, machine learning, and more. We redefine the classical definition of REV by adding stationarity condition and upper/lower bounds on its volume. While for simplicity, in this work we shall mainly focus on porous media as immediately applicable to digital rock, petrophysics, hydrology and soil physics problems, the developed mythology can be applied to other material types - composites, biological tissues, granular matter, food engineering and numerous other types of matter.

Novel Content Based Image Retrieval—Features of Correlated Visual Textons and MQLPP Descriptor

Novel Content Based Image Retrieval—Features of Correlated Visual Textons and MQLPP Descriptor

An effective ensemble learning approach for classification of glioma grades based on novel MRI features

The preoperative diagnosis of brain tumors is important for therapeutic planning as it contributes to the tumors’ prognosis. In the last few years, the development in the field of artificial intelligence and machine learning has contributed greatly to the medical area, especially the diagnosis of the grades of brain tumors through radiological images and magnetic resonance images. Due to the complexity of tumor descriptors in medical images, assessing the accurate grade of glioma is a major challenge for physicians. We have proposed a new classification system for glioma grading by integrating novel MRI features with an ensemble learning method, called Ensemble Learning based on Adaptive Power Mean Combiner (EL-APMC). We evaluate and compare the performance of the EL-APMC algorithm with twenty-one classifier models that represent state-of-the-art machine learning algorithms. Results show that the EL-APMC algorithm achieved the best performance in terms of classification accuracy (88.73%) and F1-score (93.12%) over the MRI Brain Tumor dataset called BRATS2015. In addition, we showed that the differences in classification results among twenty-two classifier models have statistical significance. We believe that the EL-APMC algorithm is an effective method for the classification in case of small-size datasets, which are common cases in medical fields. The proposed method provides an effective system for the classification of glioma with high reliability and accurate clinical findings.

Convolutional MLP orthogonal fusion of multiscale features for visual place recognition

Visual place recognition (VPR) involves obtaining robust image descriptors to cope with differences in camera viewpoints and drastic external environment changes. Utilizing multiscale features improves the robustness of image descriptors; however, existing methods neither exploit the multiscale features generated during feature extraction nor consider the feature redundancy problem when fusing multiscale information when image descriptors are enhanced. We propose a novel encoding strategy—convolutional multilayer perceptron orthogonal fusion of multiscale features (ConvMLP-OFMS)—for VPR. A ConvMLP is used to obtain robust and generalized global image descriptors and the multiscale features generated during feature extraction are used to enhance the global descriptors to cope with changes in the environment and viewpoints. Additionally, an attention mechanism is used to eliminate noise and redundant information. Compared to traditional methods that use tensor splicing for feature fusion, we introduced matrix orthogonal decomposition to eliminate redundant information. Experiments demonstrated that the proposed architecture outperformed NetVLAD, CosPlace, ConvAP, and other methods. On the Pittsburgh and MSLS datasets, which contained significant viewpoint and illumination variations, our method achieved 92.5% and 86.5% Recall@1, respectively. We also achieved good performances—80.6% and 43.2%—on the SPED and NordLand datasets, respectively, which have more extreme illumination and appearance variations.

Artificial intelligence model for tumoral clinical decision support systems

Background and ObjectiveComparative diagnostic in brain tumor evaluation makes possible to use the available information of a medical center to compare similar cases when a new patient is evaluated. By leveraging Artificial Intelligence models, the proposed system is able of retrieving the most similar cases of brain tumors for a given query. The primary objective is to enhance the diagnostic process by generating more accurate representations of medical images, with a particular focus on patient-specific normal features and pathologies. A key distinction from previous models lies in its ability to produce enriched image descriptors solely from binary information, eliminating the need for costly and difficult to obtain tumor segmentation. MethodsThe proposed model uses Artificial Intelligence to detect patient features to recommend the most similar cases from a database. The system not only suggests similar cases but also balances the representation of healthy and abnormal features in its design. This not only encourages the generalization of its use but also aids clinicians in their decision-making processes. This generalization makes possible for future research in different medical diagnosis areas with almost not any change in the system. ResultsWe conducted a comparative analysis of our approach in relation to similar studies. The proposed architecture obtains a Dice coefficient of 0.474 in both tumoral and healthy regions of the patients, which outperforms previous literature. Our proposed model excels at extracting and combining anatomical and pathological features from brain Magnetic Resonances (MRs), achieving state-of-the-art results while relying on less expensive label information. This substantially reduces the overall cost of the training process. Our findings highlight the significant potential for improving the efficiency and accuracy of comparative diagnostics and the treatment of tumoral pathologies. ConclusionsThis paper provides substantial grounds for further exploration of the broader applicability and optimization of the proposed architecture to enhance clinical decision-making. The novel approach presented in this work marks a significant advancement in the field of medical diagnosis, particularly in the context of Artificial Intelligence-assisted image retrieval, and promises to reduce costs and improve the quality of patient care using Artificial Intelligence as a support tool instead of a black box system.

A Mighty Image Retrieval Descriptor Based on Machine Learning and Gaussian Derivative Filter

The development of new image descriptor has always been an important topic to improve the efficiency of content- based image classification and retrieval. Improvements and developments in machine learning and deep learning algorithms as well as artificial intelligence algorithms are widely used by researchers to obtain effective CBIR descriptors. In our article, we will present a robust image descriptor, extended by machine learning and deep learning algorithms. The descriptor is provided through a Gaussian derivative filter scaffold named GDF-HOG with an enhanced convolutional neural network (CNN) AlexNet, to reduce the dimensions we used the principal component analysis algorithm. The experimental results were carried out on Oliva and Torralba, Caltech-101, Wang and Coil100 datasets. Experiments show that the accuracy of the proposed method is 98.23% for Coil-100%, 95.92% for Corel-1000, value 87.17 and 94.6% for Oliva and Torralba. In comparison our results with other descriptor image classifiers show that they achieved accuracy increases of 0.12% on average and up to 3.23%. These experimental results affirm the advantage of the proposed descriptor over existing systems based in terms of average accuracy. the proposed descriptor improves the precision, and also reduces the complexity of the calculation.

Efficient Crowd Anomaly Detection Using Sparse Feature Tracking and Neural Network

Crowd anomaly detection is crucial in enhancing surveillance and crowd management. This paper proposes an efficient approach that combines spatial and temporal visual descriptors, sparse feature tracking, and neural networks for efficient crowd anomaly detection. The proposed approach utilises diverse local feature extraction methods, including SIFT, FAST, and AKAZE, with a sparse feature tracking technique to ensure accurate and consistent tracking. Delaunay triangulation is employed to represent the spatial distribution of features in an efficient way. Visual descriptors are categorised into individual behaviour descriptors and interactive descriptors to capture the temporal and spatial characteristics of crowd dynamics and behaviour, respectively. Neural networks are then utilised to classify these descriptors and pinpoint anomalies, making use of their strong learning capabilities. A significant component of our study is the assessment of how dimensionality reduction methods, particularly autoencoders and PCA, affect the feature set’s performance. This assessment aims to balance computational efficiency and detection accuracy. Tests conducted on benchmark crowd datasets highlight the effectiveness of our method in identifying anomalies. Our approach offers a nuanced understanding of crowd movement and patterns by emphasising both individual and collective characteristics. The visual and local descriptors facilitate high-level analysis by closely relating to semantic information and crowd behaviour. The analysis observed shows that this approach offers an efficient framework for crowd anomaly detection, contributing to improved crowd management and public safety. The proposed model achieves accuracy of 99.5 %, 96.1%, 99.0% and 88.5% in the UMN scenes 1, 2, and 3 and violence in crowds datasets, respectively.

Convolutional Cross-View Pose Estimation.

We propose a novel end-to-end method for cross-view pose estimation. Given a ground-level query image and an aerial image that covers the query's local neighborhood, the 3 Degrees-of-Freedom camera pose of the query is estimated by matching its image descriptor to descriptors of local regions within the aerial image. The orientation-aware descriptors are obtained by using a translationally equivariant convolutional ground image encoder and contrastive learning. The Localization Decoder produces a dense probability distribution in a coarse-to-fine manner with a novel Localization Matching Upsampling module. A smaller Orientation Decoder produces a vector field to condition the orientation estimate on the localization. Our method is validated on the VIGOR and KITTI datasets, where it surpasses the state-of-the-art baseline by 72% and 36% in median localization error for comparable orientation estimation accuracy. The predicted probability distribution can represent localization ambiguity, and enables rejecting possible erroneous predictions. Without re-training, the model can infer on ground images with different field of views and utilize orientation priors if available. On the Oxford RobotCar dataset, our method can reliably estimate the ego-vehicle's pose over time, achieving a median localization error under 1 m and a median orientation error of around 1 ° at 14 FPS.

Keyframe recommendation based on feature intercross and fusion

Keyframe extraction can effectively help users quickly understand video content. Generally, keyframes should be representative of the video content and simultaneously be diverse to reduce redundancy. Aiming to find the features of frames and filter out representative frames of the video, we propose a method of keyframe recommendation based on feature intercross and fusion (KFRFIF). The method is inspired by the implied relations between keyframe-extraction problem and recommendation problem. First, we investigate the application of a recommendation framework to the keyframe extraction problem. Second, the architecture of the proposed KFRFIF is put forward. Then, an algorithm for extracting intra-frame image features based on the combination of multiple image descriptors is proposed. An algorithm for extracting inter-frame distance features based on the combination of multiple distance calculation methods is designed. Moreover, A recommendation model based on feature intercross and fusion is put forward. An ablation study is further performed to verify the effectiveness of the submodule. Ultimately, the experimental results on four datasets with five outstanding approaches indicate the superior performance of our approach.

Multi-scale fractal Fourier Ptychographic microscopy to assess the dose-dependent impact of copper pollution on living diatoms

Accumulation of bioavailable heavy metals in aquatic environment poses a serious threat to marine communities and human health due to possible trophic transfers through the food chain of toxic, non-degradable, exogenous pollutants. Copper (Cu) is one of the most spread heavy metals in water, and can severely affect primary producers at high doses. Here we show a novel imaging test to assay the dose-dependent effects of Cu on live microalgae identifying stress conditions when they are still capable of sustaining a positive growth. The method relies on Fourier Ptychographic Microscopy (FPM), capable to image large field of view in label-free phase-contrast mode attaining submicron lateral resolution. We uniquely combine FPM with a new multi-scale analysis method based on fractal geometry. The system is able to provide ensemble measurements of thousands of diatoms in the liquid sample simultaneously, while ensuring at same time single-cell imaging and analysis for each diatom. Through new image descriptors, we demonstrate that fractal analysis is suitable for handling the complexity and informative power of such multiscale FPM modality. We successfully tested this new approach by measuring how different concentrations of Cu impact on Skeletonema pseudocostatum diatom populations isolated from the Sarno River mouth.