A5029259050- Smart Agriculture and AI
- Advanced Image and Video Retrieval Techniques
- Leaf Properties and Growth Measurement
- Remote Sensing in Agriculture
- Image and Signal Denoising Methods
- Medical Image Segmentation Techniques
- Remote Sensing and LiDAR Applications
- Advanced Neural Network Applications
- Genetic and phenotypic traits in livestock
- semigroups and automata theory
- Reproductive Biology and Fertility
- Image Retrieval and Classification Techniques
- Ovarian function and disorders
- DNA and Biological Computing
- Image Enhancement Techniques
- Advanced Image Fusion Techniques
- Ovarian cancer diagnosis and treatment
- Reproductive Physiology in Livestock
- Visual Attention and Saliency Detection
- Periodontal Regeneration and Treatments
- Chromosomal and Genetic Variations
- Advanced Vision and Imaging
- Oral and Maxillofacial Pathology
- Mathematical Dynamics and Fractals
- Image Processing and 3D Reconstruction
University of Saskatchewan
2014-2023
Western University
1997-2003
With many thyroid nodules being incidentally detected, it is important to identify as malignant possible while excluding those that are highly likely be benign from fine needle aspiration (FNA) biopsies or surgeries. This paper presents a computer-aided diagnosis (CAD) system for classifying in ultrasound images. We use deep learning approach extract features Ultrasound images pre-processed calibrate their scale and remove the artifacts. A pre-trained GoogLeNet model then fine-tuned using...
Defocus blur is extremely common in images captured using optical imaging systems. It may be undesirable, but also an intentional artistic effect, thus it can either enhance or inhibit our visual perception of the image scene. For tasks, such as restoration and object recognition, one might want to segment a partially blurred into non-blurred regions. In this paper, we propose sharpness metric based on local binary patterns robust segmentation algorithm separate in- out-of-focus The proposed...
Lodging, the permanent bending over of food crops, leads to poor plant growth and development. Consequently, lodging results in reduced crop quality, lowers yield, makes harvesting difficult. Plant breeders routinely evaluate several thousand breeding lines, therefore, automatic detection prediction is great value aid selection. In this paper, we propose a deep convolutional neural network (DCNN) architecture for classification using five spectral channel orthomosaic images from canola wheat...
Lodging in agricultural crops is the permanent displacement of a plant from its upright position [2]. It may be caused by several weather and environmental conditions. Harvesting severely lodged take twice as much time results reduced yield. Plant breeders seek to identify select for lodging-resistant varieties. Currently lodging breeding trials assessed manual observation. This can both consuming subjective. The objective this work automate detection plots using drone imagery. Texture...
Polycystic ovary syndrome (PCOS) is an endocrine abnormality with multiple diagnostic criteria due to its heterogenic manifestations. One of the includes analysis ultrasound images ovaries for detection number, size, and distribution follicles within ovary. This involves manual tracing counting on determine presence a polycystic (PCO). We describe novel method that automates PCO detection. Our algorithm segmentation from images, quantifying attributes automatically segmented using...
We present a refinement of histogram equalization which uses both global and local information to remap the image grey levels. Local properties, we generally call neighborhood metrics, are used subdivide bins that would be otherwise indivisible using classical (HE). Choice metric influences how subdivided, affording opportunity for additional contrast enhancement. experimental results two specific metrics compare (LHE). find our methods can provide an improvement in enhancement versus HE,...
Unmanned Aerial Vehicles (UAVs) paired with image detection and segmentation techniques can be used to extract plant phenotype information of individual breeding or research plots. Each plot contains plants a single genetic line. Breeders are interested in selecting lines preferred phenotypes (physical traits) that increase crop yield resilience. Automated plots would enable automatic monitoring quantification phenotypes, allowing faster selection process requires much fewer person-hours...
Deep learning has shown potential in domains with large-scale annotated datasets. However, manual annotation is expensive, time-consuming, and tedious. Pixel-level annotations are particularly costly for semantic segmentation images dense irregular patterns of object instances, such as plant images. In this work, we propose a method developing high-performing deep models utilizing little annotation. As use case, focus on wheat head segmentation. We synthesize computationally dataset-using...
Rapamycin is a well-known inhibitor of the Target (TOR) signaling cascade; however, impact this drug on global genome function and organization in normal primary cells poorly understood. To explore impact, we treated human foreskin fibroblasts with rapamycin observed decrease cell proliferation without causing death. Upon treatment chromosomes 18 10 were repositioned to location similar that induced into quiescence by serum reduction. Although changes positioning occurred, comparative...
Abstract Because of the increasing global population, changing climate, and consumer demands for safe, environmentally friendly, high‐quality food, plant breeders strive higher yield cultivars by monitoring specific phenotypes. Developing new crop through current methods is time‐consuming, sometimes subjective, based on subsampling microplots. High‐throughput phenotyping using unmanned aerial vehicle‐acquired orthomosaic images breeding trials improves simplifies this labor‐intensive...
An algorithm for the automated segmentation of epithelial tissue in digital images histologic sections odontogenic cysts (cysts originating from residual epithelium) is presented. The features an image standardization process that greatly reduces variation luminance and chrominance between due to variations sample preparation. Segmentation regions uses based on binary graph cuts where weights depend probabilities obtained colour histogram models epithelium stroma regions. Algorithm training...
This paper proposes a new texture enhancement method which uses an image decomposition that allows different visual characteristics of textures to be represented by separate components in contrast with previous methods either enhance indirectly or represent all information using single component. Our is intended used as preprocessing step prior the use texture-based segmentation algorithms. modification morphological component analysis (MCA) separated into multiple each representing...
Abstract Accurate segmentation of root system architecture (RSA) from 2D images is an important step in studying phenotypic traits systems. Various approaches to image exist but many them are not well suited the thin and reticulated structures characteristic The findings presented here describe approach RSA that takes advantage inherent structural properties system, a network we call ITErRoot. We have also generated novel dataset which utilizes annotation tool developed for producing high...
Traditionally, the preservation of archaeological data has been limited by cost materials and physical space required to store them, but for last 20 years, increasing amounts digital have generated stored online. New techniques in photography document scanning dramatically increased amount that can be retained format, while at same time reducing production storage. Vast numbers hand written notes, grey literature documents, images assemblages, contexts, artefacts made available However,...
Use of digital image analysis for the identification seeds has not been recognized as a validated method. Image seed previously studied, and good recognition rates have achieved. However, data sets used in these experiments either contain very few groups non-verified specimens or little representation intra-species variations. This study considered set containing that were verified to represent species typical population variation, well look-alike share same morphological appearance,...
Gaussian modulated sinusoids are used in S-transform to extract time-local and space-local spectral information. Similar data sets recorded at neighboring spatial locations may be with cross analysis determine frequency localized velocity spectrum. The 2D is image for space wavenumber spectra. Local changes the spectrum define textural boundaries on images. This paper summarizes several of research projects involving S-transforms currently progress University Western Ontario including...
A ‘virtual histology’ can be thought of as the ‘staining’ a digital ultrasound image via processing techniques in order to enhance visualisation differences echotexture different types tissues. Several candidate image-processing algorithms for virtual histology using images bovine ovary were studied. The evaluated qualitatively ability visual intra-ovarian structures and quantitatively, standard texture description features, increase statistical ovarian Certain found create textures that...
Acquiring high-resolution images in the field for image-based crop phenotyping is typically performed by complicated, custom built "pheno-mobiles." In this paper, we demonstrate that large datasets of row can be easily acquired with consumer cameras attached to a regular tractor. Localization and labeling individual rows plants are computer vision approach, rather than sophisticated real-time geo-location hardware on We evaluate our approach cropping early-season from Brassica carinata trial...