Image analysis

In this thesis, a methodology and framework were created to detect the seawalls accurately and efficiently in low coastal areas and was evaluated in the study area of Hallandale Beach City, Broward County, Florida. Aerial images collected from the Florida Department of Transportation (FDOT) were processed using eCognition Developer software for Multi-Resolution Segmentation and Classification of objects. Two classification approaches, pixel-based image analysis, and the object-based image analysis (OBIA) method were applied for image classification. However, Pixel based classification was discarded for having less accuracy in output. Three techniques within object-based classification-machine learning technique, knowledge-based technique and machine learning followed by knowledge-based technique were used to compare the most efficient method of classification. While performing the machine learning technique, three algorithms: Random Forest, support vector machine and decision tree were applied to test the best algorithm. Of all the approaches used, the combination of machine learning and a knowledge-based method was able to map the sea wall effectively.

Machine learning has been utilized in bio-imaging in recent years, however as it is relatively new and evolving, some researchers who wish to utilize machine learning tools have limited access because of a lack of programming knowledge. In electron microscopy (EM), immunogold labeling is commonly used to identify the target proteins, however the manual annotation of the gold particles in the images is a time-consuming and laborious process. Conventional image processing tools could provide semi-automated annotation, but those require that users make manual adjustments for every step of the analysis. To create a new high-throughput image analysis tool for immuno-EM, I developed a deep learning pipeline that was designed to deliver a completely automated annotation of immunogold particles in EM images. The program was made accessible for users without prior programming experience and was also expanded to be used on different types of immuno-EM images.

Traditional sand analysis is labor and cost-intensive, entailing specialized equipment and operators trained in geological analysis. Even a small step to automate part of the traditional geological methods could substantially improve the speed of such research while removing chances of human error. Digital image analysis techniques and computer vision have been well developed and applied in various fields but rarely explored for sand analysis. This research explores capabilities of remote sensing digital image analysis techniques, such as object-based image analysis (OBIA), machine learning, digital image analysis, and photogrammetry to automate or semi-automate the traditional sand analysis procedure. Here presented is a framework combining OBIA and machine learning classification of microscope imagery for use with unconsolidated terrigenous beach sand samples. Five machine learning classifiers (RF, DT, SVM, k-NN, and ANN) are used to model mineral composition from images of ten terrigenous beach sand samples. Digital image analysis and photogrammetric techniques are applied and evaluated for use to characterize sand grain size and grain circularity (given as a digital proxy for traditional grain sphericity). A new segmentation process is also introduced, where pixel-level SLICO superpixel segmentation is followed by spectral difference segmentation and further levels of superpixel segmentation at the object-level. Previous methods of multi-resolution and superpixel segmentation at the object level do not provide the level of detail necessary to yield optimal sand grain-sized segments. In this proposed framework, the DT and RF classifiers provide the best estimations of mineral content of all classifiers tested compared to traditional compositional analysis. Average grain size approximated from photogrammetric procedures is comparable to traditional sieving methods, having an RMSE below 0.05%. The framework proposed here reduces the number of trained personnel needed to perform sand-related research. It requires minimal sand sample preparation and minimizes user-error that is typically introduced during traditional sand analysis.

The Florida Everglades ecosystem is experiencing increasing threats from anthropogenic modification of water flow, spread of invasive species, sea level rise (SLR), and more frequent and/or intense hurricanes. Restoration efforts aimed at rehabilitating these ongoing and future disturbances are currently underway through the implementation of the Comprehensive Everglades Restoration Plan (CERP). Efficacy of these restoration activities can be further improved with accurate and site-specific information on the current state of the coastal wetland habitats. In order to produce such assessments, digital datasets of the appropriate accuracy and scale are needed. These datasets include orthoimagery to delineate wetland areas and map vegetation cover as well as accurate 3-dimensional (3-D) models to characterize hydrology, physiochemistry, and habitat vulnerability.

Two photon microscopy is one of the fastest growing methods of in-vivo imaging of the brain. It has the capability of imaging structures on the scale of 1μm. At this scale the wavelength of the imaging field (usually near infra-red), is comparable to the size of the structures being imaged, which makes the use of ray optics invalid. A better understanding is needed to predict the result of introducing different media into the light path. We use Wolf's integral, which is capable of fulfilling these needs without the shortcomings of ray optics. We predict the effects of aberrating media introduced into the light path like glass cover-slips and then correct the aberration using the same method. We also create a method to predict aberrations when the interfaces of the media in the light-path are not aligned with the propagation direction of the wavefront.

How does spatial organization of objects affect the perceptual processing of a scene? Surprisingly, little research has explored this topic. A few studies have reported that, when simple, homogenous stimuli (e.g., dots), are presented in a regular formation, they are judged to be more numerous than when presented in a random configuration (Ginsburg, 1976; 1978). However, these results may not apply to real-world objects. In the current study, fewer objects were believed to be on organized desks than their disorganized equivalents. Objects that are organized may be more likely to become integrated, due to classic Gestalt principles. Consequently, visual search may be more difficult. Such object integration may diminish saliency, making objects less apparent and more difficult to find. This could explain why, in the present study, objects on disorganized desks were found faster.

High-resolution imagery is becoming readily available to the public. Private firms and government organizations are using high-resolution images but are running into problems with storage space and processing time. High-resolution images are extremely large files, and have proven cumbersome to work with and control. By resampling fine resolution imagery to a lower resolution, storage and processing space can be dramatically reduced. Fine-resolution imagery is not needed to map most features and resampled high-resolution imagery can be used as a replacement for low-resolution satellite imagery in some cases. The effects of resampling on the spectral quality of a high-resolution image can be demonstrated by answering the following questions: (1) Is the quality of spectral information on a color infrared DOQQ comparable to SPOT and TM Landsat satellite imagery for the purpose of digital image classification? (2) What is the appropriate resolution for mapping surface features using high-resolution imagery for spectral categories of information? (3) What is the appropriate resolution for mapping surface features using high-resolution imagery for land-use land-cover information?