An Improved Automatic Shape Feature Extraction Method Based on Template Matching

Image shape extraction is an important step in the image analysis, AI electronic industry and automation, as well as a significant part of content-based image retrieval(CBIR), which cannot be separated from contour extraction. However, traditional approach of the border following algorithm is susceptible to noise interference, thus the shape extracted is always complex in real images and cannot express feature of the target image well. Therefore, an improved shape feature extraction method is proposed, which converts color space into HSV model when preprocessing, filters contour by area size, merges adjacent contours by drawing convex hull and filters with template shapes. Lastly, this method is tested on UAV123 and YCB_Video dataset, which showed that the proportion of valid contour improved from less than 10% to 87.7% based on border following algorithm. In the experiment of OPenCV open source library in Visual Studio environment, we hope to improve the extraction efficiency of shape features.

Mathematical Modeling for Ceramic Shape 3D Image Based on Deep Learning Algorithm

Ceramic image shape 3D image modeling focuses on of ceramic that was obtained from the camera imaging equipment such as 2D images, by normalization, gray, filtering denoising, wavelet image sharpening edge enhancement, binarization, and shape contour extraction pretreatment processes such as extraction ceramic image shape edge profile, again, according to the image edge extraction and elliptic rotator ceramics phenomenon. The image distortion effect was optimized by self-application, and then the deep learning modeler was used to model the side edge contour. Finally, the 3D ceramic model of the rotating body was restored according to the intersection and central axis of the extracted contour. By studying the existing segmentation methods based on deep learning, the automatic segmentation of target ceramic image and the effect of target edge refinement and optimization are realized. After extracting and separating the target ceramics from the image, we processed the foreground image of the target into a three-dimensional model. In order to reduce the complexity of the model, a 3D contextual sequencing model is adopted to encode the hidden space features along the channel dimensions, to extract the causal correlation between channels. Each module in the compression framework is optimized by a rate-distortion loss function. The experimental results show that the proposed 3D image modeling method has significant advantages in compression performance compared with the optimal 2D 3D image modeling method based on deep learning, and the experimental results show that the performance of the proposed method is superior to JP3D and HEVC methods, especially at low bit rate points.

Ultrasound PIV Uncertainty Quantification

Ultrasound Particle image velocimetry (UPIV) is a non-invasive flow measurement technique where acousticopaque flow tracers are injected into a working fluid and ensonified to create ultrasound images. These images are processed using PIV cross-correlation based algorithms to measure the velocity field (Kim et al., 2004). UPIV is useful for opaque flows, primarily where complex flows exist, accordingly, it is used in many industrial and clinical research applications such as studying intracardiac flow (Crase et al., 2007). Furthermore, the measurement provides suitable temporal and spatial resolutions for improved diagnostic metrics. Mentioned applications and the sensitive diagnostic industrial and clinical decisions made based on these measurements intensifies the importance of characterizing the UPIV measurement accuracy and associated uncertainty. However, quantifying UPIV measurement uncertainty is non-trivial due to the complexity of possible uncertainty sources, their combination, and propagation through the measurement chain. The formation of a particle image by ultrasound significantly differs from optical imaging, introducing unique aspects to image quality that must be considered. Particle images are formed across several ultrasound scan lines, yielding an elliptical particle image shape. Furthermore, the particle’s reflected pressure wave is converted to a digital signal that undergoes signal modulation, and this process forms a non-Gaussian point spread function (PSF) along the scan line direction. Additionally, clusters of tracers produce a single, bright image intensity and speckle image pattern. Compared to conventional PIV images, UPIV incurs significantly higher image noise due to lack of filtration for the ultrasound reflection of the non-tracer obstacles.

Representing nonhuman animals as equals: An ecolinguistic analysis of vegan campaigns

Ecolinguistics studies the interactions between language and ecology. It investigates whether the stories created by language are destructive or beneficial to all the constituents of the environment. In search of positive stories for our environment, this article focuses on vegan campaigns which generally bring awareness about veganism that, in turn, advocates protection of nonhuman animals and abstention from their exploitation. Nonhuman animals are part of the ecosystem and the way they are portrayed in language may determine the relationship between human and nonhuman animals. As vegan campaigns refer to nonhuman animals as sentient living beings, it is important to analyze whether the language and image of these campaigns articulate their purposes and create beneficial stories for nonhuman species. This article explores the stories regarding nonhuman animals in 27 posters of the vegan campaign “Go Vegan World” and examines how these stories are shaped and whether they are aligned with vegan values. The study is approached from an ecolinguistic perspective with a focus on multimodality where the language was analyzed through van Leeuwen’s Social Actor and Social Action theory, and the image was analyzed with Kress and van Leeuwen’s Grammar of Visual Design. Further, the analysis involves the ecosophy defined as a personal ecological philosophy of relationships between human and nonhuman animals, plants, and the physical environment. The findings suggest that the campaign language and image shape three stories: salience where nonhuman animals are individuals with their own feelings and lives; conviction that nonhuman animals matter as much as humans; ideology where biocentrism is promoted. By comparing these stories with the article’s ecosophy, an ecolinguistic analysis showed that they are largely beneficial in representing nonhuman animals as sentient living beings who are equal to humans.

IMAGE SHAPE COMPARISON IN MOSAIC DIFFUSION MORPHOLOGIES

In this paper, developing the early proposed unified approach to the representation of morphological models, we show that since morphological models in the attribute representation have the same mathematical form as images, the morphological models themselves can be the subject of morphological analysis and can be directly compared in form with using morphological operators. In this case, the previously introduced formalism of mosaic diffusion morphology is used.In the framework of mosaic diffusion morphology, two alternative descriptions of the projection of the image on the form are considered, which are based on a clear model and a fuzzy model of the form respectively. It is shown that the projection operator in the second case is a one-sorted diffuse operator that makes direct comparision of model to model instead of image to model. In this case, a fuzzy mosaic model appears in this scheme as a projection of a clear mosaic model onto another clear mosaic model. Based on this shape-to-shape projection idea, we propose the new version of Pytiev morphology tools for shape comparison: the morphological shape difference map, the morphological quasi-distance between shapes, as well as the Morphological Shape Correlation Coefficient (MSCC). We show that MSCC from the resource parameters of the reciprocal model has exactly the same formula as the standard effective morphological correlation coefficient proposed earlier based on statistical averaging of projected images.

Comparison of statistical properties for various morphological filters based on mosaic image shape models

We consider the statistical properties of different mosaic filters. We demonstrate that in Pitiev's morphology, the measure of shape complexity is directly related to the shape simplicity measure based on morphological correlation coefficient (MCC). Based on MCC, we introduce the normalized morphological simplification index (NMSI). Using NMSI, we show that the simpler the mosaic shape, the more shape simplification is provided by the corresponding Pyt'ev projector. For the examples of mean and median mosaic filters, we address the problem of different operator comparison. In this context we introduce the concept of statistically simplifying morphological operators. Morphological correlation of mosaic shape and diffusion mosaic operator is considered. We prove that the NMSI for the diffusion mosaic operator is not related to the complexity for the corresponding diffusion shape kernel. Thus, a principal qualitative difference in the relationship between relational and operator models for diffuse and projective mosaic linear filters is demonstrated.

Calibration Techniques For Low-Cost Star Trackers

This study presents a series of cost-effect strategies for calibrating star trackers for microsatellite missions. We examine three such strategies that focus on the calibration of the image detector, geometric calibration of the lab setup used for ground testing, and an optical calibration due to lens aberrations. Procedures are developed for each of these strategies that emphasize speed of implementation and accuracy, while trying to minimize manual labour. For the detector calibration, an existing calibration technique was adapted and implemented to reduce fixed pattern noise and dark current. Preliminary results show reduced variations in pixel sensitivity by approximately 21%, averaged across each pixel color given the use of a color imager. Although not substantial, this reduction in pixel variation will help preserve the Gaussian illumination pattern of imaged stars, aiding in correct centroid location. Results pertaining to the lab calibration show accurate star placement, in angular terms to 0.0073º across most of the field of view. This provides an accurate low-cost, variable solution for characterizing sensor performance; specifically pattern matching techniques. Finally, we present some initial results for lens aberration characterization. Using a Gaussian model of star image shape gives trends consistence with astigmatism and field curvature aberrations. Together, these calibrations represent tools that aim to improve both development and manufacture of modern microsatellite star trackers.

Calibration Techniques For Low-Cost Star Trackers

This study presents a series of cost-effect strategies for calibrating star trackers for microsatellite missions. We examine three such strategies that focus on the calibration of the image detector, geometric calibration of the lab setup used for ground testing, and an optical calibration due to lens aberrations. Procedures are developed for each of these strategies that emphasize speed of implementation and accuracy, while trying to minimize manual labour. For the detector calibration, an existing calibration technique was adapted and implemented to reduce fixed pattern noise and dark current. Preliminary results show reduced variations in pixel sensitivity by approximately 21%, averaged across each pixel color given the use of a color imager. Although not substantial, this reduction in pixel variation will help preserve the Gaussian illumination pattern of imaged stars, aiding in correct centroid location. Results pertaining to the lab calibration show accurate star placement, in angular terms to 0.0073º across most of the field of view. This provides an accurate low-cost, variable solution for characterizing sensor performance; specifically pattern matching techniques. Finally, we present some initial results for lens aberration characterization. Using a Gaussian model of star image shape gives trends consistence with astigmatism and field curvature aberrations. Together, these calibrations represent tools that aim to improve both development and manufacture of modern microsatellite star trackers.

Image resizing by reconstruction from deep features

Traditional image resizing methods usually work in pixel space and use various saliency measures. The challenge is to adjust the image shape while trying to preserve important content. In this paper we perform image resizing in feature space using the deep layers of a neural network containing rich important semantic information. We directly adjust the image feature maps, extracted from a pre-trained classification network, and reconstruct the resized image using neural-network based optimization. This novel approach leverages the hierarchical encoding of the network, and in particular, the high-level discriminative power of its deeper layers, that can recognize semantic regions and objects, thereby allowing maintenance of their aspect ratios. Our use of reconstruction from deep features results in less noticeable artifacts than use of imagespace resizing operators. We evaluate our method on benchmarks, compare it to alternative approaches, and demonstrate its strengths on challenging images.

Two-step skeletization of binary images based on the Zhang-Suen model and the producing mask

The aim of the work is to limit excessive thinning and increase the resistance to contour noise of skeletons resulted from arbitrary binary image shape while maintaining a high skeletonization rate. The skeleton is a set of thin lines, the relative position, the size and shape, which conveys information of size, shape and orientation in space of the corresponding homogeneous region of the image. To ensure resistance to contour noise, skeletonization algorithms are built on the basis of several steps. Zhang-Suen algorithm is widely known by high-quality skeletons and average performance, which disadvantages are the blurring of diagonal lines with a thickness of 2 pixels and the totally disappear patterns of 2x2 pixels. To overcome them, a mathematical model that compensates the Zhang-Suen algorithm has proposed in this paper, along with a producing mask and two logical conditions for evaluating its elements.