Presently, many researchers are working on the variability of cultivation and inputs. In an agricultural parcel, yield levels vary considerably from one place to another. This paper presents an early indicator of the ...
详细信息
Presently, many researchers are working on the variability of cultivation and inputs. In an agricultural parcel, yield levels vary considerably from one place to another. This paper presents an early indicator of the cultivation variability. This indicator is correlated with yield and non-destructive for wheatears. The number of wheatears is chosen as an indicator, and an automatic system for counting wheatears is put into place.
This paper presents the algorithms for recognition and beautification which are used in incremental graphic design applications. These applications propose multimodal interfaces integrating handwriting, gesture, and s...
详细信息
A method is presented for building a hierarchic family of polygonal approximations to image outlines that relates the dominant point spacing to the scale-level better than other methods. The results indicate that the ...
详细信息
A method is presented for building a hierarchic family of polygonal approximations to image outlines that relates the dominant point spacing to the scale-level better than other methods. The results indicate that the influence of image scanning and quantization noise is quickly eliminated from the outline shape, yet the finer level approximations still provide more local details of the outlines than other methods. As the scale is increased, the number of polygon vertices in the outline approximation is reduced more uniformly. While at coarser levels, the global outline shape is captured without describing details and noise.
The application of image analysis techniques to the processing.of hand-drawn geometric figures used as a means of evaluating and assessing visual-motor integration development in young children is the focus of this st...
详细信息
The application of image analysis techniques to the processing.of hand-drawn geometric figures used as a means of evaluating and assessing visual-motor integration development in young children is the focus of this study. Some aspects of a program of work in progress which is seeking to investigate the potential for automated administration and evaluation of testing procedures such as that represented by the VMI test are described. In particular, some important issues concerned with sources of error in applying the VMI test and the potential advantages of automated image analysis in test evaluation are considered.
A three-step algorithm is developed to segment oil spills from a marine background on Synthetic Aperture Radar (SAR) data. First, filtering is performed to reduce speckle noise. Then fuzzy clustering is carried out to...
详细信息
A three-step algorithm is developed to segment oil spills from a marine background on Synthetic Aperture Radar (SAR) data. First, filtering is performed to reduce speckle noise. Then fuzzy clustering is carried out to obtain a preliminary partition of the pixels on the basis of their grey level intensities. A very simple cluster validity criterion is tested to determine the optimal number of clusters present in the data. In order to improve segmentation a final step involves a cluster merging procedure using edge information provided by a Sobel operator. The algorithm has been tested on SEASAT images.
Visual transformations are used to achieve size and rotation invariances in computational approaches to vision. Nature seem to have adapted the same strategy. In particular, research workers studying biological vision...
详细信息
Visual transformations are used to achieve size and rotation invariances in computational approaches to vision. Nature seem to have adapted the same strategy. In particular, research workers studying biological vision systems have known for a decade that a particular spatial transformation called Retino-Cortical (RC) Transform appears to be applied to the visual signals conveyed between the retina and the primary visual cortex in mammals. The RC transform have been recognised for some time, but its potential in machine vision applications has not yet been recognised in practical vision system implementations. This study reports an attempt to combine the invariant properties of the RC transform with a scale-space approach to image representation and image matching to develop an anthromorphic visual recognition architecture.
The study on active contour models (or snakes) carried out as part of the SAMMIE project (AIM project A2032) on the development of advanced segmentation aids for object demarcation in MR and CT images is presented. Ac...
详细信息
The study on active contour models (or snakes) carried out as part of the SAMMIE project (AIM project A2032) on the development of advanced segmentation aids for object demarcation in MR and CT images is presented. Active contour models are a special form of deformable models, characterized by their property of dynamic deformation to an image from an original given shape. This deformation is controlled through the minimization of an energy function.
This paper first analyses the differences between SHD and lower-resolution images in terms of HVS responses to spatial frequencies in images. Coding of SHD images by the Laplacian pyramid and the DCT pyramid using HVS...
详细信息
This paper first analyses the differences between SHD and lower-resolution images in terms of HVS responses to spatial frequencies in images. Coding of SHD images by the Laplacian pyramid and the DCT pyramid using HVS-based quantization is then described, with the test image coded at good quality with 0.5 bpp and perceptually transparent at 1 bpp, viewed at 1H.
The extraction of lines, words and characters from a digital document image are necessary computational steps preceding character recognition. Much has been discussed in character segmentation and recognition but litt...
详细信息
Post-processing.of compressed images is intended to reduce the effect of quantization noise that degrades the picture quality. In this paper, a method for post-processing.of low bit rate block coded images is presente...
详细信息
Post-processing.of compressed images is intended to reduce the effect of quantization noise that degrades the picture quality. In this paper, a method for post-processing.of low bit rate block coded images is presented. In this method, information extracted from locally calculated image gradient is used to identify and separate areas of high and low image detail in the decoded image. The main feature of the proposed method is that a fully adaptive Gaussian-type filter is used for the high image detail regions. Simulation results indicate that the proposed method can significantly improve the picture quality of the compressed images.
暂无评论