Abstract :
Scale space analysis is an efficient solution to the edge detection of objects in low to high contrast images. However, this approach is time consuming and computationally expensive. The parallel processing properties of a neural network provide an ideal solution to managing the large amounts of data processed in image analysis, however their application to multiscale analysis is still in its infancy. This paper reports on a new approach to detecting 2-dimensional and 3-dimensional objects in low to high contrast images. The novel idea is based on combining neural network arbitration and scale space analysis to automatically select one optimum scale for the entire image at which scale space edge detection can be applied. Thus, introducing new measures to solve many of the problems existing in the discipline of image processing, such as poor edge detection in low contrast images, speed of recognition and high computational cost. This new approach to edge detection is formalised in the Automatic Edge Detection Scheme (AEDS). The AEDS is implemented on a real-life application namely, the detection of bands within low contrast DNA autoradiograph images. Results are presented to show the AEDS overcoming the aforementioned common problems with image processing techniques
Keywords :
DNA; biology computing; edge detection; neural nets; object detection; parallel processing; radioisotope imaging; 2-dimensional object detection; 3-dimensional object detection; DNA bands; automatic edge detection; autoradiograph images; computational cost; image processing techniques; low contrast images; multiscale analysis; neural network; neural network arbitration; parallel processing; recognition speed; scale space analysis; Computer network management; DNA; Image analysis; Image edge detection; Image processing; Image recognition; Neural networks; Object detection; Parallel processing; Velocity measurement;
