Richard Beare

Mosaicing of microscope images with global geometric and radiometric corrections

Image mosaicing has found a number of applications such as panoramic imaging, digital terrain mapping, ophthalmology and virtual microscopy. In this study, we present an automated mosaicing technique for generating virtual slides from microscope images. We carried out robust image feature matching and global geometric and radiometric parameter estimation. The input images were transformed using the estimated geometric and radiometric parameters and mosaiced together, producing accurate registration of overlapping regions without visible seams.

Mosaicing of Microscope Images

Image mosaicing has found a number of applications such as panoramic imaging, digital terrain mapping, ophthalmology, and virtual microscopy. In this paper, we present an automated mosaicing technique for generating mosaics from microscope images. We carry out robust image feature matching and global geometric and radiometric parameter estimation. The images are transformed using the estimated parameters and mosaiced together, producing accurate registration of overlapping regions without visible seams.

Efficient Implementation of Thelocally Constrained Watershed Transform and Seeded Region Growing

The watershed transform and seeded region growing are well known tools for image segmentation. They are members of a class of greedy region growing algorithms that are simple, fast and largely parameter free. The main control over these algorithms come from the selection of the marker image, which defines the number of regions and a starting position for each region.

Stereoscopic flatbed scanner

3-D scanning has become increasingly popular in a wide range of applications. We present a prototype 3-D/stereoscopic scanning system based on a cheap, readily available flatbed scanner. Stereoscopic imaging is achieved by modifying the optical path of the ordinary flatbed scanner. The results of 3-D imaging using our prototype system are demonstrated, and a number of design alternatives are discussed.

Motion Segmentation Using Seeded Region Growing

Motion based segmentation techniques usually begin with an estimate of the velocity flow field and utilise clustering techniques to segment the flow field. This paper investigates the practicality of performing segmentation using the information typically applied to the velocity estimation stage. A region growing process is described that can reliably segment rigid objects observed by moving cameras without requiring shape or motion models.