Page 1. High-resolution beamforming for multibeam echo sounders using raw EM3000 data Are Rmhovde... more Page 1. High-resolution beamforming for multibeam echo sounders using raw EM3000 data Are Rmhovde LurenYang Department of Informatics Srn-ECY University of Oslo P. 0. Box 1080, Blindem N-03 16 Oslo, Norway P. 0. Box 124, Blindem N-03 14 Oslo, Norway ...
Abstract. Evaluation isanimportant stepindeveloping a segmentation algorithm for an image analysi... more Abstract. Evaluation isanimportant stepindeveloping a segmentation algorithm for an image analysis system.We rstgiveareviewofsegmentation evaluation methods, and then demonstrate howasupervised evaluation method based onshape features isusedinthedevelopment of asegmentation algorithm for uorescence imagesofwhite blood cells. 1
Green's theoremevaluatesadoubleintegralover the region ofanobject byasimpleintegration along... more Green's theoremevaluatesadoubleintegralover the region ofanobject byasimpleintegration along the boundary of the object. Ithasbeenusedinmoment computation sincetheshapeofabinary object istotally determined byitsboundary. By usingadiscrete analogue of Green's theorem, we present a new algorithm for fast computationofgeometric moments. The algorithm is faster than previous methods, and givesexact results. Theimportance ofexact computationisdiscussed byexamining the invariance ofHu's moments. A fastmethod forcomputingmomentsof regions in greylevelimage,usingdiscrete Green'stheorem,isalsopresented. 1
A new texture analysis approach is applied to the problem of classification of pathological state... more A new texture analysis approach is applied to the problem of classification of pathological states from electron microscopy images of mouse liver cell nuclei. For each pixel in the image, a region of consistent connected neighbouring pixels is extracted, forming a local texel of pixels belonging to the same gray level population. The geometric properties of each texel is described by invariant moment-based features. A recently developed method for fast and exact computation of Cartesian geometric moments is utilized. Each cell nucleus is characterized by a feature vector, giving the average feature values of both the bright and the dark structures. A leave-one-out classification is performed, using 4 dioeerent classes of cells (normal, proliferating, precancer and cancer). The results demonstrate that using gray level connected neighbour structures as texels gives us important information about the chromatin texture of the cell nuclei and thus the pathological state of the cell. Sev...
When artifical neural networks (ANN) are used in the prediction problems, it is usually desirable... more When artifical neural networks (ANN) are used in the prediction problems, it is usually desirable that some form of confidence bound is placed on the predicted value. Methods to estimate the confidence bound are available. However, these methods are valid under certain assumptions, which are rarely satisfied in practice. The behavior of the estimated confidence bound are not well understood when the assumptions are violated. We have designed some test functions to examine the behavior, and suggest how the estimated confidence bound can be corrected. The suggested method is used in the prediction of rock porosity values from seismic data for oil reservoir characterisation.
The area and the perimeter of a planar object are two useful features to describe the shape of th... more The area and the perimeter of a planar object are two useful features to describe the shape of the object, and again the motion of it. This paper deals with the estimation of the two features from a discrete binary image. The area can often be accurately estimated by counting the number of pixels inside the object. However, the estimation of the perimeter is a problem, since many possible contours, all having different lengths, correspond to a specific discrete realzzation. Thus, to develop a practical length estimator, some reasonable assumption about the original contour should be made. We assume that the boundary of a blob-like object consists of chains of circular arcs, and therefore evaluate the precision of several area and length estimators applied to circles. The problem of efficient computatron is also discussed.
The area and the perimeter of a planar object are two useful features to describe the shape of th... more The area and the perimeter of a planar object are two useful features to describe the shape of the object, and again the motion of it. This paper deals with the estimation of the two features from a discrete binary image. The area can often be accurately estimated by counting the number of pixels inside the object. However, the estimation of the perimeter is a problem, since many possible contours, all having different lengths, correspond to a specific discrete realzzation. Thus, t o develop a practical length estimator, some reasonable assumption about the original contour should be made. W e assume that the boundary of a blob-like object consists of chains of circular arcs, and therefore evaluate the precision of several area and length estimators applied to circles. The problem of efficient computatron is also discussed.
Chapter 5 provides a r e v i e w o f e d ge-based s e gmentation methods a n d t h e e xperimenta... more Chapter 5 provides a r e v i e w o f e d ge-based s e gmentation methods a n d t h e e xperimental results o f s o m e o f t he methods f o r t h e u o r e s c e nce c e l l i m a ges. Chapter 6 describes t h r e e t y pes o f g r a y l e v e l t h r e sholding t e c h n i q u e s : g l obal, local and d y namic t h r e sholding. A new d y n a m i c t h r e s h o l d i n g m e t h od is p r e s e n t e d. Chapter 7 presents a t w o-pass s patial-temporal i m age s e gmentation s y s t e m f o r t he uorescence c e l l i m a ge sequences. T e c h n i q u e s d i s c u s s e d i n C h apter 5 a n d 6 a r e used i n t h e r s t p a s s. C e l l o b j e c t t r a c k i n g , r e g i o n c l o s i n g a n d b o u ndary s m oothing are d o n e i n t he second p ass. C h apter 4 , 5 , 6 a n d 7 l e ad to a c omplete s e gmentation system, w h i c h i s e v a l u a t e d i n C h a p t e r 1 1. Chapter 8 contains a l i s t o f s h a pe features. C e l l m o t i on description b ased o n t he features i s d i scussed.
... Bottom echoes of near-normal incident angles are usu-ally detected based on the amplitude of ... more ... Bottom echoes of near-normal incident angles are usu-ally detected based on the amplitude of the received signal. ... who helped in various ways, with spe-cial thanks to Freddy P~hner, Erik Hammerstad, Ole B. Gammelseter, Even B. Lunde, Kjell Nilsen, and Inger H. Steensen. ...
... The geometry of the transducers and the beam pat-terns of a multibeam bathymetric sonar syste... more ... The geometry of the transducers and the beam pat-terns of a multibeam bathymetric sonar system is shown in Fig. 1. The transducers send narrow band acoustic signals towards the sea bottom, and receive echoes for bathymetry computation. ...
Selected Papers from the 9th Scandinavian Conference on Image Analysis, 1995
We present an image analysis system developed to measure the motion of white blood cells from a t... more We present an image analysis system developed to measure the motion of white blood cells from a temporal sequence of uorescence microscopy images. A twopass spatio-temporal segmentation system is used. Pixels are classi ed as cell and background pixels by an initial segmentation in the rst pass. Region labeling, correction and cell tracking are done in the second pass. After segmentation, shape features are estimated from binary discrete regions, and cell motion is then measured by using shape features. A supervised method based on shape features is used to evaluate the results of the segmentation.
. Evaluation is an important step in developing a segmentationalgorithm for an image analysis sys... more . Evaluation is an important step in developing a segmentationalgorithm for an image analysis system. We ?rst give a review of segmentationevaluation methods, and then demonstrate how a supervisedevaluation method based on shape features is used in the developmentof a segmentation algorithm for AEuorescence images of white blood cells.1 IntroductionAn image analysis system typically has three major components: image segmentation,feature
Page 1. High-resolution beamforming for multibeam echo sounders using raw EM3000 data Are Rmhovde... more Page 1. High-resolution beamforming for multibeam echo sounders using raw EM3000 data Are Rmhovde LurenYang Department of Informatics Srn-ECY University of Oslo P. 0. Box 1080, Blindem N-03 16 Oslo, Norway P. 0. Box 124, Blindem N-03 14 Oslo, Norway ...
Abstract. Evaluation isanimportant stepindeveloping a segmentation algorithm for an image analysi... more Abstract. Evaluation isanimportant stepindeveloping a segmentation algorithm for an image analysis system.We rstgiveareviewofsegmentation evaluation methods, and then demonstrate howasupervised evaluation method based onshape features isusedinthedevelopment of asegmentation algorithm for uorescence imagesofwhite blood cells. 1
Green's theoremevaluatesadoubleintegralover the region ofanobject byasimpleintegration along... more Green's theoremevaluatesadoubleintegralover the region ofanobject byasimpleintegration along the boundary of the object. Ithasbeenusedinmoment computation sincetheshapeofabinary object istotally determined byitsboundary. By usingadiscrete analogue of Green's theorem, we present a new algorithm for fast computationofgeometric moments. The algorithm is faster than previous methods, and givesexact results. Theimportance ofexact computationisdiscussed byexamining the invariance ofHu's moments. A fastmethod forcomputingmomentsof regions in greylevelimage,usingdiscrete Green'stheorem,isalsopresented. 1
A new texture analysis approach is applied to the problem of classification of pathological state... more A new texture analysis approach is applied to the problem of classification of pathological states from electron microscopy images of mouse liver cell nuclei. For each pixel in the image, a region of consistent connected neighbouring pixels is extracted, forming a local texel of pixels belonging to the same gray level population. The geometric properties of each texel is described by invariant moment-based features. A recently developed method for fast and exact computation of Cartesian geometric moments is utilized. Each cell nucleus is characterized by a feature vector, giving the average feature values of both the bright and the dark structures. A leave-one-out classification is performed, using 4 dioeerent classes of cells (normal, proliferating, precancer and cancer). The results demonstrate that using gray level connected neighbour structures as texels gives us important information about the chromatin texture of the cell nuclei and thus the pathological state of the cell. Sev...
When artifical neural networks (ANN) are used in the prediction problems, it is usually desirable... more When artifical neural networks (ANN) are used in the prediction problems, it is usually desirable that some form of confidence bound is placed on the predicted value. Methods to estimate the confidence bound are available. However, these methods are valid under certain assumptions, which are rarely satisfied in practice. The behavior of the estimated confidence bound are not well understood when the assumptions are violated. We have designed some test functions to examine the behavior, and suggest how the estimated confidence bound can be corrected. The suggested method is used in the prediction of rock porosity values from seismic data for oil reservoir characterisation.
The area and the perimeter of a planar object are two useful features to describe the shape of th... more The area and the perimeter of a planar object are two useful features to describe the shape of the object, and again the motion of it. This paper deals with the estimation of the two features from a discrete binary image. The area can often be accurately estimated by counting the number of pixels inside the object. However, the estimation of the perimeter is a problem, since many possible contours, all having different lengths, correspond to a specific discrete realzzation. Thus, to develop a practical length estimator, some reasonable assumption about the original contour should be made. We assume that the boundary of a blob-like object consists of chains of circular arcs, and therefore evaluate the precision of several area and length estimators applied to circles. The problem of efficient computatron is also discussed.
The area and the perimeter of a planar object are two useful features to describe the shape of th... more The area and the perimeter of a planar object are two useful features to describe the shape of the object, and again the motion of it. This paper deals with the estimation of the two features from a discrete binary image. The area can often be accurately estimated by counting the number of pixels inside the object. However, the estimation of the perimeter is a problem, since many possible contours, all having different lengths, correspond to a specific discrete realzzation. Thus, t o develop a practical length estimator, some reasonable assumption about the original contour should be made. W e assume that the boundary of a blob-like object consists of chains of circular arcs, and therefore evaluate the precision of several area and length estimators applied to circles. The problem of efficient computatron is also discussed.
Chapter 5 provides a r e v i e w o f e d ge-based s e gmentation methods a n d t h e e xperimenta... more Chapter 5 provides a r e v i e w o f e d ge-based s e gmentation methods a n d t h e e xperimental results o f s o m e o f t he methods f o r t h e u o r e s c e nce c e l l i m a ges. Chapter 6 describes t h r e e t y pes o f g r a y l e v e l t h r e sholding t e c h n i q u e s : g l obal, local and d y namic t h r e sholding. A new d y n a m i c t h r e s h o l d i n g m e t h od is p r e s e n t e d. Chapter 7 presents a t w o-pass s patial-temporal i m age s e gmentation s y s t e m f o r t he uorescence c e l l i m a ge sequences. T e c h n i q u e s d i s c u s s e d i n C h apter 5 a n d 6 a r e used i n t h e r s t p a s s. C e l l o b j e c t t r a c k i n g , r e g i o n c l o s i n g a n d b o u ndary s m oothing are d o n e i n t he second p ass. C h apter 4 , 5 , 6 a n d 7 l e ad to a c omplete s e gmentation system, w h i c h i s e v a l u a t e d i n C h a p t e r 1 1. Chapter 8 contains a l i s t o f s h a pe features. C e l l m o t i on description b ased o n t he features i s d i scussed.
... Bottom echoes of near-normal incident angles are usu-ally detected based on the amplitude of ... more ... Bottom echoes of near-normal incident angles are usu-ally detected based on the amplitude of the received signal. ... who helped in various ways, with spe-cial thanks to Freddy P~hner, Erik Hammerstad, Ole B. Gammelseter, Even B. Lunde, Kjell Nilsen, and Inger H. Steensen. ...
... The geometry of the transducers and the beam pat-terns of a multibeam bathymetric sonar syste... more ... The geometry of the transducers and the beam pat-terns of a multibeam bathymetric sonar system is shown in Fig. 1. The transducers send narrow band acoustic signals towards the sea bottom, and receive echoes for bathymetry computation. ...
Selected Papers from the 9th Scandinavian Conference on Image Analysis, 1995
We present an image analysis system developed to measure the motion of white blood cells from a t... more We present an image analysis system developed to measure the motion of white blood cells from a temporal sequence of uorescence microscopy images. A twopass spatio-temporal segmentation system is used. Pixels are classi ed as cell and background pixels by an initial segmentation in the rst pass. Region labeling, correction and cell tracking are done in the second pass. After segmentation, shape features are estimated from binary discrete regions, and cell motion is then measured by using shape features. A supervised method based on shape features is used to evaluate the results of the segmentation.
. Evaluation is an important step in developing a segmentationalgorithm for an image analysis sys... more . Evaluation is an important step in developing a segmentationalgorithm for an image analysis system. We ?rst give a review of segmentationevaluation methods, and then demonstrate how a supervisedevaluation method based on shape features is used in the developmentof a segmentation algorithm for AEuorescence images of white blood cells.1 IntroductionAn image analysis system typically has three major components: image segmentation,feature
Uploads
Papers by Luren Yang