Fuzzy Connectedness Image Segmentation in Graph Cut Formulation: A Linear-Time Algorithm and a Comparative Analysis

A deep theoretical analysis of the graph cut image segmentation framework presented in this paper simultaneously translates into important contributions in several directions. The most important practical contribution of this work is a full theoretical description, and implementation, of a novel powerful segmentation algorithm, GC(max). The output of GC(max) coincides with a version of a segmentation algorithm known as Iterative Relative Fuzzy Connectedness, IRFC. However, GC(max) is considerably faster than the classic IRFC algorithm, which we prove theoretically and show experimentally. Specifically, we prove that, in the worst case scenario, the GC(max) algorithm runs in linear time with respect to the variable M=|C|+|Z|, where |C| is the image scene size and |Z| is the size of the allowable range, Z, of the associated weight/affinity function. For most implementations, Z is identical to the set of allowable image intensity values, and its size can be treated as small with respect to |C|, meaning that O(M)=O(|C|). In such a situation, GC(max) runs in linear time with respect to the image size |C|. We show that the output of GC(max) constitutes a solution of a graph cut energy minimization problem, in which the energy is defined as the a"" (a) norm ayenF (P) ayen(a) of the map F (P) that associates, with every element e from the boundary of an object P, its weight w(e). This formulation brings IRFC algorithms to the realm of the graph cut energy minimizers, with energy functions ayenF (P) ayen (q) for qa[1,a]. Of these, the best known minimization problem is for the energy ayenF (P) ayen(1), which is solved by the classic min-cut/max-flow algorithm, referred to often as the Graph Cut algorithm. We notice that a minimization problem for ayenF (P) ayen (q) , qa[1,a), is identical to that for ayenF (P) ayen(1), when the original weight function w is replaced by w (q) . Thus, any algorithm GC(sum) solving the ayenF (P) ayen(1) minimization problem, solves also one for ayenF (P) ayen (q) with qa[1,a), so just two algorithms, GC(sum) and GC(max), are enough to solve all ayenF (P) ayen (q) -minimization problems. We also show that, for any fixed weight assignment, the solutions of the ayenF (P) ayen (q) -minimization problems converge to a solution of the ayenF (P) ayen(a)-minimization problem (ayenF (P) ayen(a)=lim (q -> a)ayenF (P) ayen (q) is not enough to deduce that). An experimental comparison of the performance of GC(max) and GC(sum) algorithms is included. This concentrates on comparing the actual (as opposed to provable worst scenario) algorithms' running time, as well as the influence of the choice of the seeds on the output.

Real-Time Multimodal Retinal Image Registration for a Computer Assisted Laser Photocoagulation System

An algorithm for the real-time registration of a retinal video sequence captured with a scanning digital ophthalmoscope (SDO) to a retinal composite image is presented. This method is designed for a computer-assisted retinal laser photocoagulation system to compensate for retinal motion and hence enhance the accuracy, speed, and patient safety of retinal laser treatments. The procedure combines intensity and feature-based registration techniques. For the registration of an individual frame, the translational frame-to-frame motion between preceding and current frame is detected by normalized cross correlation. Next, vessel points on the current video frame are identified and an initial transformation estimate is constructed from the calculated translation vector and the quadratic registration matrix of the previous frame. The vessel points are then iteratively matched to the segmented vessel centerline of the composite image to refine the initial transformation and register the video frame to the composite image. Criteria for image quality and algorithm convergence are introduced, which assess the exclusion of single frames from the registration process and enable a loss of tracking signal if necessary. The algorithm was successfully applied to ten different video sequences recorded from patients. It revealed an average accuracy of 2.47 ± 2.0 pixels (∼23.2 ± 18.8 μm) for 2764 evaluated video frames and demonstrated that it meets the clinical requirements.

Inter-observer agreement for spectral- and time-domain optical coherence tomography image grading: a prospective study

The purpose of this study was to compare inter-observer agreement of Stratus™ OCT versus Spectralis™ OCT image grading in patients with neovascular age-related macular degeneration (AMD). Thirty eyes with neovascular AMD were examined with Stratus™ OCT and Spectralis™ OCT. Four different scan protocols were used for imaging. Three observers graded the images for the presence of various pathologies. Inter-observer agreement between OCT models was assessed by calculating intra-class correlation coefficients (ICC). In Stratus™ OCT highest interobserver agreement was found for subretinal fluid (ICC: 0.79), and in Spectralis™ OCT for intraretinal cysts (IRC) (ICC: 0.93). Spectralis™ OCT showed superior interobserver agreement for IRC and epiretinal membranes (ERM) (ICC(Stratus™): for IRC 0.61; for ERM 0.56; ICC(Spectralis™): for IRC 0.93; for ERM 0.84). Increased image resolution of Spectralis™ OCT did improve the inter-observer agreement for grading intraretinal cysts and epiretinal membranes but not for other retinal changes.

A Dynamically Adaptable Image Processing Application Trading Off Between High Performance, Consumption and Dependability in Real Time

As embedded systems evolve, problems inherent to technology become important limitations. In less than ten years, chips will exceed the maximum allowed power consumption affecting performance, since, even though the resources available per chip are increasing, frequency of operation has stalled. Besides, as the level of integration is increased, it is difficult to keep defect density under control, so new fault tolerant techniques are required. In this demo work, a new dynamically adaptable virtual architecture (ARTICo3) to allow dynamic and context-aware use of resources is implemented in a high performance Wireless Sensor node (HiReCookie) to perform an image processing application.

Paris, France, ca. 4th Century (Time Series Map 1 of 8) (Raster Image)

This layer is a georeferenced raster image from the historic paper map series entitled: Lutece ... plan de la ville de Paris ..., par M.L.C.D.L.M. ; A. Coquart, delineavit et sculp. It was published by Jean & Pierre Cot in 1705. Scale [ca. 1:10,000]. This image is of map 1 entitled: Lutece, ou, premier plan de la ville de Paris: tiré de Cesar, de Strabon, de l'empereur Iulien, et d'Ammian Marcellin. The map represents Paris ca. 4th Century. Map in French.The image inside the map neatline is georeferenced to the surface of the earth and fit to the European Datum 1950, Universal Transverse Mercator (UTM) Zone 31N projected coordinate system. All map collar and inset information is also available as part of the raster image, including any inset maps, profiles, statistical tables, directories, text, illustrations, index maps, legends, or other information associated with the principal map. This map shows features such as drainage, towns and villages, roads, selected buildings, ground cover, and more. Relief shown by hachures. This layer is part of a selection of digitally scanned and georeferenced historic maps from the Harvard Map Collection. These maps typically portray both natural and manmade features. The selection represents a range of originators, ground condition dates, scales, and map purposes.

Paris, France, 5th-10th Century (Time Series Map 2 of 8) (Raster Image)

This layer is a georeferenced raster image from the historic paper map series entitled: Lutece ... plan de la ville de Paris ..., par M.L.C.D.L.M. ; A. Coquart, delineavit et sculp. It was published by Jean & Pierre Cot in 1705. Scale [ca. 1:10,000]. This image is of map 2 entitled: Lutece conquise par les François sur les Romains, ou, second plan de la ville de Paris: tiré du misopogone de l'empereur Iulien, d'Amian Marcellin, de Gregoire de Tours, de Boece, de l'abbé Suger, des ancients titres, du tresor des chartres, des archives de St. Germain, de St. Magloire, de St. Eloy, du temple et des vestiges de cette ancienne enceinte qui subsiste encore au jourd'huy ou l'on a vû de nos jours. The map represents Paris, 5th to 10th Centuries. Map in French.The image inside the map neatline is georeferenced to the surface of the earth and fit to the European Datum 1950, Universal Transverse Mercator (UTM) Zone 31N projected coordinate system. All map collar and inset information is also available as part of the raster image, including any inset maps, profiles, statistical tables, directories, text, illustrations, index maps, legends, or other information associated with the principal map. This map shows features such as drainage, towns and villages, roads, built-up areas and selected buildings, fortification, ground cover, and more. Relief shown by hachures.This layer is part of a selection of digitally scanned and georeferenced historic maps from the Harvard Map Collection. These maps typically portray both natural and manmade features. The selection represents a range of originators, ground condition dates, scales, and map purposes.

Paris, France, 1137-1180 (Time Series Map 3 of 8) (Raster Image)

This layer is a georeferenced raster image from the historic paper map series entitled: Lutece ... plan de la ville de Paris ..., par M.L.C.D.L.M. ; A. Coquart, delineavit et sculp. It was published by Jean & Pierre Cot in 1705. Scale [ca. 1:10,000]. This image is of map 3 entitled: Troisieme plan de la ville de Paris son êtendiie et les bourgs dont elle êtoit environnée sous le regne de Louis le Jeune VIIe du nom: tiré des descriptions de Fortunat, de Gregoire de Tours, Dabbon, et D'Aimoin, et des anciens titres qui sont dans les Tresor des chartres de France, les registres de la Chambre, des Comptes, les Banieres du Châtelet de Paris, les Archives de l'Archevêché celles de Saint Germain, de Ste. Genevieve, de St. Martin, de S. Eloy, et du Temple. The map represents Paris, 1137 to 1180. Map in French.The image inside the map neatline is georeferenced to the surface of the earth and fit to the European Datum 1950, Universal Transverse Mercator (UTM) Zone 31N projected coordinate system. All map collar and inset information is also available as part of the raster image, including any inset maps, profiles, statistical tables, directories, text, illustrations, index maps, legends, or other information associated with the principal map. This map shows features such as drainage, towns and villages, roads, built-up areas and selected buildings, fortification, ground cover, and more. Relief shown by hachures.This layer is part of a selection of digitally scanned and georeferenced historic maps from the Harvard Map Collection. These maps typically portray both natural and manmade features. The selection represents a range of originators, ground condition dates, scales, and map purposes.

Paris, France, 1180-1223 (Time Series Map 4 of 8) (Raster Image)

This layer is a georeferenced raster image from the historic paper map series entitled: Lutece ... plan de la ville de Paris ..., par M.L.C.D.L.M. ; A. Coquart, delineavit et sculp. It was published by Jean & Pierre Cot in 1705. Scale [ca. 1:10,000]. This image is of map 4 entitled: Quatriême plan de la ville de Paris son accroissement, et l'etat ou elle êtoit sous le regne de Philippe Auguste qui mourut l'an 1223 apres avoir regné 43 ans: tiré de Rigord, de Knobelsderf, de Rodolphe Boterée, de Raoul, de Praesles, de Paul Merula, de Guaquin, de Pithou, de Papire Masson, de Corrozet, de Dubreüil, de Duchesne, des memoriaux, et autres anciens registres de la Chambre, des Comptes, et des archives de l'Archevêché du Chaptire de N. Dame, et des anciennes abbayes. The map represents Paris, 1180 to 1223. Map in French.The image inside the map neatline is georeferenced to the surface of the earth and fit to the European Datum 1950, Universal Transverse Mercator (UTM) Zone 31N projected coordinate system. All map collar and inset information is also available as part of the raster image, including any inset maps, profiles, statistical tables, directories, text, illustrations, index maps, legends, or other information associated with the principal map. This map shows features such as drainage, towns and villages, roads, built-up areas and selected buildings, fortification, ground cover, and more. Relief shown by hachures. Includes index.This layer is part of a selection of digitally scanned and georeferenced historic maps from the Harvard Map Collection. These maps typically portray both natural and manmade features. The selection represents a range of originators, ground condition dates, scales, and map purposes.

Paris, France, 1367-1383 (Time Series Map 5 of 8) (Raster Image)

This layer is a georeferenced raster image from the historic paper map series entitled: Lutece ... plan de la ville de Paris ..., par M.L.C.D.L.M. ; A. Coquart, delineavit et sculp. It was published by Jean & Pierre Cot in 1705. Scale [ca. 1:10,000]. This image is of map 5 entitled: Cinquiême plan de la ville de Paris son accroissement et sa quatriême clôture commancée sous Charles V. l'an 1367 et finie sous Charles VI. l'an 1383: tiré des devis et marchez faits avec le ouvriers, des procez verbaux de toisez et receptions des ouvrages des comptes rendus par ceux qui en eurent la conduite, de la chronique m.s. de St. Denis et d'autres titres et manuscrits qui sont conservez en la Chambre des Comptes et dans Biblioteques. The map represents Paris, 1367 to 1383. Map in French.The image inside the map neatline is georeferenced to the surface of the earth and fit to the European Datum 1950, Universal Transverse Mercator (UTM) Zone 31N projected coordinate system. All map collar and inset information is also available as part of the raster image, including any inset maps, profiles, statistical tables, directories, text, illustrations, index maps, legends, or other information associated with the principal map. This map shows features such as drainage, towns and villages, roads, built-up areas and selected buildings, fortification, ground cover, and more. Relief shown by hachures. Includes index, text, and notes.This layer is part of a selection of digitally scanned and georeferenced historic maps from the Harvard Map Collection. These maps typically portray both natural and manmade features. The selection represents a range of originators, ground condition dates, scales, and map purposes.

Paris, France, 1589-1643 (Time Series Map 7 of 8) (Raster Image)

This layer is a georeferenced raster image from the historic paper map series entitled: Lutece ... plan de la ville de Paris ..., par M.L.C.D.L.M. ; A. Coquart, delineavit et sculp. It was published by Jean & Pierre Cot in 1705. Scale [ca. 1:10,000]. This image is of map 7 entitled: Septiême plan de la ville de Paris son acroissement [sic] et ses embelissemens sous Henry III. et Louis XIII. depuis 1589 jusqu'en 1643: tiré des lettres patentes ou arrest du conseil qui ont ordonné les ouvrages des devis et marchez faits avec les entrepreneurs et levé sur les lieux ou ils ont été construits, et ou la plus grand partie subsistent encore. The map represents Paris, 1589 to 1643. Map in French.The image inside the map neatline is georeferenced to the surface of the earth and fit to the European Datum 1950, Universal Transverse Mercator (UTM) Zone 31N projected coordinate system. All map collar and inset information is also available as part of the raster image, including any inset maps, profiles, statistical tables, directories, text, illustrations, index maps, legends, or other information associated with the principal map. This map shows features such as drainage, towns and villages, roads, built-up areas and selected buildings, fortification, ground cover, and more. Relief shown by hachures. Includes index, text, and notes.This layer is part of a selection of digitally scanned and georeferenced historic maps from the Harvard Map Collection. These maps typically portray both natural and manmade features. The selection represents a range of originators, ground condition dates, scales, and map purposes.

Paris, France, 1705 (Time Series Map 8 of 8) (Raster Image)

This layer is a georeferenced raster image from the historic paper map series entitled: Lutece ... plan de la ville de Paris ..., par M.L.C.D.L.M. ; A. Coquart, delineavit et sculp. It was published by Jean & Pierre Cot in 1705. Scale [ca. 1:10,000]. This image is of map 8 entitled: Huitieme plan plan de Paris divisé en ses vingts quartiers. The map represents Paris, 1705. Map in French.The image inside the map neatline is georeferenced to the surface of the earth and fit to the European Datum 1950, Universal Transverse Mercator (UTM) Zone 31N projected coordinate system. All map collar and inset information is also available as part of the raster image, including any inset maps, profiles, statistical tables, directories, text, illustrations, index maps, legends, or other information associated with the principal map. This map shows features such as drainage, towns and villages, roads, built-up areas and selected buildings, fortification, ground cover, and more. Relief shown by hachures. Includes index, text, and notes.This layer is part of a selection of digitally scanned and georeferenced historic maps from the Harvard Map Collection. These maps typically portray both natural and manmade features. The selection represents a range of originators, ground condition dates, scales, and map purposes.

Paris, France, 1422-1589 (Time Series Map 6 of 8) (Raster Image)

This layer is a georeferenced raster image from the historic paper map series entitled: Lutece ... plan de la ville de Paris ..., par M.L.C.D.L.M. ; A. Coquart, delineavit et sculp. It was published by Jean & Pierre Cot in 1705. Scale [ca. 1:10,000]. This image is of map 6 entitled: Sixiême plan de la ville de Paris et ses accroissements depuis le commencement du régne de Charles VII. l'an 1422 jusqu'a la fin du régne d'Henry III. l'an 1589: tiré des lettres patentes qui ont ordonné les ouvrages, des contrats passez avec les entrepreneurs, des registres de la chambre des comptes de l'histoire et des memoires du temps. The map represents Paris, 1422 to 1589. Map in French.The image inside the map neatline is georeferenced to the surface of the earth and fit to the European Datum 1950, Universal Transverse Mercator (UTM) Zone 31N projected coordinate system. All map collar and inset information is also available as part of the raster image, including any inset maps, profiles, statistical tables, directories, text, illustrations, index maps, legends, or other information associated with the principal map. This map shows features such as drainage, towns and villages, roads, built-up areas and selected buildings, fortification, ground cover, and more. Relief shown by hachures. Includes index, text, and notes.This layer is part of a selection of digitally scanned and georeferenced historic maps from the Harvard Map Collection. These maps typically portray both natural and manmade features. The selection represents a range of originators, ground condition dates, scales, and map purposes.

Image enhancement of real-time television to benefit the visually impaired

Purpose: To examine the use of real-time, generic edge detection, image processing techniques to enhance the television viewing of the visually impaired. Design: Prospective, clinical experimental study. Method: One hundred and two sequential visually impaired (average age 73.8 ± 14.8 years; 59% female) in a single center optimized a dynamic television image with respect to edge detection filter (Prewitt, Sobel, or the two combined), color (red, green, blue, or white), and intensity (one to 15 times) of the overlaid edges. They then rated the original television footage compared with a black-and-white image displaying the edges detected and the original television image with the detected edges overlaid in the chosen color and at the intensity selected. Footage of news, an advertisement, and the end of program credits were subjectively assessed in a random order. Results: A Prewitt filter was preferred (44%) compared with the Sobel filter (27%) or a combination of the two (28%). Green and white were equally popular for displaying the detected edges (32%), with blue (22%) and red (14%) less so. The average preferred edge intensity was 3.5 ± 1.7 times. The image-enhanced television was significantly preferred to the original (P < .001), which in turn was preferred to viewing the detected edges alone (P < .001) for each of the footage clips. Preference was not dependent on the condition causing visual impairment. Seventy percent were definitely willing to buy a set-top box that could achieve these effects for a reasonable price. Conclusions: Simple generic edge detection image enhancement options can be performed on television in real-time and significantly enhance the viewing of the visually impaired. © 2007 Elsevier Inc. All rights reserved.