We conduct an exhaustive survey of image thresholding methods, categorize them, express their formulas under a uniform notation, and finally carry their performance comparison. The thresholding methods are categorized according to the information they are exploiting, such as histogram shape, measurement space clustering, entropy, object attributes, spatial correlation, and local gray-level surface. 40 selected thresholding methods from various categories are compared in the context of nondestructive testing applications as well as for document images. The comparison is based on the combined performance measures. We identify the thresholding algorithms that perform uniformly better over nonde- structive testing and document image applications. © 2004 SPIE and IS&T. (DOI: 10.1117/1.1631316)

/pdf/survey-over-image-thresholding-techniques-and-quantitative-2ekadyqjhj.pdf

Survey over image thresholding techniques and quantitative performance evaluation

A survey on image segmentation

A survey on evaluation methods for image segmentation

http://www.bioimageanalysis.org/wp/wp-content/uploads/formidable/Survey_unsupervisedEvaluation.pdf

Image segmentation evaluation: A survey of unsupervised methods

Intuitionistic fuzzy information - Applications to pattern recognition

Error measures for scene segmentation

A method and apparatus is disclosed for use in performing automated classification of cells of a pap smear as well as other types of cells, such as red blood cells and blood cell neutrophiles. The apparatus generates an image that is split into two paths, (1) a high resolution image of primarily the nucleus of a cell, and (2) a low resolution image of the total cell, nucleus and cytoplasm. An image scanning device located in the first path provides output signals representing the image and circuitry performs measurement and analysis relating to texture of the nucleus. The image in the other path is demagnified which provides an enlarged, lower resolution image of the entire cell, including the nucleus and the cytoplasm and it is projected to a second image scanning device, which provides electrical signals that represent the image of the larger area and circuitry performs measurement and analysis relating to the size, density and color of the cytoplasm and the nucleus. The analysis obtained from each of the two circuits are applied to classification logic circuitry for the purpose of determining malignant cells.

Dual resolution method and apparatus for use in automated classification of pap smear and other samples

A method and apparatus for automatically classifying abnormal and normal red blood cells is disclosed. The abnormal cells may be differentiated into mutually exclusive subpopulations. The individual hemoglobin characteristic for each red blood cell is measured and a hemoglobin parameter may be reported for each subpopulation of cells. Additionally, cell sizes are measured and the mean cell size and Wintrobe indices may be reported. The interior cell structures for red blood cells are analyzed and the cells are classified based on the presence or lack of a central pallor, their shape, size, and hemoglobin content.

Method of and an apparatus for automatic classification of red blood cells

Appareil et procede de production de signaux representant le volume moyen des globules rouges d'un echantillon sanguin. Apparatus and signal producing METHOD representing the mean volume of red blood cells of a blood sample. L'appareil comprend des moyens (310) de production de signaux representant la surface des cellules et des moyens (318) de mesure de la densite optique des cellules isolees et de production de signaux representant la teneur en hemoglobine ou la masse des cellules. The apparatus includes means (310) for producing signals representing the surface of cells and means (318) for measuring the optical density of the signal and producing isolated cells representing the hemoglobin concentration or mass of cells. La paleur centrale est mesuree pour les cellules presentant un phenomene de paleur centrale et des moyens (264) produisent un signal representant la paleur centrale de ces cellules. The central pallor is measured for cells presenting a phenomenon of central pallor and means (264) produce a signal representing the central pallor of these cells. Les signaux representant la surface, les signaux representant la teneur en hemoglobine et ceux representant la paleur centrale sont envoyes a des moyens (403, 415) produisant un resultat (430) representant le volume moyen des globules. The signals representing the surface, the signals representing the content of hemoglobin and those representing the central pallor are sent to means (403, 415) producing a result (430) representing the average volume of blood cells.

Method and apparatus for measuring mean cell volume of red blood cells

A method and apparatus are disclosed for measuring characteristics of cells, such as red blood cells, and for analyzing parameters of the cell characteristics to define a patient's blood. These parameters may be compared for resemblance to predetermined reference characteristic values for a blood cell pathological condition such as a specific kind of anemia or for a normal blood. A report may be generated showing such resemblance to an anemia or to a normal blood. A report may be generated showing parameters of a multivariate dispersion of distribution for a subpopulation of biconcave cells, an indication of skewness of the distribution of the cells with regard to shape variations in central pallor size; the proportion of abnormal kinds of cells found, and closeness of blood to several specific anemias. To expedite the system, a plurality of microprocessors are employed with one microprocessor controlling the imaging means and the summing of measured characteristics while one or more additional microprocessors are measureing characteristics and analyzing the digitized image signals. An improved method of measuring central pallor size is provided.

James W. Bacus

Papers

Error measures for scene segmentation

Dual resolution method and apparatus for use in automated classification of pap smear and other samples

Method of and an apparatus for automatic classification of red blood cells

Method and apparatus for measuring mean cell volume of red blood cells

Automated method and apparatus for classification of cells with application to the diagnosis of anemia