This book-length meditation on the Hebrew alphabet offers profound insights into many important ideas found in Jewish thought. From time immemorial, the Hebrew alphabet has been considered to be more than a collection of individual letters. Indeed, the essence of each letter of the Hebrew alphabet can be seen as a fundamental building block of the world. Jewish scholars throughout the ages have meditated on these letters, deriving spiritual inspiration in the process. In The Inner Meaning of the Hebrew Letters, Robert M. Haralick looks closely at each of the Hebrew characters, helping us to gain insight from this remarkable tradition. Drawing primarily upon traditional kabbalistic and chasidic thought, Haralick combines his own insights with those of great Jewish personalities such as Moshe Chayim Luzzatto and Rabbi Nachman of Bratslav, as well as drawing upon classical texts, including the Bahir, the Zohar, the Midrash, and the Talmud. One of Haralick's main sources of inspiration is the ancient Jewish art of gematria, where each letter has a numerical value as does each combination of letters. Through this traditional methodology, Haralick shows his readers the many, often dazzling, ways that the Hebrew alphabet has been examined.

This volume is the collection of lectures and presentations of the NATO AS! On Pictorial Data Analysis, held August 1-12, 1982 in the beautiful chateau de Bonas, Bonas France. The director of the AS! was Robert M. Haralick and the Co-director was Stefano Levialdi. The papers in the book are arranged in two sections first theory and general prinicples and then applications. Local computations play a central role in image processing both when a traditional computer is used and when parallel machines are used for improving image throughput. Levialdi reviews such neighborhood operators. Hung and Kasvand discuss a line thinning application which involves detection of critical points on chain encoded data. Most low level image processing has been done using the digital raster as the basic data structure. Within the last few years many of these basic algorithms have been developed for the quadtree data structure. The quadtree permits easier access to certain kinds of spatial adjacency relationships in a variable resolution context. Rosenfeld reviews the properties of these representations and their uses in image segmentation and property measurement. Besslich discusses an expanded form of an invertible quadtree representation which permits a multiprocessor execution. Gisolfi and Vitulano discuss the C-matrix and C-filtering technique for image and texture feature extraction. O'mara et.al. discuss the application of Codel numbers to image feature extraction. Kropatsch discusses an image segmentation technique which permits the effective use of a variety of different kinds of segmentation techniques.