Speckle pattern applications have grown up beside holography making use of the highly coherent illumination derived from optical lasers. Speckle has not been well understood and in many cases treated as noise and something to suppress as far as possible. To date a few workers have used speckle monitoring to detect surface stability and others have used speckle in spatial filtering to form a displacement sensitive system. A significant property of the speckle pattern, namely its ability to carry wavefront phase information seems to have been missed until Leendertz published work at the I.C.O. Conference in 1969. In contrast to this applications of holographic interferometry have been progressively sought out, developed and used, since the early discoveries in 1964. The hologram has promised great potential in engineering metrology but has been disappointingly slow in finding its way into system design and industrial application. In the author's opinion this may be due to several basic practical difficulties with holography some of which can be overcome using the new speckle pattern methods. To establish this some of the metrological requirements must be considered together with fundamentally different recording practices which can be used for holographic and speckle pattern interferometry. It would be desirable to carry out all data processing rapidly and preferably avoid the uncertainties of the photographic process.

We describe a method of classifying or identifying a three-dimensional object from one or more of its silhouettes. The method is based on a low-cost parallel mechanism for computing the slope density, i.e., the density of slopes of the edge of the silhouette. Examples of applications of the method are microscopy and aerial navigation. In both of these applications the object's outline is often the principal clue to its identity, even though other clues, such as texture, color and interior detail may be available. The present study is based on an application of our technique to recognizing aircraft.

In this paper, a pattern recognition technique through atmospheric turbulence is discussed. We have shown that there exists no unique relationship between the target and its power spectrum. However, for some finite number of distinguishable targets, it may be possible to recognize the targets by means of the Wiener-Khinchin's theorem. The fundamental advantage of this pattern recognition technique is the elimation of the phase distortion due to the atmospheric turbulence. However, this phase elimination also results in a loss of phase information of the target. Nevertheless, if the atmospheric turbulence is strong, this pattern recognition technique by means of Wiener-Khinchin's theorem may be more advantageous than the direct imaging method, since the phase variation will not do much good for the image formation.

COMMENT With this issue of the book review section the column is completing its first year. The publishers have cooperated fully in sending material for review - in fact, we now get more books than we can possibly review. However, this is as it should be; it allows the editor of this column to be selective in the volumes reviewed. This may be a little unfair but to be realistic some selection has to take place - but we will list other titles that have been received with a brief statement on their content. Hence, the section later headed 'Other books received.'