D. A. Waterman

Pattern-Directed Inference Systems

Chapter 6 puts the information covered to that point into direct application. The authors first discuss in some detail a crop identification and acreage estimation case study. This is followed by rather brief discussions of five selected management problems: large area land use inventory and forest, snow-cover, geologic, and water-temperature mapping. Serious students will wish to supplement these with studies of problems pertinent to their own areas of special interest. While much of the information presented is valuable, I see little justification for the final chapter since most of the material in it could well have been worked into other parts of the text. A few imperfections merit comment. Reproduction of some of the aerial-photographs and images does not meet the standards which were imposed on the drawings. For example, the images in Fig. 5-39 are difficult to interpret, although that problem may relate more to the small size of each wave band illustrated than to the quality of photographic reproduction. The areas shown in Fig. 1-7 to illustrate the three spectral regions are not the same scale; further, the same areas (with the same scale problem) are shown in Fig. 5-41. Fig. 6-13 contributes the little to an understanding of the selection or appearance of training areas; nor does Fig.-6-15 to-the selection of test areas. Three chapters have brief but useful summary sections. The other four would have benefited by a similar procedure. While the selection of terms to include in a glossary is a difficult task, a few which are encountered frequently in quantitative remote sensing were omitted, e.g., band ratioing, minimum Euclidean distance elassifier, maximum likelihood classifier smoothing, vector, etc. While there are savings in printing costs to have all color plates grouped on four pages, I found this system awkward to use and disruptive of comprehension. I was surprised, too, that answers to the questions posed after the various sections are not given. Individuals using the text on a self-study basis probably would not have a background adequate to verify their answers without such assistance. These, though, are relatively minor criticisms. Overall, this is one of the best sources of information that I have encountered on the subject of quantitative remote sensing. It would serve well as the textbook for courses at various levels and for students with a wide range of backgrounds. Professionals in the field of remote sensing will wish to add this volume …