Global Mapping Of Human Settlement: Experiences, Datasets, and Prospects
Global Mapping Of Human Settlement: Experiences, Datasets, and Prospects
复制标题
DOI：
10.1111/j.1477-9730.2010.00574_3.x
复制
发表时间：
2010-06
影响因子：
2.4
通讯作者：
P. Longley
中科院分区：
文献类型：
--
作者：
P. Longley
全球人类住区测绘：经验、数据集和前景参考数据集，以及结果的呈现和解释。本书还明确指出，没有一种普遍接受和适用的准确性评估方法，并鼓励读者利用对关键问题的理解来为给定的应用构建适当的方法。通过对准确性评估中的许多关键问题进行讨论，本书极大地帮助读者深入了解如何为特定应用设计准确性评估程序。尽管这本书有积极的特点，但也存在一些严重的担忧。尽管作者明确地打算关注关键问题，并因此保持简短易懂的文字，但这本书有时确实感觉有点肤浅。人们可能合理地预期，许多读者会从对抽样设计和提供与流行方法相关的关键方程等问题的更全面的讨论中受益。同样，虽然很高兴看到有关模糊分类的讨论，但尽管有关该主题的文献丰富，但只考虑了有限范围的问题和方法。更一般地说，这本书也有点令人失望，因为它提倡的方法越来越被认为不适合许多应用。例如，尽管最近对这种测量精度评估应用的批评颇多，但本书还是提倡使用 kappa 一致性系数。同样，文本仍然鼓励混淆矩阵的标准化，尽管这个过程有许多不良影响（尤其是用户和生产者准确性的均衡）。此外，尽管自第一版出版以来，有关准确性评估的文献已大大增加，但引用的参考文献似乎相当过时；从关于准确性评估的大量文献来看，似乎仅参考了第一版出版后十年内发表的 10 篇文章。最后，应该指出的是，文本有时在某些地方有点尴尬（例如，标题可能表明这本书是关于图像辐射测量的准确性，而不是图像衍生地图），尽管公平地说，作者的意思总体上是明确的。因此，这本书有许多值得称赞的特点，特别是在其目标和覆盖范围方面，但它也有一些地方缺乏深度，感觉有点过时。它的积极属性足以让我立即将其添加到我所教授的研究生课程的阅读清单中，但它的使用伴随着“健康警告”和一系列广泛的进一步参考。尽管如此，本书在传达有关重要主题的有用信息方面对社区具有巨大价值，并且是对成功的第一版的明显进步。虽然我相信这本书会被广泛使用，但遗憾的是作者没有抓住机会就这个主题提供更深入、更严格和更现代的讨论。也许这些担忧将在第三版中得到解决，从而成为一本精彩的书。
Global Mapping Of Human Settlement: Experiences, Datasets, and Prospects reference data-sets, as well as the presentation and interpretation of results. The book is also explicit in stating that there is no one universally accepted and applicable method of accuracy assessment and encourages readers to use an understanding of the key issues to construct an appropriate approach for a given application. By providing a discussion on many of the key issues in accuracy assessment, this book greatly helps provide the reader with insight into how an accuracy assessment programme might be designed for a particular application. Despite the positive features of the book there are some serious concerns. Although the authors have clearly set out to focus on key issues, and so keep to a short and accessible text, the book does feel at times a little superficial. One might reasonably expect that many readers would have benefited from a fuller discussion on issues such as sampling designs and the provision of key equations associated with popular methods. Similarly, while it is good to see a discussion in relation to fuzzy classifications, only a limited range of issues and approaches are considered despite the rich literature on the topic. More generally, the book is also a little disappointing in that it promotes methods that are increasingly seen as inappropriate for many applications. For example, the book promotes the use of the kappa coefficient of agreement despite the considerable recent criticism of this measure for accuracy assessment applications. In a similar vein, the text still encourages normalisation of the confusion matrix, even though this process has many undesirable effects (not least the equalisation of the user’s and producer’s accuracy). Moreover, although the literature on accuracy assessment has grown enormously since the publication of the first edition, the references cited seem quite dated; from the vast literature on accuracy assessment, it seems that reference is made to just 10 articles published in the decade after the publication of the first edition. Finally, it should be noted that the text is sometimes a little awkward in places (for example, the title might suggest that the book is about the accuracy of image radiometric measurements rather than of image-derived maps), although, to be fair, the authors’ meaning is generally clear.This book, therefore, has many commendable features, notably in relation to its aims and scope of coverage, but it also lacks depth in places and feels a little out of date. Its positive attributes were sufficient for me immediately to add it to the reading list for a postgraduate course that I teach, but its use is accompanied with ‘‘health warnings’’and an extensive set of further references. Nonetheless, this book has great value to the community in conveying useful information on important topics and is a clear advance on the successful first edition. While I am confident that the book will be widely used, it is, however, a shame that the authors did not seize the opportunity to provide a deeper, more rigorous and more contemporary discussion on the subject. Perhaps the concerns will be addressed in a third edition, resulting in a spectacular book.