Taking the role taken today by the image as a means of communication, this book aims to reflect on existing new ways of representing and retrieving images. We are accustomed to both the representation of the images for later retrieval and finding the same to be made through the use of words. However, in these pages studying a type of automated accounting systems is addressed and for retrieving images without resorting to textual attributes that define them. The advantages involved in these systems are evident for both the user and the professional documentation. A priori, among others solve the problem of the time taken to index quote pictures, inconsistency between indexers or graphic qualities difficulty of expressing in words. For example, when a user does not have a clear idea of ​​the image you want, the system will offer a sample from which visual similarity can browse until you find something relevant; moreover, no need to consider the most appropriate words to represent your idea. In this paper these systems, known in abbreviated form as CBIR systems are defined, and its origin is at the time the research productivity is discussed about the characteristics and main constraints faced by these systems are specified and marked the typologies we can classify existing CBIR applications, discussing some examples within each category. The author devotes a section to the study of the applications most directly related to the professional field of library and information science. Finally, the interaction problems that affect these systems in relation to the user, while a number of minimum requirements applicable to any of these systems are examined usability aims.