Image Labelling using an Associative Memory
نویسنده
چکیده
This paper presents an application of an associative memory neural network to the complex task of labelling the parts of an image. The network identiies features in the image, and recalls associations between features and the objects which they comprise. The identiication of features is tolerant of noise and clutter in the image. Features are associated with parameterised descriptions of objects, and the parameter transform has been extended to allow the simultaneous labelling of many diierent classes of objects.
منابع مشابه
Neural Associative Processing of Document Images
Binary neural associative memories are attractive for image processing because of their speed of operation for learning associations and for recalling them. We have added feedback to a feed-forward associa-tive memory, to produce a pattern completion network which performs translation-invariant pattern completion, and at the same time resolves contradictory image information. The image completi...
متن کاملDocument Feature Recognition using a Mesh of Associative Memories
This paper describes a new approach to the problem of identiication of complex objects in document images. The novelty of the approach lies in its use of a distributed representation for objects which seeks to overcome some of the problems of data noise and incompleteness. Objects are modelled as a set of features and the relationship between neighbouring features. Feature recognition is perfor...
متن کاملImage Object Labelling and Classification Using an Associative Memory
An essential part of image analysis is the location and identiication of objects within the image. Noise and clutter make this identiication problematic, and the size of the image may present a computational problem. To overcome these problems, we use a window onto the image to focus onto small areas. Conventionally we still need to know the size of the object we are searching for in order to s...
متن کاملContent-Based Image Retrieval Using Associative Memories
The rapid growth in the number of large-scale repositories has brought the need for efficient and effective content-based image retrieval (CBIR) systems. The state of the art in the CBIR systems is to search images in database that are “close” to the query image using some similarity measure. The current CBIR systems capture image features that represent properties such as color, texture, and/o...
متن کاملObject Recognition in Image Sequences and Robust Associative Image Memory using the Multilevel Hypermap Architecture
The introduced system for object recognition and tracking uses an associative memory for storing prototypes of objects. The Multilevel Hypermap Architecture (MHA), a self-organizing neural network approach, is used, to construct a robust system. To process form variant objects the MHA is extended to work with masked input data. Because of using scaled input objects, the system is invariant to t...
متن کامل