JSR 34: Java^TM Advanced Imaging API 1.1

This JSR proposes the revision of the Java Advanced Imaging API 1.0 specification to version 1.1. The purpose of this revision is to add functionality which has been requested by customers, revise existing functionality in the light of customer feedback, and improve the interoperability of the Java Advanced Imaging API with other Java standard extensions (e.g., Java 3D).

• Volumetric imaging. True three-dimensional images and data structures will be added to the API. A standard mechinism for implementing operators for processing three-dimensional images will be defined. A number of sample operators will be specified and included as part of the core functionality.
• Image analysis. Image analysis is a subset of the discipline generally known as image understanding, machine vision, or pattern recognition. Image analysis provides the fundamental framework for extracting features of interest from an image, creating data structures to describe those features, and operating on those data structures to extract meaningful information. It excludes the higher-level processes which are used for formulating an overall synthesis of image content based on the feature information extracted.

  The 1.1 specification will standardize a framework for the development of image analysis prototypes and applications, allowing workers in this area to count on a standard set of concepts, terminology, and features. Developing a more detailed description of the work to be done in this area will be part of the architectural process.

• Interoperabilty between the Java Advanced Imaging API and the Java 3D API. Although the 1.0 API was designed with interoperability with Java 2D and Java 3D in mind, the integration of the Java Advanced Imaging API and the Java 3D API is incomplete. In the 1.1 API, the necessary architectural concepts and operators will be introduced to allow Java 3D and the Java Advanced Imaging API to smoothly interoperate.

  In particular, it is assumed that the Java Advanced Imaging API will most often be used as a 'back-end' engine by Java 3D to process an image data set for inclusion in a Java 3D scene graph. Joint work between the Java Advanced Imaging API and Java 3D API teams will define the necessary architectural features.

• Enhanced Network imaging. The 1.0 API provides for distributed processing of image data sets and operators anywhere on the network. In the 1.1 API, these capabilities will be enhanced to include the ability to more easily utilize additional compute resources and to define arbitrary wire protocols and compression techniques for the exchange of image tiles. These additions will also ease the implementation requirements for the input/output of certain image file formats. This work will be carried out in close coordination with the image file i/o standard extension work.
• New image operator specifications. New image operators. The core image operator functionality in the 1.0 API was chosen to address the most common and widely useful methods. In the 1.1 API, additional operators will be added to the core and some previous operators modified in light of the feedback we have received from customers.
• Core modifications. Some aspects of the 1.0 API core architecture may need to be modified in light of the new 1.1 functionality. Feedback from customers has also indicated some desirable additions to the architecture which will enhance potential performance and ease the implementation of new operators by application developers and third parties. Backward compatibility of 1.1 with 1.0 is a major goal; imcompatible changes will be made only when their advantages are overwhelming.

This JSR specifically excludes modifications to areas of the architecture not named above.

It is possible that the revisions to the Java Advanced Imaging API may make revisions in the Java 2D API desirable; any such revisions will be negotiated with the Java 2D API team in the proper timeframe.