This paper presents the tool chain, methodology, and initial results of a study to provide a thorough, objective, and quantitative analysis of the design alternatives for space software defined radio (SDR) transceivers. The approach taken was to develop a set of models and tools for describing communications requirements, the algorithm resource requirements, the available hardware, and the alternative software architectures, and generate analysis data necessary to compare alternative designs. The space transceiver analysis tool (STAT) was developed to help users identify and select representative designs, calculate the analysis data, and perform a comparative analysis of the representative designs. The tool allows the design space to be searched quickly while permitting incremental refinement in regions of higher payoff

We describe the results of a series of psychophysical experiments that investigated the relationships among defect visibility, content importance, and perceived impairment in digital video. Various types of controlled defects were inserted into normal video and shown to our test subjects. The defects varied in their strength, location, appearance, size, and duration. We measured three specific subjective quantities: the defect detection probability, the perceived impairment, and the content importance. Impairment was found to be tightly related to defect visibility but only weakly related to content importance.

A compact color descriptor and an efficient indexing method for this descriptor are presented. The target application is similarity retrieval in large image databases using color. Colors in a given region are clustered into a small number of representative colors. The feature descriptor consists of the representative colors and their percentages in the region. A similarity measure similar to the quadratic color histogram distance measure is defined for this descriptor. The representative colors can be indexed in the three-dimensional (3-D) color space thus avoiding the high-dimensional indexing problems associated with the traditional color histogram. For similarity retrieval, each representative color in the query image or region is used independently to find regions containing that color. The matches from all of the query colors are then combined to obtain the final retrievals. An efficient indexing scheme for fast retrieval is presented. Experimental results show that this compact descriptor is effective and compares favorably with the traditional color histogram in terms of overall computational complexity.

The signal-dependent rank order mean (SD-ROM) filter is effective at removing high levels of impulse noise from 2-D scalar-valued signals. Excellent results have been presented for both a two-state and a multi-state version of the filter. However, implementation of the two-state SD-ROM filter requires the selection of a set of threshold values. We propose a method for choosing the thresholds based on a statistical characterization of an input image. The method approximates the histogram of an image with a weighted sum of Gaussian distributions. Using the statistical model and the input distributions, the likelihood of correctly identifying impulses is estimated as a function of the thresholds. By maximizing the likelihood of correct detection, optimal thresholds are predicted. The performance of the algorithm using the predicted thresholds is compared to the optimal performance found using a brute-force search.

We propose a new class of mapping-based nonlinear quadratic Volterra filters for image enhancement. The input signal is mapped prior to filtering by a memoryless one-to-one nonlinear mapping function. The mapping function can be described by any arbitrary function and implemented by a simple look-up table. If necessary, an inverse mapping function is applied after filtering. Using this general approach, the quadratic Volterra filter output can be adjusted to a desired response. Some results are presented where a high-pass quadratic Volterra filter is used to enhance an image by unsharp masking.

MIRTIS is an environment which employs model-based synthesis techniques to generate real-time image processing applications. The system is capable of creating very high performance implementations of a large class of image processing computations. It automatically data parallelizes the computations using the split-and-merge processing model and executes them on a parallel hardware architecture, a network of C40 DSPs. MIRTIS provides a high level programming interface which masks the complexities of the underlying parallel implementation. Graphical tools are used for building models and controlling the running applications.

We describe a programming environment which is being developed for the automatic generation of parallel image processing applications. Through the use of model-based software synthesis, we transparently create large grained data parallel applications which can be executed on arbitrary processor networks. The high-level abstractions provided by the modeling paradigm isolates the user from the complexity of the underlying implementation, allowing developers with little or no experience in parallel programming to rapidly create parallel applications. The data parallel modeling facilities perform the same tasks as the data alignment and distribution compiler directives of High Performance Fortran and the aggregate objects of pC++. However, we have found that by introducing the parallelism on the system level, instead of in the algorithm, we can use traditional compilers and leave the application specific code unchanged. This allows us to take advantage of existing well developed image processing code libraries. Here we describe a system which generates data parallel versions of applications created in Khoros, the popular image processing package developed by the University of New Mexico. This system retains the best qualities of Khoros: its interactive and experimental nature, and its visual interface, but adds the capability for automatically generating much higher performance parallel implementations when needed. This system demonstrate the suitability of the model-based approach for developing parallel imaging software.>