One-dimensional image processing
One-dimensional image processing, or signal processing, forms the basis of some image-processing techniques. But in many applications, designers do not reconstruct images from collected data. Instead, they make pass/fail decisions by applying signal-processing techniques to information as or after it is collected.
One-dimensional image processing
Andy Wilson Editor
andyw@pennwell.com
One-dimensional image processing, or signal processing, forms the basis of some image-processing techniques. But in many applications, designers do not reconstruct images from collected data. Instead, they make pass/fail decisions by applying signal-processing techniques to information as or after it is collected.
One such technique, bi-spectral analysis, had limited use in imaging sunspots, seismic activity, and military targets. But, as contributing editor Larry Brown reports, high-speed, low-cost computers have allowed bi-spectral analysis to be used in the development of electroencephalogram systems. Using this technique, doctors can correlate a patient`s brain-wave activity to anesthetic dosage.
Barcodes are familiar as an application of one-dimensional imaging. But in manufacturing tracking applications, barcodes are often not up to the task. Contributing editor John Haystead reports on the latest two-dimensional barcodes and how they are easier to use in industrial applications.
While vision systems such as the solid-state system of the Galileo spacecraft use solid-state cameras to capture images, transmitting those images to earth is accomplished via serial radio transmission, explains John Haystead. Once downloaded to ground computers, signal- and image-processing techniques enhance, analyze, and interpret the images.
In every real-time signal- or image-processing application, processing speed is crucial. But for many years, pipelined or distributed computer architectures were the only means to increase throughput in vision systems. With the advent of field-programmable gate arrays, which can be reconfigured in real time, designers can embed image-processing algorithms to create reconfigurable vision systems. Contributing editor Barry Phillips explains how these devices are being used in compression, filtering, and target-tracking systems.
To help reconfigure such systems, computer-based hosts must rapidly respond to changing input data. That`s why designers are turning to real-time operating systems. As contributing editor Rick Nelson explains, these high-speed operating systems provide a way for designers of machine-vision systems to focus on their customers` high-level applications.
