At the Botanical Institute of the University of Karlsruhe (Karlsruhe, Germany), C.

At the Botanical Institute of the University of Karlsruhe (Karlsruhe, Germany), C. Boxler-Baldomà is using a PC-based image-processing package to examine the influence of air contaminants on spruce trees. With the image-analyzing functions of the Windows-based analySIS package from Soft Imaging System Corp. (Lakewood, CO), she is characterizing forest damage by interpreting the yellowing of the spruce needles. This discolorization, caused by the degradation of the photosynthesis mechanism, is being analyzed by measuring the chlorophyll content in the needles` chloroplasts.

However, because of varying image-contrast ranges, brightness, and illumination, a fully automatic image analysis has not been possible. "Because the structural changes in the chloroplasts are complex, statistical image-processing methods are too coarse and time-consuming to quantify gradual changes of individual structures," says Boxler-Baldomà.

"Therefore, we developed a method to measure only the most important chloroplast parameters, so that judgments can be made about the seasonal changes of individual chloroplast structures in healthy and yellow-spruce needles. In this method, transmission-electron-microscope im ages of 12,000X magnification were taken of the chloroplast areas. Then, specially defined analySIS macros were used to analyze the images.

"Because these [macros] execute contrast optimization, binarization, filtering, detection classification, and output of the measured results, we are assured a high throughput of samples in a relatively short period of time," says Boxler-Baldomà. However, certain operations, such as changes in the gray values, splitting of the brightness areas, or determination of threshold values for the creation of binary images, must be determined for each image because of varying image qualities.

Per-pixel processor aims for real-time motion detection

Mobile robots and self-guided vehicles often use image motion information to track targets and obtain depth information from scenes. Unfortunately, traditional motion algorithms running on Von-Neumann processing architectures are computationally intensive, preventing their use in real-time applications. Consequently, researchers developing image motion systems often turn to faster, more unconventional processing architectures.

One such architecture is the processor-per-pixel design, an approach recently proposed by Marc Tremblay and his colleagues at the Computer Vision and Systems Laboratory of the Université Laval Sainte-Foy (Québec, Canada). Fabricated in 1.5-mm CMOS and 0.8-mm BiCMOS, the low-resolution prototypes currently under development integrate a 50 x 50 smart-sensor array with integrated signal-processing capabilities.

"Each pixel integrates a photo detector, an analog signal-processing module, and a digital interface," says Tremblay. "Pixels are sensitive to temporal illumination changes produced by edges in motion. If a pixel detects an illumination change, it signals its position to an external digital module. Here, time stamps from a temporal reference are assigned to each sensor request. These time stamps are stored in local RAM and are later used to compute velocity vectors. The digital module also controls the sensor`s analog I/O signals and interfaces the system to a host computer through a serial link," he adds.