"Seeing is believing". Microscope images or high-speed image sequences often provide enough information of the studied process or phenomenon. However, image analysis can be used to yield quantitative measurement information from the image data. There are a plenty of techniques and algorithms available for various measurement tasks. In the following, some principal analysis techniques are presented. Object recognition and overlapping object detection Image analysis can be used to recognize objects from the image. One of the challenges in object recognition is the detection and separation of overlapping objects. In the images with high object density, the neighbouring objects are likely to form cluster in the image. Pixact has developed very effective algorithms to separate individual objects from the clusters. As a result of overlapping object detection the number of detected objects increases significantly improving the measurement reliability. Image power spetral density Measurement signals are typically characterized by their instant values and statistics such as mean and variance. The power spectral density (PSD) represents the variance of the signal as a function of frequency (or wavelengths). The distribution of variance over the frequency spectrum allows detailed analysis time or spatial scales of the signal variation. This type of analysis is very useful in many measurement applications, such as paper analysis and flow dynamics. Read more... info@pixact.fi    tel. +358 (0)10 439 2400 (C) Pixact Ltd. 2009. All rights reserved