Project Gallery

youFinder is a successful product developed at BoulderLabs to identify groups of families and friends in collections of up to 50,000 images. The process begins with the use of facial recognition to extract and match faces from the images. We then supplement the facial recognition data with a number of additional metrics before making decisions about what groups of images belong together.

Technology highlights include:

• Efficient data structures and graph theoretic algorithms, with hundreds of thousands of nodes and hundreds of millions of edges.
• Multi processor, multi-threaded application to process the work load in reasonable clock time.
• Comprehensive Graphical User Interface that creates and caches thousands of scaled pictures for efficient presentation.

Screenshot:

Boulder Labs worked with the University of Colorado to develop a three-dimensional computer-aided design tool to address the needs of researchers in the field of atomic physics. The resulting product, LiveAtom, is currently being used by several universities and government laboratories to assist with their research.

Since the creation of the Bose-Einstein condensate in 1995, physicists have been performing increasingly sophisticated experiments with cold and ultra-cold atoms. One of the major challenges with these experiments is the absence of a standard tool to assist with experimental design or atomic modeling; until now, each research group or laboratory had to create their own tools.

Boulder Labs' CAD software addresses this challenge. The tool is fast, intuitive and easy to use, yet it performs sophisticated physical calculations and provides advanced 3D-visualization of the results.

See more in this video:

Screenshots:

Boulder Labs developed algorithms to check the grammar of children's writing for a client. With our expertise in linguistics, statistical learning, and software architecture, we are catching some of the most common errors made by 3rd-5th graders. We have tackled some of the most challenging tasks in natural language processing with results superior to competing products. We detect over thirty kinds of grammatical errors, including subject-verb agreement errors, run-on sentences, and more.

Our client has integrated our work with their production software that is currently in use in hundreds of homes and schools across the country.

The cell voltage monitor monitors the voltages of an array of up to 48 batteries and sends the data to a nearby computer.

Large arrays of low voltage batteries are commonly used for storing and using electricity generated by renewable sources, such as solar or wind. One problem with this approach is that a single bad cell will degrade performance of the system, and may damage the other cells. By making it easy to monitor the voltages of all the cells over time, BoulderLabs' Cell Voltage Monitor will allow the operator to quickly detect a bad cell and replace it with a good one.

Using a combination of off the shelf hardware and custom components we designed a circuit board that may be hooked up to an array of 48 batteries, where each cell carries a maximum of 5 volts. Using class 1 bluetooth we are able to transmit the voltage readings to a computer up to 100 meters away for processing.

The purpose of this project is to manage an array of high speed cameras and the data that each camera captures. The cameras capture at a variety of speeds up to 50,000 frames per second. The software provides a LabVIEW interface to initialize the cameras, display real time synchronized images, and store capture data to disk. We also provide a Matlab interface to easily access the stored capture data in a format that is synchronized across multiple camera frame rates.

The Delta Smelt is an endangered species of fish endemic to the Sacramento river delta. This small (2-3" long), slender fish is very susceptible to changing conditions in the delta. In 2007, a federal court imposed significant limitations on the amount of water from the delta that could be used for irrigation in order to protect the Delta Smelt.

SureWorks LLC asked Boulder Labs to help develop the software for SmeltCam -- a fully-submersible, non-lethal, vision-based product to monitor Delta Smelt populations. We developed algorithms to track unique objects through video captured on a high-speed camera, determine if the objects were fish, and to perform fully-automated species identification.

Learn more about what we did in this video:

Marine biologists using SealFinder confirm and identify seals on arctic ice about 100 times faster than using manual inspection.

Our software uses image processing techniques and machine learning to present biologists with candidate seals on a probability basis. By weighting 14 characteristics, we have achieved a level of success far better than anything used by the NOAA experts.