This report presents results from the National Institute of Justice's Electronic Crime Technology Center of Excellence's (ECTCoE) evaluation and testing of RedLight (Beta) Software Version 0.1.0.0, which is an application designed to detect pornographic image content within a computer file system.
The Redlight software uses sophisticated image analysis techniques in detecting pornography through characteristics such as high concentrations of skin tone and edges indicative of humans. The software will only run on Windows XP or later. The evaluation concludes that RedLight would be a "great asset to the investigative tool box." In one of the tests, for example, out of 29,721 images, RedLight identified 358 as possible pornographic images. Although those images were subsequently determined not to be pornographic, they met RedLight's criteria for skin tone and shapes. This finding means that the forensics examiner would be saved the time of searching each of the 29,721 images to determine if any were pornographic. Thus RedLight is useful in assisting an examiner in identifying potential evidence in a child pornography case by minimizing the amount of graphic files that may constitute evidence. Once RedLight identifies potential pornography, a more in-depth examination can be conducted. RedLight also enables the user to narrow the search for date and time blocks or files of a certain size, which can be useful if the approximate date and time of the downloading of the images is known. In addition to presenting the evaluation findings, this report describes the download and installation of Red Light, configuration of RedLight, and the objectives and procedures of the four tests that were conducted on RedLight. 6 figures
Downloads
Similar Publications
- Introducing the NIJ Forensic Intelligence Framework: Pillars and Guiding Principles for Successful Implementation
- Enhancing Genetic and Epigenetic Sample Preparation with Microfluidics
- Stage Transitions in Lucilia sericata and Phormia regina (Diptera: Calliphoridae) and Implications for Forensic Science