Video classifications are usually tailored towards categorizing videos into one or more predefined categories (e.g. genres) using the contexts associated with such categories. This limits their application to only ldquoproduction videosrdquo (i.e. video produced and edited for a viewing audience). We seek to make the classification criteria more flexible by classifying videos using low-level computable features that can be determined for any type of video independent of the context associated with its predetermined genre. The methodology adopted was based on choosing unrestricted computable features for developing a classification scheme. It extracted and analyzed the low-level components (key frames) and computable features (such as dominant color, lighting condition, and color dynamics) from sample videos. It then generated a model SVM classifier that was able to discriminate between tested videos to be classified. It finally, developed an interactive application to automate the extraction and analysis process.