Both caption and scene texts which are found in images and video frames containrnvaluable information. These texts can be used for many applications to answer questionsrnlike what, when, where, and by who to give context to the images and video frames.rnSo, automatic text detection enhances the user's understanding of the media content.rnIn Ethiopia, most street posts and promotional boards are written in multi-lingualrncharacters such as Latin (English, Afaan Oromo etc.) and Ethiopic (Amharic, Tigrignarnetc.). In this work, we have studied Ethiopic and Latin multilingual text detection andrnscript identi cation from videos and images for both caption and scene texts.rnAfter the images and video frames are pre-processed, maximally stable extremal regionrn(MSER) algorithm, aspect ratio and stroke width transform (SWT) algorithm are usedrnto extract text regions and discriminate non-text patterns from texts, respectively. Thenrntexture features are computed using local binary pattern (LBP) from the extractedrnregions. Finally, support vector machine (SVM) is used to classify text region vsrnnon-text using the computed LBP features. In the next phase of our work, which isrnscript identi cation, the detected text regions are binarized using Niblack's algorithm.rnRadon transform was applied on the binarized text regions to detect and correct skew.rnSegmentation of lines using horizontal projection pro le followed by word segmentationrnusing vertical projection pro le is done when the text region contains more than onernline of text. From the resulting text words, texture features are computed again usingrnLBP and the text words are categorized to their respective script classes using SVM.rnWe used the International Conference on Document Analysis and Recognition(ICDAR)rn2003 data set as well as prepared a new multilingual Ethiopic and Latin script imagerndataset to evaluate our method. Our text detection method performs better compared with the state of the art method with precision 5%, recall of 10% and 8% f-measure onrnICDAR 2003 dataset. The text detection was also evaluated on our dataset, where 81%rnprecision,74% recall with a f- measure of 77% was obtained. The overall system givesrn79.9% accuracy of script identification.