Malaria is a life-threatening infectious disease transmitted through mosquitoes bites. According to the World Health Organization, it is a cause of more than half of million human deaths each year, some due to late or erroneous diagnosis. There are a few different techniques that are used to diagnose malaria with manual microscopy considered to be the most accurate. However, the manual assessment requires a lot of steps and time, leading to late diagnosis and is prone to human mistakes, even in experienced hands. With our prototype we aimed to develop a robust, unsupervised and sensitive malaria screening technique that is not only less expensive, and much faster, but also does not rely on human ability, which in turn would minimize the number of misdiagnoses.We propose the use of the segmentation based approach that applies various machine learning and computer vision algorithms for the detection of malaria infected red blood cells. From the computer vision point of view, diagnosis of malaria is a multi-part problem. A complete system must be equipped with functions to perform: image acquisition, pre-processing, segmentation (candidate object localization), and classification tasks. All features/parameters are computed from the data obtained by the digital images of the blood cells and is given as input to k-means clustering which classifies the cell as the infected one or otherwise.