The increase in spatial resolution has enabled analysis and detection of images obtained through satellite.To increase the accuracy in image processing, an unwanted shadow needs to be processed effectively. A quality check mechanism governs the false reconstruction problems. In the proposed paper we would like to suggest use of Lib SVM algorithm. The original image obtained from satellite is first supervised to identify shadow and non shadow regions. The shadow region is then filtered using morphological filter and a border is created around shadow. With the help of Lib SVM algorithm the shadow image can be removed using post classification methods and shadow free image can be obtained. The remote sensing of an object has become possible because of very high resolution (VHR) satellite images. The increase of spatial resolution demands new analysis, classification, and change in detection techniques. The objects, like small building structures, trees, vehicles, and roofs can be revealed with the help of VHR images. In urban areas the presence of high rise structures such as buildings, towers causes large change in surface elevations and results in longer shadows. Though shadow characteristics are used in recognition of building position, height and other useful parameters it may cause a high risk to produce false color tones, to distort the shape, to merge, or to lose objects. Due to the partial or total loss of information in the image, shadows may impact adversely in the exploitation of VHR images as it influences detailed mapping that leads in flawed classification/interpretation. To soothe these drawbacks and to increase image credulous ability, two steps are necessary.