Missing children‟s becoming a concern in the globe particularly in India, as more number of missing children is reported in day today life. This issue switches the research community to spotlight on this era. Several theories and models have been proposed to mitigate this issue, however almost all techniques use conventional methods such as register complaints, face recognition using image processing algorithms. No doubt these methods fulfill the gap of missing child, but may not be optimum because of detection delay and success rate. Moreover few authors presented deep learning concept to improve the detection process of missing child, though deep learning is used to overcome the issue of delay and success rate, but needs huge date set to train the model which is not available in the respective police satiation at the time of registering FIR. In this paper we use the face recognition module of deep learning to extract the landmarks from the face and encode it to give the feature vector, latter KNN is used to classify the feature vectors of the image extracted by deep learning. The presented approach minimizes the delay and improves the success rate to identify the missing child. We validate our approach using Python programming.