<p>Common Law System in India considers analysis and interpretation of previous judgments very important. Judgments are complex documents as they are unstructured, written in natural language with legal terms, various details about the case, laws, have improper sentences with abnormal punctuation, irrelevant data, lots of acronyms and abbreviations. Legal practitioners require information from these documents for several legal tasks like sentence identification, recommendation, similarity of legal documents and many more. Owing to the need for information, complexity and huge number of judgments, understanding and retrieving information from these documents manually is time-consuming and laborious. This raises the need for automated processing of documents. Several efforts have been taken to automate legal tasks, but lack of processed corpus produce major challenge. These characteristics of Indian legal documents make them unsuitable for traditional Natural Language Processing (NLP) based text. To overcome these research challenges, we develop a novel dictionary for mapping acronyms and abbreviations, Indian Legal Acronyms and Abbreviations Dictionary (ILAAD) and propose a text preprocessing technique, Indian Legal Text Preprocessing (ILTP) explicitly for Indian court judgments. The proposed ILTP is applied on documents from unprocessed corpus to generate processed corpus. The documents from processed corpus is then used for generating summaries utilizing NLP models. The summarized text by models is then compared with the summaries by legal experts through ROUGE and BERT Score. Both the metrics show better scores for summaries generated from processed corpus.</p>