Table detection is a crucial step in several document anal- ysis applications as tables are used to present essential information to the reader in a structured manner. In companies that deal with a large amount of data, administrative documents must be processed with rea- sonable accuracy, and the detection and interpretation of tables are cru- cial. Table recognition has gained interest in document image analysis, particularly in unconstrained formats (absence of rule lines, unknown information of rows and columns). This problem is challenging due to the variety of table layouts, encoding techniques, and the similarity of tabular regions with non-tabular document elements. In this, paper, we make use of the location, context, and content type, thus it is purely a structure perception approach, not dependent on the language and the quality of the text reading. We evaluate our model on invoice-like doc- uments and the proposed method showed good results for the task of table extraction.