The project is developed using the U2Net model and working with an accuracy of 95% with all types of images. The project idea is based on U2-Net(U square net) published in Pattern Recognition 2020. Based on 𝐓𝐞𝐧𝐬𝐨𝐫𝐅𝐥𝐨𝐰, 𝐏𝐲𝐭𝐨𝐫𝐜𝐡, 𝐎𝐩𝐞𝐧𝐂𝐕, 𝐌𝐱𝐍𝐞𝐭, 𝐆𝐥𝐮𝐨𝐧𝐂𝐕, 𝐔𝟐𝐍𝐄𝐓, 𝐘𝐎𝐋𝐎 (𝐀𝐥𝐥 𝐕𝐞𝐫𝐬𝐢𝐨𝐧), 𝐀𝐖𝐒 𝐚𝐧𝐝 𝐅𝐥𝐚𝐬𝐤.