This project aims to develop a defect detection and analytic system. Defect detection act as an important role for quality control in die casting industry due to imperfect manufacturing process. Die casting is an automated metal casting process in which molten metal is pressed into a mold cavity under high pressure and at a high filling speed. Due to imperfect manufacturing processes, some of the product might have various defect such as deformed, pinhole, crack, dent, bump and others. Instead using human vision for defect detection, computer vision and image processing techniques are proposed to detect defection for different types of product. The camera capture the image of the product and the image is undergoes prepossessing before it is compared with the perfect product for defect detection. The system gives the detection result and type of defect if the product is defected. The system also provides an overview of the number of defected product detected, type of defect detected and the overall percentage of defective.