The work aims to propose a method of inkjet defect detection based on machine vision, so as to improve the accuracy of inkjet detection in food packaging process. The structure and process flow of the automatic inkjet system were analyzed, including carrying manipulator, conveying device, inkjet device, detection device, etc. With scanning code detection as the key research object, the image was acquired by machine vision, and the inkjet defect detection was realized by image processing algorithm, including template matching algorithm and vertical projection method. At the same time, the defect detection process was presented, which was mainly composed of image segmentation, character correction and segmentation, character segmentation, defect detection and other steps. The experimental results showed that, the recognition success rate of the proposed inkjet detection method could reach 99%, and the recognition success rate was high. The proposed method can effectively deal with the defects such as missed printing and achieve the automatic sorting of food packaging in place of the manual work.