A Combination of Corrugated Cardboard Images Using Image Stitching Technique

This paper presents image stitching technique to combine corrugated cardboard images in order to have an image of the whole pile of cardboard. Image stitching algorithm is consist of the corner feature detection using Forstner method, feature matching while inconsistent features are eliminated by random sample consensus (RANSAC) and using bundle adjustment for image composition by refining the output image position. After that, this algorithm is tested to combine cardboard images on cut-off side type BC, 5 sheets simulated using HALCON program. The experimental results show that the proposed algorithm can combine cut-off side images correctly when they are overlapped about 25%. Then the output of the combined image is investigated by manually counting compared with the real number of the cardboard and it is found that the corresponding result is given.