Blur detection in image sequences recorded by a wearable camera

A new method based on the Discrete Cosine Transform (DCT) and the Otsu method for blur detection in image sequences is proposed in this paper. In the first step, the standard deviation (STD) and the DCT coefficients are utilized to detect blurred and homogeneous areas in each image. Then, the Otsu method is used to calculate an adaptive threshold in each segment of the image sequence. Our experiments illustrate that the proposed method is robust and effective.