The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Publications Copernicus
Articles | Volume XLVIII-1/W1-2023
25 May 2023
 | 25 May 2023


G. Scarmana

Keywords: Image restoration, image blurring, image compression

Abstract. Mobile mapping technology has transformed the way in which we capture and map our surroundings. The widespread use of mobile devices such as smart phones and drones has made data collection more efficient and accessible than ever before. However, the image quality of this data is often compromised due to the static or motion blurs resulting from the device being stationary or moving during the data collection process. This can lead to a loss of information, making the data less useful.

To address this issue, an inverse filtering or deblurring method based on the theory of least squares is examined. This method implements a deconvolution process in the space domain and can restore the original image with a high degree of accuracy if the model of the function or filter that blurred the image is known or can be established.

The accuracy and validity of the deblurring results are presented in terms of the root mean square error (RMSE) of the differences of pixel intensity values between the original and de-blurred images. In tests using grey-scale aerial images of varying entropies and different types of blurring, the RMSE value never exceeded ±5 pixel intensity values. This discrepancy is due to the rounding of pixel values resulting from image operations.

The deblurring method presented in this work is an adaptation and extension of a previously described process, tailored specifically for filtering and restoring images - particularly aerial imagery - affected by static and motion blur. This process could also be applied in image compression processes and techniques of transmission over digital links, where blurring filters can suppress noise and increase the dependence between neighbouring pixel values, thereby improving the compression ratio (CR).