Precise deterministic change detection for smooth surfaces

Abstract

We introduce a precise deterministic approach for pixelwise change detection in images taken of a scene of interest over time. Our motivation is for applications such as artefact condition monitoring and structural inspection, where a common problem is the need to efficiently and accurately identify subtle signs of damage and deterioration. The approach we describe is designed to compensate for the three most common sources of nuisance variation encountered when tackling the problem of change detection, namely: viewpoint variation due to camera motion between images, photometric variation due to lighting differences, and changes in image resolution/focal settings.To tackle viewpoint variation, particularly in areas of low texture, we propose the use of the generalised PatchMatch (PM) correspondence algorithm to compute a dense flow field. The flow field is regularized using a Thin Plate Spline (TPS) model which assumes a smooth underlying geometry and allows registration to be interpolated precisely through areas of low texture or uncertain flow. To compensate for low-frequency lighting variation, we fit a second TPS model to the photometric differences between registered images. Finally, to account for changes in focal settings, we estimate and apply a blurring kernel via optimisation over image differences.We provide a thorough evaluation of the performance of our method on an illustrative toy dataset and on two recent, real-world inspection datasets. Our approach performs favourably versus state-of-the-art baselines in both cases, while remaining relatively transparent to understand and simple to compute.

Change detection with cross enhancement of high‐ and low‐level change‐related features

“…To conquer the effects of camera pose misalignment and lighting difference, Feng et al [22] propose to jointly optimize camera pose alignment, lighting correction and low-rank based change detection from small scale to large scale. Image difference can also be used after the images are aligned in camera pose and lighting [23]. Although traditional CD methods are easy to be implemented, they still have problems in facing complex scenes for real-world applications.…”

Selecting change image for efficient change detection

“…Feng et al [6] proposed an iteration optimisation change detection by modelling the changes as iteration of camera pose alignment, lighting correction, and low-rank from two observations. Stent et al [18] proposed using the generalised Patch-Match correspondence algorithm to align images and using the thin plate spline model to estimate the illumination variation to conquer the camera pose variation and lighting difference between reference and query images. Gharbia et al [19] used the log ratio of two images after being registered with Scale-Invariant Feature Transform (SIFT) to detect the change.…”

“…Stent et al. [18] proposed using the generalised Patch‐Match correspondence algorithm to align images and using the thin plate spline model to estimate the illumination variation to conquer the camera pose variation and lighting difference between reference and query images. Gharbia et al.…”

Dual Compositional Learning in Interactive Image Retrieval

“…Difference Detection between Images. Whereas learn-ing the pixel-level difference and/or flows of consecutive images have been extensively studied (Stent et al 2016;Khan et al 2017) (Liu et al 2018) incorporate image retrieval to generate more diverse and richer sentences and (Liu et al 2020) add a text retrieval module to improve the quality of generated captions. (Qiao et al 2019;Joseph et al 2019) jointly train an image generation model with a captioning model for better text-to-image generation.…”