Leveraging Blur Information for Plenoptic Camera Calibration

Abstract

This paper presents a novel calibration algorithm for plenoptic cameras, especially the multi-focus configuration, where several types of micro-lenses are used, using raw images only. Current calibration methods rely on simplified projection models, use features from reconstructed images, or require separated calibrations for each type of micro-lens. In the multi-focus configuration, the same part of a scene will demonstrate different amounts of blur according to the micro-lens focal length. Usually, only micro-images with the smallest amount of blur are used. In order to exploit all available data, we propose to explicitly model the defocus blur in a new camera model with the help of our newly introduced Blur Aware Plenoptic (BAP) feature. First, it is used in a pre-calibration step that retrieves initial camera parameters, and second, to express a new cost function to be minimized in our single optimization process. Third, it is exploited to calibrate the relative blur between micro-images. It links the geometric blur, i.e., the blur circle, to the physical blur, i.e., the point spread function. Finally, we use the resulting blur profile to characterize the camera’s depth of field. Quantitative evaluations in controlled environment on real-world data demonstrate the effectiveness of our calibrations.

Improved blur circle detection method for geometric calibration of multifocus light field cameras

“…Calibrations are conducted on the calibration datasets (R12-A, R12-B, and R12-C) of the Raytrix R12 camera provided by Labussière et al 17 For comparison, 10 light field images of the checkerboard with the relative step of 10, 10, and 50 mm are extracted from these datasets. In addition, the calibration experiments of the Raytrix R29 light field camera are performed.…”

“…Figure 14 shows the detection results by the Geiger corner detector and method. 17 As shown in Fig. 14, although the resolution of the subimages is low, both Geiger corner detector and method 17 can detect most of the corner points, but the accuracy of Geiger corner detector is better than that of method, 17 indicating that the Geiger corner detector is applicable to the corner detection in subimages of the multifocal light field camera.…”

“…The geometric parameters of R12-A, R12-B, and R12-C are obtained from the RxLive software in the literature. 17 For dataset R29-D, the geometric parameters are taken from the literature. 27 The focal length of the main lens is 50 mm for the R12-A, R12-B, and R12-C, and 100 mm for the R29-D. Table 3 summarizes the initial geometric parameters for each dataset and the geometric parameters obtained by the improved blur circle detection method and the blur circle detection method in Ref.…”

“…In multifocus light field cameras, the microlenses with different focal lengths form the blur circles with different radii in the white raw image. Based on these characteristics, Labussière et al 16,17 introduced the geometric parameters of multifocus light field cameras into the calibration model through the blur circle radii of the image corners formed by the checkerboard corners, and realized the geometric parameter calibration of multifocal light field cameras. They also proposed a precalibration step using the blur circle radii of the subimages in white raw images to provide a robust initial estimate of geometric parameters.…”

“…19 However, none of the above methods consider the effect of the incorrect corners on the calibrations. The projection points of the checkerboard corners are mainly determined by the conjugate relationship between the image corners and projections with respect to the main lens, 15 the centers of gravity of corner clusters, 8,16,17 or the center points of corner clusters. 19 All the above projection point detection methods of the checkerboard corners are greatly affected by the uneven distribution of image corners.…”