Jitter noise modeling and its removal using recursive least squares in shape from focus systems.

Three-dimensional shape recovery from the set of 2D images has many applications in computer vision and related fields. Passive techniques of 3D shape recovery utilize a single view point and one of these techniques is Shape from Focus or SFF. In SFF systems, a stack of images is taken with a single camera by manipulating its focus settings. During the image acquisition, the inter-frame distance or the sampling step size is predetermined and assumed constant. However, in a practical situation, this step size cannot remain constant due to mechanical vibrations of the translational stage, causing jitter. This jitter produces Jitter noise in the resulting focus curves. Jitter noise is invisible in every image, because all images in the stack are exposed to the same error in focus; thus, limiting the use of traditional noise removal techniques. This manuscript formulates a model of Jitter noise based on Quadratic function and the Taylor series. The proposed method, then, solves the jittering problem for SFF systems through recursive least squares (RLS) filtering. Different noise levels were considered during experiments performed on both real as well as simulated objects. A new metric measure is also proposed, referred to as depth distortion (DD), which calculates the number of pixels contributing to the RMSE in percentage. The proposed measure is used along with the RMSE and correlation, to compute and test the reconstructed shape quality. The results confirm the effectiveness of the proposed scheme.

A new focus measure operator for enhancing image focus in 3D shape recovery.

Measuring the image focus is an important issue in Shape from Focus methods. Conventionally, the Sum of Modified Laplacian, Gray Level Variance (GLV), and Tenengrad techniques have been used frequently among various focus measure operators for estimating the focus levels in a sequence of images. However, they have various issues such as fixed window size and suboptimal focus quality. To solve these problems, a new focus measure operator based on the adaptive sum of weighted modified Laplacian is proposed. First, the adaptive window size selection algorithm based on the GLV is applied. Next, appropriate weights are assigned to the Modified Laplacian values in the image window based on the distance between the center pixel and neighboring pixels. Finally, the Weighted Modified Laplacian values in the image window are summed. Experimental results demonstrate the effectiveness of the proposed method.

A simplified approach using deep neural network for fast and accurate shape from focus.

Three-dimensional shape recovery is an important issue in the field of computer vision. Shape from Focus (SFF) is one of the passive techniques that uses focus information to estimate the three-dimensional shape of an object in the scene. Images are taken at multiple positions along the optical axis of the imaging device and are stored in a stack. In order to reconstruct the three dimensional shape of the object, the best-focused positions are acquired by maximizing the focus curves obtained via application of a focus measure operator. In this article, Deep Neural Network (DNN) is employed to extract the more accurate depth of each object point in the image stack. The size of each image in the stack is first reduced and then provided to the proposed DNN network to aggregate the shape. The initial shape is refined by applying a median filter, and later the reconstructed shape is sized back to original by utilizing bi-linear interpolation. The results are compared with commonly used focus measure operators by employing root mean squared error (RMSE), correlation, and image quality index (Q). Compared to other methods, the proposed SFF method using DNN shows higher precision and low computational time consumption.

Unorthodox approach toward microscopic shape from image focus using optical microscopy.

Shallow depth-of-field is an inherent property of optical microscope. Because of this limitation, it is usually impossible to image large three-dimensional (3D) objects entirely in focus. However, the in-focus information of the object's surface can be acquired over a range of images by optical sectioning of the object in consideration. These images can then be processed to generate a single in-focus image and further for 3D shape reconstruction using methods like Shape from focus (SFF). SFF represents a passive technique for recovering object shapes. Although numerous methods for SFF have been recently proposed, all follow similar precedent of focus measure application and depth recovery by maximizing the focus curves. As the conventional techniques assume the presence of prominent texture in the scene, the shape of weak textured surfaces are not recovered properly. In this manuscript, we have followed an unorthodox approach to recover shapes of microscopic objects using SFF. At first, the in-focus image is obtained, pursued by computing depth along the edges and their neighbors present in scene. Empty spaces in the final depth map are then calculated by surface interpolation. The proposed approach works well even for objects with weak textures.