Performance Analysis of Focus Measures in a SFF-Inspired Approach for Sparse Scene Reconstruction

Senthilnathan, R and Subhasree, P and Sivaramakrishnan, R and Srinivasan, C R and Srividhya, R (2015) Performance Analysis of Focus Measures in a SFF-Inspired Approach for Sparse Scene Reconstruction. IJCTA, 20 (10). pp. 1-11.

[img] PDF
3.pdf - Published Version
Restricted to Registered users only

Download (814kB) | Request a copy


Shape From Focus (SFF) is a popular technique in the field of computer vision for scene reconstruction. The SFF technique is based on the fact that the focus levels of the pixels of the image preserves depth information. The usage of telecentric lenses for conventional SFF limits its application for only small objects so as to preserve magnification constancy. In the current research work a new SFF-inspired algorithm is developed which utilizes a wide angle lens in place of a telecentric lens. This extends the range of object that the system can deal with, though severe magnification changes occur when a stack of images are acquired with respect to the scene. This problem is addressed using a variable window approach when focus measures are computed. This paper is a segment of the larger research work which aims at evaluation of 15 different focus measures. The paper presents significant results of performance evaluation of four different focus measures most commonly used in SFF and auto-focus algorithms. The evaluation is carried out based on two different performance evaluation criteria namely root mean square error and computation time. The analysis of focus measures are carried out under various operating conditions such as different spatial resolution, window size, contrast changes, gray level saturation and camera noise.

Item Type: Article
Uncontrolled Keywords: Shape from Focus, Variable Magnification, Variable Window Size, Focus Measures
Subjects: Engineering > MIT Manipal > Mechatronics
Depositing User: MIT Library
Date Deposited: 16 Jan 2016 14:11
Last Modified: 16 Jan 2016 14:11

Actions (login required)

View Item View Item