Object Sorting Mechanism based on Image Analysis Using LBP and Color Texture Approach

Nikita, Shankar and Kamath, Ravishankar H and Sharma, Dheeraj (2018) Object Sorting Mechanism based on Image Analysis Using LBP and Color Texture Approach. Journal of Advanced Research in Dynamical and Control Systems, 10 (03). pp. 1396-1399. ISSN 1943-023X

[img] PDF
4639.pdf - Published Version
Restricted to Registered users only

Download (177kB) | Request a copy

Abstract

This research work was carried in IATC Siemens, Panchkula. The purpose of this research paper was to design and analyze images which differ in texture (by real-time acquisition system) to gather 2-D information of the object in non-linear form converted to a binary output response (which decides whether the objects should be passed or rejected). Its main objective was to optimize the filters namely SSIM, edge-gradient and co-relational and for classifying objects travelling on conveyor belt with short travelling distance (1m). Acquisition of image of moving object is done on horizontal and vertical camera. Then the images are analyzed under the gradient filters to separate the images and the captured images are sent to the cross-referenced metrics. Along with this, the colors of the objects are also cross-referenced to the pre-fed system values. For Process Control once the filter’s output response is obtained, the binary values are then cross-referenced with external hardware consisting of microcontroller interfaced with servomotors and IR sensors. Using GUI interface of MATLAB, a counting mechanism is also developed. Effectiveness of the acquisition is detected and control response is analyzed through real time object detection and sorting or removal followed by time-based response calculation of processing filter. I.

Item Type: Article
Uncontrolled Keywords: data acquisition, processing and process control systems
Subjects: Engineering > MIT Manipal > Instrumentation and Control
Depositing User: MIT Library
Date Deposited: 06 Jul 2018 08:45
Last Modified: 06 Jul 2018 08:45
URI: http://eprints.manipal.edu/id/eprint/151494

Actions (login required)

View Item View Item