Prediction of treated Bambusa blumeana bamboo properties using computer vision and machine learning

Date of Publication

7-2025

Document Type

Master's Thesis

Degree Name

Bachelor of Science in Civil Engineering (Honors) - Ladderized

Subject Categories

Civil Engineering

College

Gokongwei College of Engineering

Department/Unit

Civil Engineering

Thesis Advisor

Jason Maximino C. Ongpeng

Defense Panel Chair

Maria Emilia P. Sevilla

Defense Panel Member

Cheryl Lyne C. Roxas

Daniel Nichol R. Valerio

Abstract (English)

Growing concern over environmental degradation and the construction industry’s damaging impact raises the need for sustainable materials and methods. Bambusa Blumeana is identified as a sustainable construction material that is easily accessible in the Philippines at a low cost. Its mechanical properties were investigated considering the influence of treatment methods and indicating properties (IP). Treatment methods include fire treatment and Copper-Chrome-Boron (CCB) chemical treatment. The mean compressive strength observed was 37.08 MPa, 39.42 MPa, and 40.62 MPa for untreated, fire-treated, and CCB-treated, respectively. While the mean tensile strength values were seen to be 85.18 MPa, 65.11 MPa, and 90.69 MPa for untreated, fire-treated, and CCB-treated, respectively. Through regression analysis, the observed independent IP with the highest predictability in terms of compressive capacity was linear mass at 12% MC (q12) followed by wall thickness (δ) and outer diameter (D). The developed artificial neural network (ANN) used these IPs as input with a measured performance of 17.38 and 13.44 for root mean squared error (RMSE) and mean absolute error (MAE), respectively. Lastly, the computer vision measurement system measured the outer diameter (D) and wall thickness (δ) with mean percentage error of 2.42% to 5.44% and 3.23% to 5.24%, respectively. Integrating these computer vision measurements into the trained ANN, a performance of 12.55 and 9.24 was observed for RMSE and MAE, respectively. With the identified IPs and its implementation with computer vision, the study contributes to the selection process of bamboo by having an inference of its mechanical properties considering treatment methods.

Abstract Format

html

Abstract (Filipino)

Ang lumalaking problema mula sa pagkasira ng kapaligiran at nakakapinsalang epekto ng industriya ng konstruksyon ang nagpapaigting ng pangangailangan para sa likas-kayang mga materyales at pamamaraan. Ang Bambusa Blumeana ay kinikilala bilang isang likas-kayang materyales para sa konstruksyon na madaling mahanap sa Pilipinas sa murang halaga. Ang kanyang mekanikal na mga katangian ay inimbestigahan na isinasaalang-alang ang impluwensiya ng mga paraan ng treatment at indicating properties (IP). Ang mga paraan ng treatment na sakop ay gumagamit ng apoy at kemikal treatment na gumagamit ng Copper-Chrome-Boron (CCB). Ang mean compressive strength na naobserbahan ay 37.08 MPa, 39.42 MPa, at 40.62 MPa para sa walang treatment, apoy, at CCB, ayon sa pagkakabanggit. Para naman sa mean tensile strength, ang mga nasukat ay 85.18 MPa, 65.11 MPa, at 90.69 MPa para sa walang treatment, apoy, at CCB, ayon sa pagkakabanggit. Gamit ang regression analysis, ang naobserbahang indicating properties na may pinakamataas na predictability ayon sa compressive capacity ay linear mass sa 12% MC (q12) sinundan ng wall thickness (δ) at outer diameter (D). Ang nagawang artificial neural network (ANN) ay ginamit itong mga indicating properties bilang input at ang nasukat na root mean squared error (RMSE) at mean absolute error (MAE) ay 17.38 at 13.44, ayon sa pagkakabanggit. Ang nabuong computer vision measurement system ay sinukat ang outer diameter (D) at wall thickness (δ) na may mean percentage error na 2.42% hanggang 5.44% at 3.23% hanggang 5.24%, ayon sa pagkakabanggit. Ang pagsama ng computer vision sa nabuong ANN, ang naobserbahang RMSE at MAE ay 12.55 at 9.24, ayon sa pagkakabanggit. Mula sa mga kinilalang mga IP at paggamit ng computer vision, ang pananaliksik na ito ay nagbibigay ambag sa proseso ng pagpili ng kawayan sa pagkakaroon ng hinuha sa kanyang mekanikal na katangian na isinasaalang-alang ang mga paraan ng treatment.

Abstract Format

html

Language

English

Format

Electronic

Keywords

Bamboo construction

Upload Full Text

wf_yes

This document is currently not available here.

Share

COinS