Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

WoodYOLO: A Novel Object Detector for Wood Species Detection in Microscopic Images

Lars Nieradzik, Henrike Stephani, Jördis Sieburg-Rockel, Stephanie Helmling, Andrea Olbrich, Stephanie Wrage, Janis Keuper

TL;DR

This paper introduces WoodYOLO, a novel object detection algorithm specifically designed for microscopic wood fiber analysis that significantly outperforms state-of-the-art models and contributes to enhancing regulatory compliance, supporting sustainable forestry practices, and promoting biodiversity conservation efforts globally.

Abstract

Wood species identification plays a crucial role in various industries, from ensuring the legality of timber products to advancing ecological conservation efforts. This paper introduces WoodYOLO, a novel object detection algorithm specifically designed for microscopic wood fiber analysis. Our approach adapts the YOLO architecture to address the challenges posed by large, high-resolution microscopy images and the need for high recall in localization of the cell type of interest (vessel elements). Our results show that WoodYOLO significantly outperforms state-of-the-art models, achieving performance gains of 12.9% and 6.5% in F2 score over YOLOv10 and YOLOv7, respectively. This improvement in automated wood cell type localization capabilities contributes to enhancing regulatory compliance, supporting sustainable forestry practices, and promoting biodiversity conservation efforts globally.

WoodYOLO: A Novel Object Detector for Wood Species Detection in Microscopic Images

TL;DR

This paper introduces WoodYOLO, a novel object detection algorithm specifically designed for microscopic wood fiber analysis that significantly outperforms state-of-the-art models and contributes to enhancing regulatory compliance, supporting sustainable forestry practices, and promoting biodiversity conservation efforts globally.

Abstract

Wood species identification plays a crucial role in various industries, from ensuring the legality of timber products to advancing ecological conservation efforts. This paper introduces WoodYOLO, a novel object detection algorithm specifically designed for microscopic wood fiber analysis. Our approach adapts the YOLO architecture to address the challenges posed by large, high-resolution microscopy images and the need for high recall in localization of the cell type of interest (vessel elements). Our results show that WoodYOLO significantly outperforms state-of-the-art models, achieving performance gains of 12.9% and 6.5% in F2 score over YOLOv10 and YOLOv7, respectively. This improvement in automated wood cell type localization capabilities contributes to enhancing regulatory compliance, supporting sustainable forestry practices, and promoting biodiversity conservation efforts globally.

Paper Structure

This paper contains 22 sections, 6 equations, 4 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Materials and methods
  4. Architecture
  5. Loss Function
  6. Regression Loss
  7. Classification Loss
  8. Metric
  9. Additional Approaches
  10. Center Sampling and Multi-Positives
  11. Label Assignment
  12. Auxiliary Head Loss
  13. Anchor Boxes
  14. NMS-Free Detection
  15. Training Strategies
  16. ...and 7 more sections

Figures (4)

  • Figure 1: Microscope image of macerated hardwood cells including vessel elements. Blue boxes indicate the correctly localized vessel locations by WoodYOLO, the light purple box indicate one false negative and red boxes denote false positives that were not annotated by wood anatomists. WoodYOLO significantly speeds up the manual annotation process by automatically identifying hundreds of vessel elements.
  • Figure 2: Detection architecture based on YOLOv7-tiny wang2022yolov7trainablebagoffreebiessets. "c" = Convolution with BN and ReLU, "+" = Concatenation, "m" = MaxPooling, "u" = Upsampling, "b" = Concatenation Block, "o" = single convolution with 5 outputs (x, y, width, height, confidence). Orange denotes an output in the neck of the model, which is given to the head. There are in total three outputs.
  • Figure 3: The "b" concatenation block consists of convolutions of kernel size 3x3 and 1x1. Each convolution is followed by batch normalization and ReLU activation. The "+" means concatenation.
  • Figure 4: Comparison of predicted bounding boxes (blue) and ground truth boxes (green). A high IoU threshold can result in both predicted boxes being rated as errors. (A) The overlap is below 0.5. Due to incorrect annotations, the predicted bounding boxes are sometimes more accurate. (B) Imperfect prediction as the end of the object (vessel element) is not detected.