Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

PlantCamo: Plant Camouflage Detection

Jinyu Yang, Qingwei Wang, Feng Zheng, Peng Chen, Aleš Leonardis, Deng-Ping Fan

Abstract

Camouflaged Object Detection (COD) aims to detect objects with camouflaged properties. Although previous studies have focused on natural (animals and insects) and unnatural (artistic and synthetic) camouflage detection, plant camouflage has been neglected. However, plant camouflage plays a vital role in natural camouflage. Therefore, this paper introduces a new challenging problem of Plant Camouflage Detection (PCD). To address this problem, we introduce the PlantCamo dataset, which comprises 1,250 images with camouflaged plants representing 58 object categories in various natural scenes. To investigate the current status of plant camouflage detection, we conduct a large-scale benchmark study using 20+ cutting-edge COD models on the proposed dataset. Due to the unique characteristics of plant camouflage, including holes and irregular borders, we developed a new framework, named PCNet, dedicated to PCD. Our PCNet surpasses performance thanks to its multi-scale global feature enhancement and refinement. Finally, we discuss the potential applications and insights, hoping this work fills the gap in fine-grained COD research and facilitates further intelligent ecology research. All resources will be available on https://github.com/yjybuaa/PlantCamo.

PlantCamo: Plant Camouflage Detection

Abstract

Camouflaged Object Detection (COD) aims to detect objects with camouflaged properties. Although previous studies have focused on natural (animals and insects) and unnatural (artistic and synthetic) camouflage detection, plant camouflage has been neglected. However, plant camouflage plays a vital role in natural camouflage. Therefore, this paper introduces a new challenging problem of Plant Camouflage Detection (PCD). To address this problem, we introduce the PlantCamo dataset, which comprises 1,250 images with camouflaged plants representing 58 object categories in various natural scenes. To investigate the current status of plant camouflage detection, we conduct a large-scale benchmark study using 20+ cutting-edge COD models on the proposed dataset. Due to the unique characteristics of plant camouflage, including holes and irregular borders, we developed a new framework, named PCNet, dedicated to PCD. Our PCNet surpasses performance thanks to its multi-scale global feature enhancement and refinement. Finally, we discuss the potential applications and insights, hoping this work fills the gap in fine-grained COD research and facilitates further intelligent ecology research. All resources will be available on https://github.com/yjybuaa/PlantCamo.

Paper Structure

This paper contains 21 sections, 4 equations, 11 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. COD in Biology & Evolution
  4. COD in Computer Vision
  5. Dataset Construction
  6. Data Collection
  7. Data Annotation
  8. Data Statistics
  9. Benchmark Study
  10. Benchmark Settings
  11. Benchmarking SoTA COD Models
  12. Discussion
  13. Method
  14. Multi-scale Global Feature Enhancement
  15. Multi-scale Feature Refinement
  16. ...and 6 more sections

Figures (11)

  • Figure 1: Visualized examples and corresponding mask annotations in PlantCamo dataset. Our dataset covers different kinds of plant camouflage: (a) background matching; (b) disruptive coloration; (c) masquerade; and (d) decoration.
  • Figure 2: Annotated examples in the proposed PlantCamo. For each image, we offer different annotations, which include image-level attributes (1st row), bounding boxes (2nd row), object annotation (3rd row), and instance annotation (4th row). Zoom in for details.
  • Figure 3: Histogram distribution for the plant categories in PlantCamo. We visualize some representative camouflaged plants.
  • Figure 4: Dataset statistics of the proposed PlantCamo. BM = background matching, DC = disruptive coloration, MQ = masquerade, DR = decoration, MO = multiple objects, SC = shape complexity, OC = occlusion, BO = big object, SO = small object, and OV = out-of-view.
  • Figure 5: Image resolution distribution.
  • ...and 6 more figures