3D-CNN with Multi-Scale Fusion for Tree Crown Segmentation and Species Classification

Natural secondary forests play a crucial role in global ecological security, climate change mitigation, and biodiversity conservation. However, accurately delineating individual tree crowns and identifying tree species in dense natural secondary forests remains a challenge. This study combines deep...

Full description

Saved in:
Bibliographic Details
Main Authors: Jiayao Wang, Zhen Zhen, Yuting Zhao, Ye Ma, Yinghui Zhao
Format: Article
Language:English
Published: MDPI AG 2024-12-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/16/23/4544
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849220439618879488
author Jiayao Wang
Zhen Zhen
Yuting Zhao
Ye Ma
Yinghui Zhao
author_facet Jiayao Wang
Zhen Zhen
Yuting Zhao
Ye Ma
Yinghui Zhao
author_sort Jiayao Wang
collection DOAJ
description Natural secondary forests play a crucial role in global ecological security, climate change mitigation, and biodiversity conservation. However, accurately delineating individual tree crowns and identifying tree species in dense natural secondary forests remains a challenge. This study combines deep learning with traditional image segmentation methods to improve individual tree crown detection and species classification. The approach utilizes hyperspectral, unmanned aerial vehicle laser scanning data, and ground survey data from Maoershan Forest Farm in Heilongjiang Province, China. The study consists of two main processes: (1) combining semantic segmentation algorithms (U-Net and Deeplab V3 Plus) with watershed transform (WTS) for tree crown detection (U-WTS and D-WTS algorithms); (2) resampling the original images to different pixel densities (16 × 16, 32 × 32, and 64 × 64 pixels) and inputting them into five 3D-CNN models (ResNet10, ResNet18, ResNet34, ResNet50, VGG16). For tree species classification, the MSFB combined with the CNN models were used. The results show that the U-WTS algorithm achieved a recall of 0.809, precision of 0.885, and an F-score of 0.845. ResNet18 with a pixel density of 64 × 64 pixels achieved the highest overall accuracy (OA) of 0.916, an improvement of 0.049 over the original images. After incorporating MSFB, the OA improved by approximately 0.04 across all models, with only a 6% increase in model parameters. Notably, the floating-point operations (FLOPs) of ResNet18 + MSFB were only one-eighth of those of ResNet18 with 64 × 64 pixels, while achieving similar accuracy (OA: 0.912 vs. 0.916). This framework offers a scalable solution for large-scale tree species distribution mapping and forest resource inventories.
format Article
id doaj-art-17a7d9f8aeb346b4af6d5acb752358cc
institution Kabale University
issn 2072-4292
language English
publishDate 2024-12-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-17a7d9f8aeb346b4af6d5acb752358cc2024-12-13T16:31:13ZengMDPI AGRemote Sensing2072-42922024-12-011623454410.3390/rs162345443D-CNN with Multi-Scale Fusion for Tree Crown Segmentation and Species ClassificationJiayao Wang0Zhen Zhen1Yuting Zhao2Ye Ma3Yinghui Zhao4School of Forestry, Key Laboratory of Sustainable Forest Ecosystem Management-Ministry of Education, Northeast Forestry University, Harbin 150040, ChinaSchool of Forestry, Key Laboratory of Sustainable Forest Ecosystem Management-Ministry of Education, Northeast Forestry University, Harbin 150040, ChinaSchool of Forestry, Key Laboratory of Sustainable Forest Ecosystem Management-Ministry of Education, Northeast Forestry University, Harbin 150040, ChinaSchool of Forestry, Key Laboratory of Sustainable Forest Ecosystem Management-Ministry of Education, Northeast Forestry University, Harbin 150040, ChinaSchool of Forestry, Key Laboratory of Sustainable Forest Ecosystem Management-Ministry of Education, Northeast Forestry University, Harbin 150040, ChinaNatural secondary forests play a crucial role in global ecological security, climate change mitigation, and biodiversity conservation. However, accurately delineating individual tree crowns and identifying tree species in dense natural secondary forests remains a challenge. This study combines deep learning with traditional image segmentation methods to improve individual tree crown detection and species classification. The approach utilizes hyperspectral, unmanned aerial vehicle laser scanning data, and ground survey data from Maoershan Forest Farm in Heilongjiang Province, China. The study consists of two main processes: (1) combining semantic segmentation algorithms (U-Net and Deeplab V3 Plus) with watershed transform (WTS) for tree crown detection (U-WTS and D-WTS algorithms); (2) resampling the original images to different pixel densities (16 × 16, 32 × 32, and 64 × 64 pixels) and inputting them into five 3D-CNN models (ResNet10, ResNet18, ResNet34, ResNet50, VGG16). For tree species classification, the MSFB combined with the CNN models were used. The results show that the U-WTS algorithm achieved a recall of 0.809, precision of 0.885, and an F-score of 0.845. ResNet18 with a pixel density of 64 × 64 pixels achieved the highest overall accuracy (OA) of 0.916, an improvement of 0.049 over the original images. After incorporating MSFB, the OA improved by approximately 0.04 across all models, with only a 6% increase in model parameters. Notably, the floating-point operations (FLOPs) of ResNet18 + MSFB were only one-eighth of those of ResNet18 with 64 × 64 pixels, while achieving similar accuracy (OA: 0.912 vs. 0.916). This framework offers a scalable solution for large-scale tree species distribution mapping and forest resource inventories.https://www.mdpi.com/2072-4292/16/23/4544hyperspectral imageconvolutional neural networkindividual tree crown segmentationtree species classification
spellingShingle Jiayao Wang
Zhen Zhen
Yuting Zhao
Ye Ma
Yinghui Zhao
3D-CNN with Multi-Scale Fusion for Tree Crown Segmentation and Species Classification
Remote Sensing
hyperspectral image
convolutional neural network
individual tree crown segmentation
tree species classification
title 3D-CNN with Multi-Scale Fusion for Tree Crown Segmentation and Species Classification
title_full 3D-CNN with Multi-Scale Fusion for Tree Crown Segmentation and Species Classification
title_fullStr 3D-CNN with Multi-Scale Fusion for Tree Crown Segmentation and Species Classification
title_full_unstemmed 3D-CNN with Multi-Scale Fusion for Tree Crown Segmentation and Species Classification
title_short 3D-CNN with Multi-Scale Fusion for Tree Crown Segmentation and Species Classification
title_sort 3d cnn with multi scale fusion for tree crown segmentation and species classification
topic hyperspectral image
convolutional neural network
individual tree crown segmentation
tree species classification
url https://www.mdpi.com/2072-4292/16/23/4544
work_keys_str_mv AT jiayaowang 3dcnnwithmultiscalefusionfortreecrownsegmentationandspeciesclassification
AT zhenzhen 3dcnnwithmultiscalefusionfortreecrownsegmentationandspeciesclassification
AT yutingzhao 3dcnnwithmultiscalefusionfortreecrownsegmentationandspeciesclassification
AT yema 3dcnnwithmultiscalefusionfortreecrownsegmentationandspeciesclassification
AT yinghuizhao 3dcnnwithmultiscalefusionfortreecrownsegmentationandspeciesclassification