PSegNet: Simultaneous Semantic and Instance Segmentation for Point Clouds of Plants

<div class="floats-full-page floats-full-page-other"><h2>Table 6</h2><div class="floats-full-page-caption"><div>Quantitative comparison of instance segmentation performances of six networks including PSegNet.</div></div><div class="floats-full-page-content"><table class="table-group" id="tab6"><tr><td><table class="table"><tr><td class="thead-hr" colspan="6"><hr /></td></tr><tr class="thead"><td class="align_left"></td><td class="align_center">Methods</td><td class="align_center">Tobacco leaf</td><td class="align_center">Tomato leaf</td><td class="align_center">Sorghum leaf</td><td class="align_center">Mean</td></tr><tr><td class="thead-hr" colspan="6"><hr /></td></tr><tr><td class="align_left" rowspan="3">mPrec (%)</td><td class="align_center">ASIS</td><td class="align_center">78.54</td><td class="align_center">80.21</td><td class="align_center">79.04</td><td class="align_center">79.26</td></tr><tr><td class="align_center">PlantNet</td><td class="align_center">87.74</td><td class="align_center">85.50</td><td class="align_center">79.39</td><td class="align_center">84.21</td></tr><tr><td class="align_center">PSegNet (ours)</td><td class="align_center"><b>87.80</b></td><td class="align_center"><b>89.92</b></td><td class="align_center"><b>86.68</b></td><td class="align_center"><b>88.13</b></td></tr><tr><td class="align_center" colspan="6"><hr /></td></tr><tr><td class="align_left" rowspan="3">mRec (%)</td><td class="align_center">ASIS</td><td class="align_center">56.27</td><td class="align_center">64.84</td><td class="align_center">72.08</td><td class="align_center">64.40</td></tr><tr><td class="align_center">PlantNet</td><td class="align_center">69.36</td><td class="align_center">76.63</td><td class="align_center">81.83</td><td class="align_center">75.94</td></tr><tr><td class="align_center">PSegNet (ours)</td><td class="align_center"><b>77.09</b></td><td class="align_center"><b>78.62</b></td><td class="align_center"><b>82.13</b></td><td class="align_center"><b>79.28</b></td></tr><tr><td class="align_center" colspan="6"><hr /></td></tr><tr><td class="align_left" rowspan="3">mCov (%)</td><td class="align_center">ASIS</td><td class="align_center">62.88</td><td class="align_center">76.61</td><td class="align_center">74.26</td><td class="align_center">71.25</td></tr><tr><td class="align_center">PlantNet</td><td class="align_center">71.98</td><td class="align_center">83.34</td><td class="align_center">82.63</td><td class="align_center">79.32</td></tr><tr><td class="align_center">PSegNet (ours)</td><td class="align_center"><b>80.88</b></td><td class="align_center"><b>84.89</b></td><td class="align_center"><b>84.27</b></td><td class="align_center"><b>83.35</b></td></tr><tr><td class="align_center" colspan="6"><hr /></td></tr><tr><td class="align_left" rowspan="3">mWCov (%)</td><td class="align_center">ASIS</td><td class="align_center">73.95</td><td class="align_center">82.73</td><td class="align_center">77.31</td><td class="align_center">78.00</td></tr><tr><td class="align_center">PlantNet</td><td class="align_center">84.83</td><td class="align_center">89.48</td><td class="align_center">85.68</td><td class="align_center">86.66</td></tr><tr><td class="align_center">PSegNet (ours)</td><td class="align_center"><b>90.72</b></td><td class="align_center"><b>90.45</b></td><td class="align_center"><b>87.46</b></td><td class="align_center"><b>89.54</b></td></tr><tr class="table-tr"><td colspan="6"><hr class="tbody-hr" /></td></tr></table></td></tr><tr class="table-fn"><td><div>The best values are in boldface.<br /></div></td></tr></table></div></div>

Table 6 | PSegNet: Simultaneous Semantic and Instance Segmentation for Point Clouds of Plants

title

Table 6: PSegNet: Simultaneous Semantic and Instance Segmentation for Point Clouds of Plants

description

Table 6

About this journal

<p>The open access journal <em>Plant Phenomics</em>, published in association with NAU, publishes novel research that advances plant phenotyping and connects phenomics with other research domains.</p>


Journal profile

Meet the editorial board

<p><em>Plant Phenomics'</em> editorial board is led by Seishi Ninomiya (University of Tokyo), Frédéric Baret (French National Research Institute for Agriculture, Food and Environment), and Zong-Ming Cheng (Nanjing Agricultural University/University of Tennessee) and is comprised of leading experts in the field.</p>


Editorial board

More special issue details

<p>Open to submit: <strong><a href="https://spj.sciencemag.org/plant-phenomics-special-issues/plant-functional-phenotyping/">Plant Functional Phenotyping</a></strong>; <strong><a href="https://spj.sciencemag.org/plant-phenomics-special-issues/time-series-phenotyping/">Advances in Proximal Sensing and Image Analysis for Time-Series Phenotyping</a></strong>​​​​​; <strong><a href="https://spj.sciencemag.org/plant-phenomics-special-issues/pma2022/">PMA2022</a></strong>;&nbsp;<strong><a href="https://spj.sciencemag.org/plant-phenomics-special-issues/object-detection/">Object Detection and Image Segmentation for Plant Phenotyping</a></strong></p>

<p>&nbsp;</p>

<p>Read our <a href="https://spj.sciencemag.org/plant-phenomics-special-issues/#published-special-issues">published special issues</a>!</p>


Special issues

Author guidelines

Databases and indexing

Plant Phenomics | Science Partner Journal

The open access journal Plant Phenomics, published in association with NAU, publishes novel research that advances plant phenotyping and connects phenomics with other research domains.