Volume XLI-B3
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLI-B3, 931-935, 2016
https://doi.org/10.5194/isprs-archives-XLI-B3-931-2016
© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLI-B3, 931-935, 2016
https://doi.org/10.5194/isprs-archives-XLI-B3-931-2016
© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

  10 Jun 2016

10 Jun 2016

IMAGE LABELING FOR LIDAR INTENSITY IMAGE USING K-NN OF FEATURE OBTAINED BY CONVOLUTIONAL NEURAL NETWORK

Masaki Umemura1, Kazuhiro Hotta1, Hideki Nonaka2, and Kazuo Oda2 Masaki Umemura et al.
  • 1Meijo University, 1-501 Shiogamaguchi, Tempaku-ku, Nagoya 468-8502, Japan
  • 2Asia Air Survey Co.,Ltd., 1-2-2 Manpukuji, Asao-ku, Kawasaki, Kanagawa, Japan

Keywords: Image labeling, Convolutional neural network, K-nearest neighbour and LIDAR intensity image

Abstract. We propose an image labeling method for LIDAR intensity image obtained by Mobile Mapping System (MMS) using K-Nearest Neighbor (KNN) of feature obtained by Convolutional Neural Network (CNN). Image labeling assigns labels (e.g., road, cross-walk and road shoulder) to semantic regions in an image. Since CNN is effective for various image recognition tasks, we try to use the feature of CNN (Caffenet) pre-trained by ImageNet. We use 4,096-dimensional feature at fc7 layer in the Caffenet as the descriptor of a region because the feature at fc7 layer has effective information for object classification. We extract the feature by the Caffenet from regions cropped from images. Since the similarity between features reflects the similarity of contents of regions, we can select top K similar regions cropped from training samples with a test region. Since regions in training images have manually-annotated ground truth labels, we vote the labels attached to top K similar regions to the test region. The class label with the maximum vote is assigned to each pixel in the test image. In experiments, we use 36 LIDAR intensity images with ground truth labels. We divide 36 images into training (28 images) and test sets (8 images). We use class average accuracy and pixel-wise accuracy as evaluation measures. Our method was able to assign the same label as human beings in 97.8% of the pixels in test LIDAR intensity images.