The Japanese government has put forth its "Space Industry Vision 2030" in hopes of expanding applicability beyond the space equipment industry to other relevant industries as well, with the aim of fast-paced doubling of the overall space industry's size by 2030. Amidst these circumstances, the Ministry of Economy, Trade and Industry launched Japan’s first satellite data platform “Tellus” intended for industry use on February 21, 2019. Tellus facilitates the use of traditionally difficult-to-handle satellite data by everyone from private companies, universities and research institutions through to individuals while also aiding in the creation of businesses in a broad range of industries and business categories made possible with this free access to space data.

The Tellus Satellite Challenge is a data analysis contest aimed at promoting the use of Tellus, namely in the form of producing a visual representation of satellite data use case models, uncovering outstanding human resources specializing in data analysis, and disseminating information and educating others on the types of satellite data and its various formats. This is the third time the event has been held.

The theme of this third event is the “detection of areas of sea ice using SAR data.”

The First Regional Coast Guard Headquarters currently provides information on sea ice formations to prevent maritime accidents caused by sea ice around Hokkaido. To do this, in addition to visual observations of sea ice from the air and patrol boats, the First Regional Coast Guard Headquarters also analyzes satellite observation images taken of sea ice.

Satellites are capable of capturing images that provides useful information about the sea surface over a broad area at a glance. Of these, synthetic aperture radar (SAR) data, which allows images to be captured at any time—night or day—and in any weather conditions, demands a particularly high skill set in order to interpret sea ice coverage. This contest challenges participants to develop an algorithm to detect areas of sea ice with a high degree of accuracy from SAR data.

Observation data from PALSAR-2 onboard the Advanced Land Observing Satellite 2 “Daichi-2” (ALOS-2) taken of the sea area around Hokkaido will be used for this contest. PALSAR-2 is a synthetic aperture radar (SAR) sensor that retrieves data by bouncing radio waves down to the Earth’s surface and receiving the radio waves that are reflected back. SAR is unique in that it can capture images regardless of the time of day and weather conditions. See here for more details.

This Tellus Satellite Challenge is conducted based on data supplied by the First Regional Coast Guard Headquarters and the Japan Aerospace Exploration Agency.

Task Description

Participants are tasked with predicting and detecting areas of sea ice from SAR data (HH polarization and HV polarization).

Data Description

・PALSAR-2 (©JAXA)

    Size (Height × Width): No set size

    No. of images:

         For learning: 80 HH polarized and HV polarized images (scenes) for a total of 160 images
         For evaluation: 40 HH polarized and HV polarized images (scenes) for a total of 80 images

    Observation mode: Broad area observation mode (ScanSAR) *Two patterns using modes (1) and (2)
 Mode (1)
     Polarization: HH, HV
     Area of observation: 350 km
     Pixel spacing (distance on the ground per pixel): 25 m
 Mode (2)
     Polarization: HH, HV
     Area of observation: 490 km
     Pixel spacing (distance on the ground per pixel): 25 m

    Further descriptions of polarization and the relationship between the typical features that need to be taken into account when detecting ice formations and polarization can be found on the “Polarization” tab.

    *While the original images are 16-bit images, the conversion process used converts the data to 8-bit images.

・Label images
    Format: Greyscale image

    Size (Height × Width): Same as the corresponding SAR data

    No. of images:
        For learning: 80 images

    Category details:
        0: Sea ice density 0 (no sea ice detected)
        1: Sea ice density 1–3
        4: Sea ice density 4–6
        7: Sea ice density 7–8
        9: Sea ice density 9–10
        11: Lake
        12: Land mass

        Each pixel value represents a category, and areas with a sea ice density of 1–10 (in pixel value terms, 1, 4, 7, 9) are interpreted as sea ice formations.

Evaluation Function

・“IoU” (intersection over union) is used as the evaluation function.
・An evaluation value between 0 and 1 will be assigned. The higher the accuracy, the higher the value will be.

Details

　(1) Calculate the overlap (IoU) of the prediction area and the correct area (at a pixel level) for certain scenes
　(2) Calculate the average of (1) for all scenes subject to evaluation

    *Set areas with a sea ice density of 1–10 as sea ice formations to be included in predictions. For more information on labels, refer to the description.txt file contained in the readme.zip file.
    *For more information on submission formats, refer to the figure.pdf and description.txt files contained in the readme.zip file provided.
    *The script (make_submit.py) used for creating submission files from prediction results (a binary image where sea ice formations show a pixel value of 1, and all other areas show a pixel value of 0) is included in the readme.zip file.

Determining Winners

1. The final ranking will be determined using evaluation values based on a correct data set that differs from that used during the contest. Score board rankings and scores will automatically change to a final evaluation when the contest ends.
2. In the event of a tie, the participant that submitted their entry first will be declared the higher ranking contestant.
3. Participants on top of the final rankings chart will be contacted by the secretariat as winner candidates.
4. Winner candidates will be required to submit the following on the condition that the submitting party guarantees that the secretariat shall bear no financial burden for such, and shall not be considered to have infringed on any rights.
　・Source code for model (or the equivalent of such)
　・Learning-complete model
　・Procedure description document enabling reproduction of the prediction process (must clearly explain preprocessing, learning, and prediction processes) The source (include details as to the location of the data, links to pertinent theses, etc.) of any data other than the submitted data, learning-complete models, APIs, etc., if used 
　・Computer/equipment operating environment information, including OS version, software utilized and analysis methods
　・Random number seed (if Random Forest or similar was used to generate random numbers)
　・Degree of model contribution for all explanatory variables (when using a method that enables calculation of degree of prediction)
　・Insights gained through data interpretation, manipulation and modeling
5. Winner candidates will be disqualified should any of the following apply to the winner candidates, or their submitted models, during the reproducibility verification period.
　・The contestant does not respond within the specified timeframe to contact from, or a request/demand from, the contest administrative office regarding contest-related procedures
　・The contestant fails to adhere to participant requirements, rules, etc.
　・The model's prediction results cannot be reproduced
　・Predictions cannot be made when using new data
　・Any other reason for which the secretariat deems the candidate unworthy of winning the contest
6. Winners will be selected from candidates for which reproducibility can be verified.

Contestant Attitude

・We ask that all contestants strive to achieve practical, useful approaches with top priority placed on achieving corporate goals, helping to solve social problems, sharing research results with the public, and other overarching goals.

System Usage

・Each person is limited to one account only. However, team-based participation is allowed as long as the one-account rule is followed.
・Persons wishing to participate as a team can do so provided that teams are made up of a maximum of five members, including the team representative.
    - When forming a team, click the “Create Team” button at the top right of the competition page to have the representative fill out the entry details.
    - Entry applications must be made by Monday November 4, 2019 to enter a team in this competition.
    - If the team creator (representative) cancels their SIGNATE membership, the team will be treated as being disbanded.
    - Be aware that when a team disbands, any user that participated in the team up to that point cannot submit another entry into this competition, either as part of another team, or as individuals.
    - See here for more details about team functions.

Handling of Contest-related Information

・Sharing of prediction-related data and source codes with other contestants is prohibited.

Data Usage

・Submission of manually labeled result data is prohibited.
・Pseudo-labeling evaluation data to use the evaluation data again for learning is prohibited.
・Rewriting of learning data labels and then using said data for model learning is prohibited. However, automated data augmentation via learning-use image data processing for use as learning data is allowed.
・Model learning using any data aside from that distributed to contestants is prohibited.
・Use of Google Cloud Vision API and other such APIs is prohibited. However, use of open-source learning-complete models, libraries and similar is allowed (source code publicly open and available to anyone, and not designed for commercial use; e.g., Tensorflow, Chainer, Keras, etc.). When using a learning-complete model, clearly cite the source using a hyperlink to relevant documentation or similar. 

Implementation Method

・Learning tools for models are limited to open-source, free tools (Python, R, etc.).
・Divide source code in three parts as shown below (preprocessing, learning and prediction), and carry out implementation for each.
        1. Preprocessing
            Module that reads supplied data, carries out preprocessing, and outputs each file in a format that can be input into the model. Divide into files for learning and files for evaluation ("get_train_data" and "get_test_data" or similar), and define a function for preprocessing. 
        2. Learning
            Module that reads files created in step 1 to enable model learning. Define a function that can output the learning-complete model, feature quantity importance level and cross validation assessment results.
        3. Prediction
            Module that reads the test data created in step 1 and model created in step 2 and outputs prediction results in file format.

Contest sponsor:

Japanese Ministry of Economy, Trade and Industry (METI)

Support:

SAKURA Internet Inc.

Present:

First Regional Coast Guard Headquarters
Japan Aerospace eXploration Agency

What is Polarization

    Polarization is an indicator of radio wave properties that represents the direction of oscillation of an electric field. HH/HV polarization represents the combination of the direction of polarization for sending/receiving (H = Horizontal, V = Vertical). HH indicates a combination of data sent and received as horizontally polarized waves, and HV indicates a combination of data sent as horizontally polarized waves and received as vertically polarized waves. As the reflective properties of radio waves varies based on the surface of the ground (forest, soil, water areas, etc.), different information can be captured based on the combination of polarization used. For more information, see Link.

Relationship Between the Typical Features that Need to be Taken into Account when Detecting Ice Formations and Polarization

First-year ice and multi-year ice

    First-year ice refers to newly formed ice, while multi-year ice that has developed over repeated formations over time. First-year ice has a high amount of salinity and has a high relative permittivity along the surface of the sea ice. This causes scattering across the sea ice surface (surface scattering), leading to a stronger HH polarization component. This results in these areas showing up brighter on HH polarization images. Conversely, multi-year ice has a low amount of salinity and has a high relative permittivity inside sea ice. This causes scattering inside the ice (volume scattering), leading to a stronger HV polarization component. This results in these areas showing up brighter on HV polarization images. The amount of salinity is inversely proportional to the thickness of the ice.

Surface coarseness

    Reflected radio waves get stronger the rougher the surface. This results in these areas appearing brighter in images.

Thawing

    The surface of the ice smooths out when it begins to thaw. This weakens the reflection of radio waves for both HH and HV components, making these areas appear darker in images. This has a particularly pronounced impact on HV polarization images.

Deposited snow

    Reflected radio waves get stronger the greater the snow fall. This strengthens HV polarization components, resulting in these areas showing up brighter on HV polarization images.

Potential for erroneous detection

    Strong winds strengthen the waves of the ocean, and also strengthen the radio waves reflecting off the sea surface. This results in the shape of the waves appearing brighter in image, and this is a reason for errors in detecting sea ice.

If you get the chance, take the time to review the SIGNATE “satellite image-based sea ice detection” competition, its objectives and points to consider when reviewing images.

【Additional Notes】
Use of ALOS-2 data:
When using ALOS-2 data, contestants must understand and adhere to the following.
(1) Regarding ALOS-2 data received from SIGNATE Inc., the intellectual property rights and all other rights for this data belong to the Japan Aerospace eXploration Agency (JAXA).
(2) Neither SIGNATE Inc. nor JAXA provides any guarantee regarding the quality of data supplied for this competition, and neither organization will take responsibility for results stemming from ALOS-2 data usage or for any damages incurred through use of said data.

【Matters to be updated】
(The provisions of paragraph 1 of Article 3 [Reward and Vesting of Rights] of the Terms below shall be amended as follows [including the surviving provisions].):
1. Unless otherwise provided for, any Participant shall satisfy the following requirements in order to be entitled to receive a reward in any Competition that offers a reward:
(1) To be a winner;
(2) To grant to the Host and its supporting companies (hereinafter referred to as the “Licensees”) a non-exclusive license without restriction as to term and territory and free of charge to use all rights necessary for modifying the Final Submissions and Submissions for Final Judgment for commercial or non-commercial purposes;
(3) To grant to the Licensees a non-exclusive license without restriction as to term and territory and free of charge to use all rights necessary for publishing and use, the Final Submissions and Submissions for Final Judgment and the Final Submissions and Submissions for Final Judgment as modified on the satellite data platform “Tellus” to be started to be provided in future (hereinafter referred to as “Tellus”) for commercial or non-commercial purposes, and for granting a sublicense to any contractor for the operation of Tellus and user of Tellus (together with the rights as prescribed in the preceding item, hereinafter collectively referred to as the “Rights”);
(4) To agree not to exercise moral rights to the Rights against the Licensees and their sub-licensees;
(5) To enter into agreements for the grant of license of the Rights with the Licensees including the consent to the preceding three (3) items and any other reasonable provisions;
(6) If the Host or any supporting company desires to receive an exclusive license or transfer of the rights contained in the Final Submissions and Submissions for Final Judgment, to accept the consultation about the agreement on such license or transfer;
(7) To have the personal identity of such Participant verified by the Company; and
(8) Not to breach any provision of these Terms and the Terms of Use.

(The provisions of Article 4 [Confidentiality] of the Terms below shall be amended as follows [including the surviving provisions].):
1. Any Participant shall treat any information and data that they receive from the Company in relation to each Competition (hereinafter referred to as the “Company-Provided Information”) as confidential information and shall not disclose the same to any third party and use the same for any purpose other than for such Competition and purpose specified by the Company separately; provided, however, that the confidential information shall not include any information that falls under any of the following items:
(1) Information that is known to the public at the time of the disclosure;
(2) Information that is already possessed by the Participant at the time of the disclosure (only in the case where such Participant may demonstrate such fact by reasonable means);
(3) Information that becomes known to the public without the fault of the Participant after the disclosure;
(4) Information that is independently developed by the Participant without reference to any information as disclosed; or
(5) Information that is rightfully disclosed by any third party having a right to do so without the obligations of confidentiality (only in the case where such Participant may demonstrate such fact by reasonable means).
2. Any Participant shall delete or return to the Company the Company-Provided Information immediately after the completion of each Competition.
3. If there is any separate arrangement in relation to the confidential information in each Competition, the provisions of such arrangement shall prevail over the provisions of these Terms.
4. If any dispute occurs between the Host or other third party and the Company due to the breach by any Participant of the provisions of this Article and such other party makes any claim against the Company, such Participant shall compensate for any damage, loss, expenses (including, but not limited to, attorneys’ fees), lost profits and lost revenues, etc. incurred by the Company.
5. The provisions of this Article shall survive the termination of the relevant Competition or the Participant’s completion of the procedures for withdrawal from the service of the Company, with respect to the Company-Provided Information and the Winner’s Final Submissions and Submissions for Final Judgment for a period of five (5) years thereafter.