Upgrade to Pro — share decks privately, control downloads, hide ads and more …




Andres Root

April 25, 2016

Other Decks in Technology


  1. None
  2. Santiago Aljure - Luis Bejarano - Christian Campos Juan Cantor

    - Andres Lujan - Karen Navarrete
  3. "We created a portable, scalable, software and hardware solution that

    classifies fish even if it is still alive."
  4. The classification is one of the most important factors in

    sustainable fishering. Currently, all process are do manually without an efficient technological support. Our system is involved not only in the acquisition of data, too in the automating process of classification on site. Introduction
  5. Traditional Collecting Data *Manual measurement of individual. *In sea the

    observation human or scientist imply costing valuable time or killing the fish. * Cameras REQUIRED SOLUTION * without human intervention and internet access. Key Steps 1. Image capture 2. Identification of fish species 3. Automated measurement 4. Append records with position, time, date 5. Record User/ Fishing Vessel/ Fishing Gear 6. Transer/ Upload of data when internet access available 7. A cache to store and export the data to fisheries management agencies and fisheries scientists PROBLEM STATEMENT
  6. What’s wrong with the fishing industry? There’s a serious issue

    in the fishing field related to the capture of fish, i.e. capturing fish before its reproduction period, capturing protected species. If this keeps on happening, long term ALL fishermen are going to lose profitability in their business. We could automate a crucial process for fisheries, the fishing classification. We noted that the amount of fish that needs to be classified is huge so we believe that using image acquisition could be useful and it could be automated as well. That would decrease the time of classification considerably and benefit all fishermen who use this solution. Is there a solution to this problem? YES!
  7. What it is? In order to fully automate the classification

    process for fisheries our solution is able to classify and physically separate fish (either dead or alive) and store relevant information about its features.
  8. 1. Recognition phase: We use a cellphone camera and a

    mobile app equipped with a computer vision system that detects specific features of the fish (i.e. color, length, fish fin) This information is sent to a computer. 2. Classification Phase: We implemented machine learning techniques and software to build a classifier that can make more accurate predictions about the fish species. For the physical classification we created two stages: the first stage works with water tubes that is able to separate fish keeping them alive. The second stage works with a conveyor belt and is capable of separating dead fish, it also works as a rectification phase of the first stage, this process is made with additional cameras. We built this system thinking about keeping alive and returning to the sea protected species, fish that is before its reproduction period or other protected animals that are not even fish. i.e. Turtles, Sharks.
  9. HARDWARE/ SOFTWARE SOLUTION Recognition phase We designed a 3-stage-solution that

    involves identifying by image analysis if the fish captured are ready (i.e. beyond their reproductive age). Those who are not eligible are sent back to the sea, the rest continue to a second stage. We use a cellphone camera and a mobile app equipped with a computer vision system that detects specific features of the fish (i.e. color, length, fish fin) This information is sent to a computer. In the second stage a set cameras make sure that the first analysis was right taking and processing more images. In the case of a mistake the system LEARNS so each time less mistakes are made in the future. The system is also able to take feedback from a worker so it can also learn from humans.
  10. The first stage (classification stage) is made of a recollection

    tank, piping and valves. The tank is filled by seawater so the fish captured remain alive. Then they are forced through a pipe where cameras are placed in order to do the analysis. After the pipe with cameras, they arrive to a 3-way valve system with opens 2 possible paths: Mature fish path or non-mature fish path. If a mature fish is detected, then the first path opens (or keeps opened). If a non- mature fish is detected, the second path is opened and lets the fish return to the sea. The second stage (rectification stage) is located in the mature fish path. More cameras analyse the fish classified into this stage and verifies that they were adequately chosen. Finally fish are stored. HARDWARE/PHYSICAL SOLUTION
  11. Analysis Users can log in into our webpage in order

    to look at their own fishing and other fisheries statistics.
  12. ACCOMPLISHMENTS THAT WE’RE PROUD OF We found a solution that

    not only to improves work conditions for fisheries and fishermen by creating a faster fishing selection process but it also keeps alive and returns to the sea the fish that is too young to be fished. We managed, in less that 48 hours, to develop a minimum-viable-product integrating software and hardware. We’re proud to say that this solution is applicable to other different industries and uses.
  13. WHAT WE LEARNED Interdisciplinary teamwork is huge in terms of

    problem solving. As the brainstorming process was being done, each discipline representant was arguing why each idea was strong or not from their point of view. This was crucial to consider not only one-point-of-view- solutions but holistic solutions, which are more robust.
  14. WHAT’S NEXT FOR FISH-CLASSIFIER? Physical prototypes are required to validate

    our first design and keep on redesigning until there’s a feasible and profitable solution. We consider it’s important to do this in collaboration with fishermen in order to receive qualified feedback from the future real users of our solution system. For the software, the next step is to build a more robust fish detector system. We consider important to include additional features and also a larger amount of data is needed to improve machine learning.
  15. Thanks from Bogotá!!