SRPNet

Instructions

1 . Take an image Stand on soil where you need to calculate soil cover. Open the cora app and click on upload image button to open the camera Hold the mobile camera 1m above ground level. Capture image

2 . Analysed image Calculated the soil cover percentages will show on top of the application. Application will be using color schemes to visualized the analysed data.

3 . Download image Click on the download button to download the images.

About..!

Quantify the soil cover in classes of soil,residue and plant using semantic segmentation is important researcher and agronomist. The application utilizes a The Seg (semantic segmentation) Net (Network) architecture , a well-known encoder-decoder based architecture model which vastly used in area of biomedicine and computer vision.

The model has been trained image set of 33000 images which are collected from the experimental plots in Kansas State University across Kansas. The trained model then extracted to JSON format at implemented in this web app with the help of tensorflow js.

The model takes an input of a digital image of a crop field and quantify the image and returns percentages of the soil cover. The user can view and download the classification of the analysed image.

References

Badrinarayanan, V., Kendall, A., & Cipolla, R. (2017). SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(12), 2481–2495. https://doi.org/10.1109/TPAMI.2016.2644615

Collaborators