Server API Documentation - Gachon-Graduation-work/Muzzle_Detection_Project GitHub Wiki

Function

serverAPI

load_image()

def load_img(path):

function to load image which comes from client and run YoloV3 network model and send it to post_process function.

  • parameter: path of image.
  • return: information about input picture if it is dangerous or not.

post_process() at yolo.py

def post_process(img, outputs, conf):

function which finds dogs inside the picture using YoloV3 result, if there is dog, run load_image2 function to find if there is muzzle or not.

  • parameter: image, result of YoloV3 network model, confidence.
  • return: result of all-processed image.

load_image2()

def load_img(path):

function to find out if there is muzzle or not by using YoloV3 network model we made. if there is no muzzle then run load_image3 function to classify if dog is fierce dog or not.

  • parameter: path of image
  • return: information about the dogs inside the image.

post_process() at muzzle_yolo.py

def post_pocess(img, outputs, conf):

function which finds muzzle from dog images.

  • parameter: image of dog, result of YoloV3 network model that we made, confidence.
  • return: result of analysis of dog images.

load_image3()

def load_img3(path):

function to classify if input dog image is fierce dog or not by using ResNet model we made. parameter: path of image of dog

  • parameter: path of image of dog
  • return: result of classification of dog image.