345 lines
12 KiB
Python
345 lines
12 KiB
Python
import os
|
|
import shutil
|
|
import time
|
|
from fastapi import FastAPI, File, UploadFile,HTTPException ,Request
|
|
# from fastapi.responses import FileResponse
|
|
from fastapi.middleware.cors import CORSMiddleware
|
|
|
|
import cv2
|
|
import numpy as np
|
|
import mediapipe as mp
|
|
import matplotlib.pyplot as plt
|
|
import json
|
|
|
|
|
|
# First step is to initialize the Hands class an store it in a variable
|
|
mp_hands = mp.solutions.hands
|
|
|
|
# Now second step is to set the hands function which will hold the landmarks points
|
|
hands = mp_hands.Hands(static_image_mode=True, max_num_hands=2, min_detection_confidence=0.3)
|
|
|
|
# Last step is to set up the drawing function of hands landmarks on the image
|
|
mp_drawing = mp.solutions.drawing_utils
|
|
|
|
app = FastAPI()
|
|
|
|
app.add_middleware(
|
|
CORSMiddleware,
|
|
allow_origins=["*"],
|
|
allow_credentials=True,
|
|
allow_methods=["*"],
|
|
allow_headers=["*"],
|
|
)
|
|
|
|
|
|
@app.get("/")
|
|
def read_root():
|
|
return {"Hello": "World"}
|
|
|
|
@app.post("/")
|
|
async def image(image: UploadFile = File(...)):
|
|
content_type = image.content_type
|
|
print(content_type)
|
|
if(content_type != "image/jpeg" and content_type != "image/png" and content_type != "image/jpg") :
|
|
raise HTTPException(status_code=404, detail="Fail bukan foto")
|
|
image_name = image.filename
|
|
with open("temp/"+image_name, "wb") as buffer:
|
|
shutil.copyfileobj(image.file, buffer)
|
|
time.sleep(2)
|
|
sample_img = cv2.imread("temp/"+image_name)
|
|
sample_img = cv2.flip(sample_img, 1)
|
|
results = hands.process(cv2.cvtColor(sample_img, cv2.COLOR_BGR2RGB))
|
|
if not results.multi_hand_landmarks:
|
|
if os.path.exists("temp/"+image_name):
|
|
os.remove("temp/"+image_name)
|
|
raise HTTPException(status_code=404, detail="Foto harus ada telapak tangan")
|
|
|
|
if len(results.multi_handedness) > 1:
|
|
if os.path.exists("temp/"+image_name):
|
|
os.remove("temp/"+image_name)
|
|
raise HTTPException(status_code=404, detail="Hanya satu telapak tangan yang bisa diramal")
|
|
|
|
tangan = results.multi_handedness[0].classification[0].label
|
|
if(tangan == 'Right'):
|
|
if os.path.exists("temp/"+image_name):
|
|
os.remove("temp/"+image_name)
|
|
raise HTTPException(status_code=404, detail="Hanya Tangan Kiri Yang Bisa Diramal")
|
|
|
|
thumb = None
|
|
pinky = None
|
|
for hand_no, hand_landmarks in enumerate(results.multi_hand_landmarks):
|
|
thumb = hand_landmarks.landmark[mp_hands.HandLandmark.THUMB_TIP].x
|
|
pinky = hand_landmarks.landmark[mp_hands.HandLandmark.PINKY_TIP].x
|
|
cek_tapak_tangan = thumb > pinky
|
|
if(cek_tapak_tangan == False):
|
|
if os.path.exists("temp/"+image_name):
|
|
os.remove("temp/"+image_name)
|
|
raise HTTPException(status_code=404, detail="Sila foto telapak tangan kiri anda")
|
|
|
|
# if os.path.exists("temp/"+image_name):
|
|
# os.remove("temp/"+image_name)
|
|
|
|
change_background_mp = mp.solutions.selfie_segmentation #untuk hapus background
|
|
change_bg_segment = change_background_mp.SelfieSegmentation() #untuk hapus background
|
|
|
|
sample_img1 = sample_img
|
|
sample_img = cv2.cvtColor(sample_img, cv2.COLOR_BGR2RGB)
|
|
sample_img = change_bg_segment.process(sample_img)
|
|
sample_img = sample_img.segmentation_mask > 0.9
|
|
sample_img = np.dstack((sample_img,sample_img,sample_img))
|
|
sample_img = np.where(sample_img, sample_img1, 255)
|
|
sample_img = cv2.resize(sample_img, (350,450), interpolation = cv2.INTER_AREA)
|
|
shape = sample_img.shape
|
|
results = hands.process(cv2.cvtColor(sample_img, cv2.COLOR_BGR2RGB))
|
|
image_height, image_width, _ = sample_img.shape
|
|
|
|
|
|
result=None
|
|
print(results.multi_hand_landmarks)
|
|
for hand_landmarks in results.multi_hand_landmarks:
|
|
annotated_image = sample_img.copy()
|
|
palm_center_y = (hand_landmarks.landmark[mp_hands.HandLandmark.THUMB_CMC].y +
|
|
hand_landmarks.landmark[mp_hands.HandLandmark.PINKY_MCP].y)/2.1
|
|
palm_center_x = (hand_landmarks.landmark[mp_hands.HandLandmark.THUMB_CMC].x +
|
|
hand_landmarks.landmark[mp_hands.HandLandmark.PINKY_MCP].x)/2.1
|
|
myradius = int(image_height/4.9)
|
|
#annotated_image = makeCircle(annotated_image,palm_center_y,palm_center_x,myradius)
|
|
y = int(palm_center_y * image_height)
|
|
x = int(palm_center_x * image_width)
|
|
circle_coordinates = (x,y)
|
|
mask = np.zeros(sample_img.shape, dtype=np.uint8)
|
|
cv2.circle(mask, circle_coordinates, myradius, (255,255,255), -1)
|
|
ROI = cv2.bitwise_and(sample_img, mask)
|
|
mask = cv2.cvtColor(mask, cv2.COLOR_BGR2GRAY)
|
|
x,y,w,h = cv2.boundingRect(mask)
|
|
result = ROI[y:y+h,x:x+w]
|
|
mask = mask[y:y+h,x:x+w]
|
|
result[mask==0] = (255,255,255)
|
|
|
|
cv2.imwrite("temp/"+image_name+"cropped.png", result)
|
|
cv2.waitKey(0)
|
|
|
|
sample_img = cv2.imread("temp/"+image_name+"cropped.png")
|
|
width = 450
|
|
height = 450
|
|
dim = (width, height)
|
|
sample_img = cv2.resize(sample_img, dim, interpolation = cv2.INTER_AREA)
|
|
# cv2.imshow("palm",image) #to view the palm in python
|
|
# cv2.waitKey(0)
|
|
gray = cv2.cvtColor(sample_img,cv2.COLOR_BGR2GRAY)
|
|
edges = cv2.Canny(gray,25,45,apertureSize = 3)
|
|
# cv2.imshow("edges in palm",edges)
|
|
# cv2.waitKey(0)
|
|
edges = cv2.bitwise_not(edges)
|
|
# cv2.imshow("edges in palm1",edges)
|
|
cv2.imwrite("temp/"+image_name+"lines.png", edges)
|
|
cv2.waitKey(0)
|
|
|
|
TARGET_FILE = "temp/"+image_name+"lines.png"
|
|
IMG_DIR = os.path.abspath(os.path.dirname(__file__)) + '/images/'
|
|
IMG_SIZE = (200, 200)
|
|
|
|
target_img = cv2.imread(TARGET_FILE)
|
|
target_img = cv2.resize(target_img, IMG_SIZE)
|
|
|
|
print('TARGET_FILE: %s' % (TARGET_FILE))
|
|
|
|
bf = cv2.BFMatcher(cv2.NORM_HAMMING)
|
|
detector = cv2.AKAZE_create()
|
|
(target_kp, target_des) = detector.detectAndCompute(target_img, None)
|
|
|
|
hasil_ramalan = None
|
|
|
|
datas = None
|
|
|
|
with open('dataset.json') as f:
|
|
datas = json.load(f)
|
|
|
|
pilihan = 2000
|
|
image_ramalan = None
|
|
|
|
files = os.listdir(IMG_DIR)
|
|
for file in files:
|
|
|
|
comparing_img_path = IMG_DIR + file
|
|
try:
|
|
comparing_img = cv2.imread(comparing_img_path, cv2.IMREAD_GRAYSCALE)
|
|
comparing_img = cv2.resize(comparing_img, IMG_SIZE)
|
|
(comparing_kp, comparing_des) = detector.detectAndCompute(comparing_img, None)
|
|
matches = bf.match(target_des, comparing_des)
|
|
dist = [m.distance for m in matches]
|
|
ret = sum(dist) / len(dist)
|
|
except cv2.error:
|
|
ret = 100000
|
|
|
|
print(file, ret)
|
|
if(ret < pilihan):
|
|
pilihan =ret
|
|
image_ramalan = file
|
|
|
|
print(image_ramalan)
|
|
theindex = None
|
|
for index,data in enumerate(datas):
|
|
if(data["id"] == image_ramalan):
|
|
theindex= index
|
|
|
|
hasil_ramalan = datas[theindex]['datanya']
|
|
if os.path.exists("temp/"):
|
|
os.remove("temp/"+image_name+"lines.png")
|
|
os.remove("temp/"+image_name+"cropped.png")
|
|
os.remove("temp/"+image_name)
|
|
|
|
return {"message": hasil_ramalan}
|
|
|
|
# return {"message": "hasil_ramalan"}
|
|
|
|
|
|
# @app.post("/ramalan")
|
|
# async def image( request: Request):
|
|
|
|
# body = await request.form()
|
|
# if body:
|
|
# print(body['image'].content_type)
|
|
# return {"filename": "image.filename"}
|
|
# else :
|
|
# raise HTTPException(status_code=404, detail="error")
|
|
|
|
def makeCircle(img,circle_y,circle_x,radius):
|
|
image_height, image_width, _ = img.shape
|
|
y = int(circle_y * image_height)
|
|
x = int(circle_x * image_width)
|
|
circle_coordinates = (x,y)
|
|
color = (255, 0, 0)
|
|
thickness = 2
|
|
return cv2.circle(img,circle_coordinates,radius,color,thickness)
|
|
|
|
@app.post("/ramalan")
|
|
async def ramalan(request: Request):
|
|
body = await request.form()
|
|
if(body == False) :
|
|
raise HTTPException(status_code=404, detail="error")
|
|
|
|
if(body['image'] == None and body['image'] == '' ):
|
|
raise HTTPException(status_code=404, detail="error")
|
|
|
|
# print(body['image'])
|
|
image_file = body['image'] #ini file yang akan digunakan
|
|
image_src = "temp/"+body['image'] #ini file yang akan digunakan
|
|
size = len(image_file)
|
|
image_name = image_file[:size - 4]
|
|
|
|
change_background_mp = mp.solutions.selfie_segmentation #untuk hapus background
|
|
change_bg_segment = change_background_mp.SelfieSegmentation() #untuk hapus background
|
|
|
|
with mp_hands.Hands(
|
|
static_image_mode=True,
|
|
max_num_hands=1,
|
|
min_detection_confidence=0.5) as hands:
|
|
|
|
image = cv2.flip(cv2.imread(image_src), 1)
|
|
sample_img = cv2.flip(cv2.imread(image_src), 1)
|
|
|
|
|
|
image = cv2.cvtColor(sample_img, cv2.COLOR_BGR2RGB)
|
|
image = change_bg_segment.process(image)
|
|
image = image.segmentation_mask > 0.9
|
|
image = np.dstack((image,image,image))
|
|
image = np.where(image, sample_img, 255)
|
|
|
|
shape = image.shape
|
|
|
|
results = hands.process(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
|
|
image_height, image_width, _ = image.shape
|
|
|
|
for hand_landmarks in results.multi_hand_landmarks:
|
|
palm_center_y = (hand_landmarks.landmark[mp_hands.HandLandmark.THUMB_CMC].y +
|
|
hand_landmarks.landmark[mp_hands.HandLandmark.PINKY_MCP].y)/2.1
|
|
palm_center_x = (hand_landmarks.landmark[mp_hands.HandLandmark.THUMB_CMC].x +
|
|
hand_landmarks.landmark[mp_hands.HandLandmark.PINKY_MCP].x)/2.1
|
|
myradius = int(image_height/4.9)
|
|
y = int(palm_center_y * image_height)
|
|
x = int(palm_center_x * image_width)
|
|
circle_coordinates = (x,y)
|
|
mask = np.zeros(image.shape, dtype=np.uint8)
|
|
cv2.circle(mask, circle_coordinates, myradius, (255,255,255), -1)
|
|
ROI = cv2.bitwise_and(image, mask)
|
|
mask = cv2.cvtColor(mask, cv2.COLOR_BGR2GRAY)
|
|
x,y,w,h = cv2.boundingRect(mask)
|
|
result = ROI[y:y+h,x:x+w]
|
|
mask = mask[y:y+h,x:x+w]
|
|
result[mask==0] = (255,255,255)
|
|
cv2.imwrite("temp/"+image_name+"cropped.png", result)
|
|
cv2.waitKey(0)
|
|
|
|
image = cv2.imread("temp/"+image_name+"cropped.png")
|
|
width = 450
|
|
height = 450
|
|
dim = (width, height)
|
|
image = cv2.resize(image, dim, interpolation = cv2.INTER_AREA)
|
|
# cv2.imshow("palm",image) #to view the palm in python
|
|
# cv2.waitKey(0)
|
|
gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
|
|
edges = cv2.Canny(gray,25,45,apertureSize = 3)
|
|
# cv2.imshow("edges in palm",edges)
|
|
# cv2.waitKey(0)
|
|
edges = cv2.bitwise_not(edges)
|
|
# cv2.imshow("edges in palm1",edges)
|
|
cv2.imwrite("temp/"+image_name+"lines.png", edges)
|
|
cv2.waitKey(0)
|
|
|
|
TARGET_FILE = "temp/"+image_name+"lines.png"
|
|
IMG_DIR = os.path.abspath(os.path.dirname(__file__)) + '/images/'
|
|
IMG_SIZE = (200, 200)
|
|
|
|
target_img = cv2.imread(TARGET_FILE)
|
|
target_img = cv2.resize(target_img, IMG_SIZE)
|
|
|
|
print('TARGET_FILE: %s' % (TARGET_FILE))
|
|
|
|
bf = cv2.BFMatcher(cv2.NORM_HAMMING)
|
|
detector = cv2.AKAZE_create()
|
|
(target_kp, target_des) = detector.detectAndCompute(target_img, None)
|
|
|
|
hasil_ramalan = None
|
|
|
|
datas = None
|
|
|
|
with open('dataset.json') as f:
|
|
datas = json.load(f)
|
|
|
|
pilihan = 2000
|
|
image_ramalan = None
|
|
|
|
files = os.listdir(IMG_DIR)
|
|
for file in files:
|
|
|
|
comparing_img_path = IMG_DIR + file
|
|
try:
|
|
comparing_img = cv2.imread(comparing_img_path, cv2.IMREAD_GRAYSCALE)
|
|
comparing_img = cv2.resize(comparing_img, IMG_SIZE)
|
|
(comparing_kp, comparing_des) = detector.detectAndCompute(comparing_img, None)
|
|
matches = bf.match(target_des, comparing_des)
|
|
dist = [m.distance for m in matches]
|
|
ret = sum(dist) / len(dist)
|
|
except cv2.error:
|
|
ret = 100000
|
|
|
|
print(file, ret)
|
|
if(ret < pilihan):
|
|
pilihan =ret
|
|
image_ramalan = file
|
|
|
|
print(image_ramalan)
|
|
theindex = None
|
|
for index,data in enumerate(datas):
|
|
if(data["id"] == image_ramalan):
|
|
theindex= index
|
|
|
|
hasil_ramalan = datas[theindex]['datanya']
|
|
if os.path.exists("temp/"):
|
|
os.remove("temp/"+image_name+"lines.png")
|
|
os.remove("temp/"+image_name+"cropped.png")
|
|
os.remove("temp/"+body['image'])
|
|
|
|
return {"message": hasil_ramalan}
|
|
|