## Algorithme KNN - Reconnaissance de panneaux

def Distance_uv(u,v):
    n = len(u)
    Dst = 0
    for i in range(n):
        di = u[i]-v[i]
        Dst += di**2
    Dst = Dst**(1/2)
    return Dst

def Distance(u,Lv):
    Ld = []
    for i in range(len(Lv)):
        v = Lv[i]
        Dst = Distance_uv(v,u)
        Res = [Dst,i]
        Ld.append(Res)
    return Ld

def Proches(u,Lv,k):
    Ld = Distance(u,Lv)
    Ld.sort()
    Res = Ld[:k]
    return Res

## Lecture des images

# Affichage

import matplotlib.pyplot as plt
plt.close('all')

def Affiche(image):
    plt.figure()
    plt.imshow(image)
    plt.axis('off')
    plt.show()
    plt.pause(0.00001)

# Lecture des images

def Lecture(Chemin):
    Image = plt.imread(Chemin)
    return Image

## Fonctions d'analyse des images

def Analyse(Image):
    Nl,Nc = Image.shape[0:2]
    N = Nl * Nc
    L_RGB = []
    for l in range(Nl):
        for c in range(Nc):
            R,G,B = Image[l,c]
            R = float(R)
            G = float(G)
            B = float(B)
            L_RGB += [R,G,B]
    return L_RGB

''' Vérification
len(Analyse(Lecture('Sources\\0\\0.bmp')))
'''

def Analyse_Globale(L_Chemin):
    Res = []
    N = len(L_Chemin)
    for i in range(len(L_Chemin)):
        Chemin = L_Chemin[i]
        print("Apprentissage image",i+1,"sur",N)
        Image = Lecture(Chemin)
        Image = Image[:,:,:3] # Si RGBA
        Analyse_im = Analyse(Image)
        Res.append(Analyse_im)
    return Res

## Création de la base des données

# Ouverture des images sources

Sources = "Sources\\"
Dossiers = [0,1,2,3,4,5,6,7]
Nb_Dossiers = len(Dossiers)
Nb_Images_Dossiers = [5,5,5,5,5,5,5,5]

Liste_Chemin = []
Liste_Dossier = []
Liste_Num = []
for d in Dossiers:
    Nb_Im = Nb_Images_Dossiers[d]
    for im in range(Nb_Im):
        Chemin = Sources + str(d) + "\\" + str(im) + ".bmp"
        Liste_Chemin.append(Chemin)
        Liste_Dossier.append(d)
        Liste_Num.append(im)

# Analyse des images

Donnees = Analyse_Globale(Liste_Chemin)