![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/capture-1-1024x576.jpg\")
<\/figure><\/div>\n\n\n\n<\/p>\n\n\n\n
L’appareil ne fait que pr\u00e9senter les donn\u00e9es sur un \u00e9cran LCD. Une cam\u00e9ra capture ces donn\u00e9es. Un programme, en Python, bas\u00e9 sur les r\u00e9seaux de neurones artificiels interpr\u00e8te l’affichage des segments clairs, les transformant en cha\u00eene de caract\u00e8res compatibles \u00e0 l’importation dans Excel.<\/p>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/20220227_210044c-1024x643.jpg\")
<\/figure><\/div>\n\n\n\n<\/p>\n\n\n\n
La cam\u00e9ra se place \u00e0 courte distance de d\u00e9tecteur, l’image ne doit pas \u00eatre d\u00e9form\u00e9e. Les cam\u00e9ras web communes produisent une image en barillet (bomb\u00e9e), ce qui n’est pas ad\u00e9quat. La cam\u00e9ra de marque ELP, mod\u00e8le ELP-USBFHD01M-MFV, plac\u00e9e \u00e0 15 cm devant le d\u00e9tecteur, g\u00e9n\u00e8re une image tr\u00e8s peu d\u00e9form\u00e9e, de bonne qualit\u00e9, facilement traitable. L’ouverture du diaphragme de la cam\u00e9ra s’op\u00e8re manuellement, ce qui offre un ajustement de la luminosit\u00e9 parfaite et stable.<\/p>\n\n\n\n
\n\n\n\n
\u00c9quipements<\/h3>\n\n\n\n
D\u00e9tecteur de qualit\u00e9 de l’air JSM-131 SE<\/strong><\/p>\n\n\n\n Cam\u00e9ra<\/strong><\/p>\n\n\n\n Pour s’initier au traitement par r\u00e9seau de neurones, je vous sugg\u00e8re de visionner le diaporama que j’ai produit sur ce sujet. Voir la partie \u00ab Simulation d’un r\u00e9seau de neurones\u00bb :<\/p>\n\n\n\n Nous allons simuler, par programmation, un r\u00e9seau de neurones pr\u00e9cabl\u00e9, fixe, configur\u00e9 pour ex\u00e9cuter une t\u00e2che bien particuli\u00e8re. Ce r\u00e9seau sera le plus simple possible, ayant une seule couche.<\/p>\n\n\n\n <\/p>\n\n\n\n Le traitement des images se fait \u00e0 partir de la biblioth\u00e8que de fonctionnalit\u00e9s d’OpenCV<\/a>.<\/p>\n\n\n\n Examinons, par \u00e9tape, la programmation<\/p>\n\n\n\n <\/p>\n\n\n\n <\/p>\n\n\n\n <\/p>\n\n\n\n Maintenant, il s’agit de situer la position des segments dans l’image. Nous d\u00e9coupons la fen\u00eatre en partie. Chaque partie correspond \u00e0 une valeur de la mesure; HCHO, TVOC, PM2.5, PM10, Co2. Nous cherchons \u00e0 quel pixel correspond le coin sup\u00e9rieur gauche du premier chiffre de la mesure. Par la m\u00e9thode essais-erreurs, nous trouvons pour HCHO, le pixel y=69, x= 185 que nous pouvons aussi d\u00e9finir comme ligne 69, colonne 185. <\/p>\n\n\n\n Par cet affichage, nous pouvons constater aussi que la cam\u00e9ra est bien align\u00e9e par rapport au d\u00e9tecteur. L’image ne penche pas, le haut des segments sont tous align\u00e9s sur l’horizontal.<\/p>\n\n\n\n Pour PM10, le pixel y=246, x= 185.<\/p>\n\n\n\n Pour Co2, le pixel y=305, x= 185.<\/p>\n\n\n\n Il reste \u00e0 trouver la position des points d\u00e9cimaux.<\/p>\n\n\n\n Pour le HCHO, le pixel 104, 220. Comme la valeur du HCHO sur cet appareil se situe entre 0 et 1.999, il n’y a qu’une position pour le point d\u00e9cimal.<\/p>\n\n\n\n Premi\u00e8re position du point d\u00e9cimal TVOC, pixel y=162, x=220 et pour la deuxi\u00e8me position, pixel y=162, x=272<\/p>\n\n\n\n Il n’y a pas de point d\u00e9cimal pour les autres mesures; PM2.5, PM10, Co2. La position en x du premier pixel de chaque \u00abdigit\u00bb est d\u00e9termin\u00e9e encore par essais-erreurs.<\/p>\n\n\n\n <\/p>\n\n\n\n Deux boucles sont imbriqu\u00e9es pour retrouver chaque \u00abdigit\u00bb de chaque nombre qui correspond \u00e0 une mesure.<\/p>\n\n\n\n Chaque \u00abdigit\u00bb est compos\u00e9 de 7 segments. Les segments sont allum\u00e9s selon le chiffre \u00e0 afficher.<\/p>\n\n\n\n Chaque segment est analys\u00e9 selon l’\u00e9tat de 4 pixels qui en font parti. Les segments sont nomm\u00e9s a,b,c,d,e,f,g.<\/p>\n\n\n\n Quatre pixels de chaque segment sont \u00e9valu\u00e9s pour \u00e9viter une fausse lecture si un pixel est brouill\u00e9 par une poussi\u00e8re, un probl\u00e8me de cam\u00e9ra, de transmission, etc. C’est l’avantage du r\u00e9seau de neurones, la lecture des chiffres est possible m\u00eame si une partie de l’information est erron\u00e9e.<\/strong><\/p>\n\n\n\n La position du pixel a1, du segment a, du premier nombre (HCH0), premier \u00abdigit\u00bb se calcul ainsi:<\/p>\n\n\n\n line<\/strong> = posNumber[0]<\/strong> \u00e9gal 69<\/strong> ; col<\/strong>=posDigit[0]<\/strong> \u00e9gal 185<\/strong><\/p>\n\n\n\n donc dans le mask[69+1,185+9], soit mask[ligne 70, colonne194]<\/p>\n\n\n\n <\/p>\n\n\n\n La valeur de i correspond au num\u00e9ro du neurone qui a la valeur maximum de sortie. Le chiffre du \u00abdigit\u00bb trait\u00e9 correspond au num\u00e9ro de neurone.<\/strong> Par exemple number[m,j]=i, si m=1, j=3 et i=7, number[1,3]=7, m=1 correspond \u00e0 TVOC, j \u00e0 la position du dernier \u00abdigit\u00bb, celui compl\u00e8tement \u00e0 droite, sa valeur affich\u00e9e est 7.<\/p>\n\n\n\n <\/p>\n\n\n\n Impression du nombre trouv\u00e9 dans la console Python. Attente de 1 seconde (1000 millisecondes) avant de traiter une autre capture de la cam\u00e9ra. Sortir de la boucle si la touche \u00abesc\u00bb, key==27, est appuy\u00e9e. Terminaison de la capture et fermeture des fen\u00eatres.<\/p>\n\n\n\n <\/p>\n\n\n\n Il faut v\u00e9rifier que chaque \u00abdigit\u00bb est bien d\u00e9cod\u00e9 pour les valeurs de 0 \u00e0 9. Le truc rapide est de faire \u00ab inhaler \u00bb un peu de Purel au d\u00e9tecteur ou souffler sur la flamme d’une chandelle pr\u00e8s du d\u00e9tecteur. Les \u00abdigits\u00bb HCH0, TVOC passe par tous les chiffres \u00e9ventuellement. C’est quand m\u00eame un bon test pour le Co2 et les autres.<\/p>\n\n\n\n Comme l’image peut tout de m\u00eame \u00eatre d\u00e9form\u00e9e par la cam\u00e9ra (les \u00abdigits\u00bb n’ont pas exactement la m\u00eame dimension et le m\u00eame espacement), il est pr\u00e9f\u00e9rable de trouver la position exacte de chaque \u00abdigit\u00bb par essais-erreurs et de modifier le programme en cons\u00e9quence.<\/p>\n\n\n\n La cam\u00e9ra a \u00e9t\u00e9 boug\u00e9e, donc les rep\u00e8res initiaux sont aussi modifi\u00e9s. Les changements importants sont :<\/p>\n\n\n\n Vue globale du d\u00e9but de la boucle<\/p>\n\n\n\n S\u00e9lectionner \u00ab Ouvrir \u00bb, puis parcourir pour aller dans le r\u00e9pertoire du fichier \u00ab .txt \u00bb. <\/p>\n\n\n\n S\u00e9lectionner \u00ab Tous les fichiers (*.*) \u00bb, puis le fichier \u00ab .txt \u00bb.<\/p>\n\n\n\n Suivre les \u00e9tapes 1,2,3<\/p>\n\n\n\n S\u00e9lection du point comme s\u00e9parateur d\u00e9cimal, puis Terminer R\u00e9sultats avec l’option \u00ab Filtrer\u00bb active<\/p>\n\n\n\n Fonction d’insertion de graphiques <\/p>\n\n\n\n En g\u00e9n\u00e9ral, les lectures sont bonnes, mais comme la largeur des segments est mince , le moindre d\u00e9placement ou une vibration peut causer une erreur de lecture. La cam\u00e9ra doit \u00eatre bien centr\u00e9e par rapport \u00e0 l’\u00e9cran de l’appareil de mesure et bien perpendiculaire verticalement et horizontalement pour r\u00e9duire la d\u00e9formation de l’image. Une erreur de lecture peut aussi se produire lors de la lecture d’un \u00abdigit\u00bb pendant de sa phase de changement du chiffre. Ces erreurs sont en g\u00e9n\u00e9ral simples \u00e0 d\u00e9tecter, nous les voyons facilement sur les graphiques, elles sortent du lot.<\/p>\n\n\n\n L’utilisation du \u00abdeep learning\u00bb augmenterait la pr\u00e9cision de la lecture, le syst\u00e8me serait moins sensible aux d\u00e9formations et vibrations, mais sa mise en oeuvre est plus complexe, elle demande plus de travail. C’est une avenue \u00e0 regarder.<\/p>\n\n\n\n En assemblant la cam\u00e9ra et le d\u00e9tecteur de qualit\u00e9 de l’air dans un montage rigide, les lectures seraient suffisamment fiables et exploitables. Le r\u00e9seau de neurones artificiel fixe, pr\u00e9d\u00e9terminer, \u00e0 une seule couche est simple \u00e0 mettre en oeuvre pour cette application et rempli son objectif.<\/p>\n","protected":false},"excerpt":{"rendered":" But Le but est de d\u00e9montrer le fonctionnement de base d’un r\u00e9seau de neurones artificiels. Cette d\u00e9monstration passe par la lecture de l’affichage du d\u00e9tecteur de qualit\u00e9 de l’air JSM-131 SE \u00e0 partir d’un ordinateur et d’une cam\u00e9ra. Les donn\u00e9es captur\u00e9es sont enregistr\u00e9es dans un fichier texte. Le formatage du fichier permettra d’importer ces donn\u00e9es […]<\/p>\n","protected":false},"author":2,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"_links":{"self":[{"href":"https:\/\/espacerm.com\/webgen\/wp-json\/wp\/v2\/pages\/4415"}],"collection":[{"href":"https:\/\/espacerm.com\/webgen\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/espacerm.com\/webgen\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/espacerm.com\/webgen\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/espacerm.com\/webgen\/wp-json\/wp\/v2\/comments?post=4415"}],"version-history":[{"count":248,"href":"https:\/\/espacerm.com\/webgen\/wp-json\/wp\/v2\/pages\/4415\/revisions"}],"predecessor-version":[{"id":4715,"href":"https:\/\/espacerm.com\/webgen\/wp-json\/wp\/v2\/pages\/4415\/revisions\/4715"}],"wp:attachment":[{"href":"https:\/\/espacerm.com\/webgen\/wp-json\/wp\/v2\/media?parent=4415"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}VIVOSUN Moniteur de qualit\u00e9 de l’air – Testeur de CO2 – D\u00e9tecteur de pollution int\u00e9rieur – HCHO\/TVOC\/PM2.5\/PM10 – Testeur avec donn\u00e9es en temps r\u00e9el et enregistrement de valeur moyenne pour la maison et le bureau : Amazon.ca: Commerce, Industrie et Science<\/a><\/h6>\n\n\n\n
Mini webcam USB \u00e0 objectif varifocal 2,8-12 mm haute d\u00e9finition USB avec cam\u00e9ra, mise au point manuelle, compatible avec la plupart des OS, mise au point r\u00e9glable, webcam USB 2.0 haute vitesse : Amazon.ca: \u00c9lectronique<\/a><\/h6>\n\n\n\n
\n\n\n\nR\u00e9f\u00e9rence<\/h3>\n\n\n\n
https:\/\/espacerm.com\/Rnia\/rnia_cible.php?numero_image=001<\/a><\/h6>\n\n\n\n
\n\n\n\nFonctionnement du r\u00e9seau de neurones artificiels<\/h3>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/Diapositive6.jpg\")
<\/figure><\/div>\n\n\n\n
<\/p>\n\n\n\n![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/Diapositive7.jpg\")
<\/figure><\/div>\n\n\n\n
<\/p>\n\n\n\n![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/Diapositive8.jpg\")
<\/figure><\/div>\n\n\n\n
\n\n\n\nProgrammation<\/h3>\n\n\n\n
# D\u00e9fini le port de capture vid\u00e9o \ncap = cv2.VideoCapture(0)\n<\/pre>\n\n\n\n
# Lecture en boucle de l'image vid\u00e9o\nwhile True:\n _, frame = cap.read()<\/pre>\n\n\n\n
L’image contenue dans \u00abframe\u00bb est encod\u00e9e en valeurs RGB<\/a> (Rouge-Green-Blue) ou RVB (Rouge-Vert-Bleu)
Pour le traitement ult\u00e9rieur, il faut que l’image soit encod\u00e9e en valeurs HSV<\/a> (Hue-Saturation-value) ou TSV (Teinte-Saturation-Valeur)<\/p>\n\n\n\n# Conversion de l'image RGB en HSV\nhsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)<\/pre>\n\n\n\n
Cr\u00e9ation d’un masque qui servira \u00e0 diff\u00e9rentier les segments allum\u00e9s du reste de l’image. Comme l’\u00e9cran est en majorit\u00e9 de couleur bleue et que les segments allum\u00e9s sont blancs, nous allons convertir ce qui est bleu (selon une certaine marge) en blanc, ce qui est blanc en noir.<\/p>\n\n\n\n lower_blue = np.array([38, 112, 87])\n upper_blue = np.array([179, 255, 255])\n \n mask = cv2.inRange(hsv, lower_blue, upper_blue)<\/pre>\n\n\n\n
Affichage \u00e0 l’\u00e9cran de l’image originale et de l’image masque<\/p>\n\n\n\n cv2.imshow(\"Frame\", frame)\n cv2.imshow(\"Mask\", mask)<\/pre>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/frame-et-mask_00-642x1024.jpg\")
<\/figure><\/div>\n\n\n\n
<\/p>\n\n\n\nRecherche de la position des segments<\/h5>\n\n\n\n
\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\n crop_img = source[y: y+h, x: x+w]\n \"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\n HCHO = mask[69: 125:, 185: 480]\n cv2.imshow(\"HCHO\", HCHO)<\/pre>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/setup-alignement_HCHO-1.jpg\")
<\/figure><\/div>\n\n\n\n
Nous trouvons pour TVOC, le pixel y=127, x= 185. <\/p>\n\n\n\nTVOC = mask[127: 183:, 185: 480]\ncv2.imshow(\"TVOC\", TVOC)<\/pre>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/setup-alignement_TVOC-1.jpg\")
<\/figure><\/div>\n\n\n\n
Pour PM2.5, le pixel y=186, x= 185.<\/p>\n\n\n\nPM2_5 = mask[186: 242:, 185: 480]\ncv2.imshow(\"PM2_5\", PM2_5)<\/pre>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/setup-alignement_PM2_5-1.jpg\")
<\/figure><\/div>\n\n\n\n PM10 = mask[246: 302:, 185: 480]\n cv2.imshow(\"PM10\", PM10)<\/pre>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/setup-alignement_PM10-1.jpg\")
<\/figure><\/div>\n\n\n\n CO2 = mask[305: 361:, 185: 480]\n cv2.imshow(\"CO2\", CO2)<\/pre>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/setup-alignement_Co2-1.jpg\")
<\/figure><\/div>\n\n\n\n
<\/p>\n\n\n\nRecherche de la position des points d\u00e9cimaux<\/h5>\n\n\n\n
# HCHO maximum = 1.999\n HCHO_dot1 = mask[104: 109:, 220: 226]\n cv2.imshow(\"HCHO_dot1\", HCHO_dot1)<\/pre>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/setup-alignement_dot-HCHO-1.jpg\")
<\/figure><\/div>\n\n\n\n
Pour le TVOC, sa valeur se situe entre 0 et 12, il y a deux positions possibles pour le point d\u00e9cimal.<\/p>\n\n\n\n # TVOC maximum = 12\n TVOC_dot1 = mask[162: 168:, 220: 226]\n cv2.imshow(\"TVOC_dot1\", TVOC_dot1)\n \n TVOC_dot2 = mask[162: 168:, 272: 278]\n cv2.imshow(\"TVOC_dot2\", TVOC_dot2)<\/pre>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/setup-alignement_pd-TVOC-1.jpg\")
<\/figure><\/div>\n\n\n\n
Vue globale sur l’\u00e9cran de l’ordinateur<\/p>\n\n\n\n![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/aligne_00-1024x681.jpg\")
<\/figure><\/div>\n\n\n\n
Le mot \u00abdigit\u00bb sera employ\u00e9 dans la suite plut\u00f4t que \u00abchiffre\u00bb pour r\u00e9f\u00e9rer \u00e0 l’ensemble des 7 segments qui composent un chiffre. Cela a pour but de diff\u00e9rentier entre chiffre la valeur 0,1,2,3,4,5,6,7,8,9 et chiffre en tant qu’unit\u00e9, dizaine, centaine…<\/p>\n\n\n\n# Assignement m\u00e9moire pour 5 nombres de 4 \u00abdigits\u00bb\nnumber=np.array([[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4]])\n\n # Position en y (ligne) du premier pixel de chaque nombre\n posNumber=np.array([69,127,186,246,305])\n # Position en x (col) du premier pixel de chaque \u00abdigit\u00bb\n posDigit=np.array([185,239,289,340])<\/pre>\n\n\n\n
La partie \u00abr\u00e9seau de neurones\u00bb<\/h5>\n\n\n\n
for m in range(0, 5): # 5 Nombres (position 0,1,2,3,4)\n line=posNumber[m]\n for j in range(0, 4): # 4 \u00abdigits\u00bb (position 0,1,2,3)\n col=posDigit[j]<\/pre>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/Diapositive7b.jpg\")
<\/figure>\n\n\n\n
Le pixel nomm\u00e9 a1 vaut 1 s’il est noir ( == 0) sur le mask sinon il vaut -1 <\/p>\n\n\n\n # un pixel blanc sur le mask a comme valeur 255\n # un pixel noir sur le mask a comme valeur 0\n a1 = 1 if mask[line+1,col+9]== 0 else -1<\/pre>\n\n\n\n
# Assignement m\u00e9moire pour 5 nombres de 4 \u00abdigits\u00bb\n number=np.array([[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4]])\n \n # Position en y (ligne) du premier pixel de chaque nombre\n posNumber=np.array([69,127,186,246,305])\n # Position en x (col) du premier pixel de chaque \u00abdigit\u00bb\n posDigit=np.array([185,239,289,340])\n\n for m in range(0, 5): # 5 Nombres (position 0,1,2,3,4)\n line=posNumber[m]\n for j in range(0, 4): # 4 \u00abdigits\u00bb (position 0,1,2,3)\n col=posDigit[j]\n\n # un pixel blanc sur le mask a comme valeur 255\n # un pixel noir sur le mask a comme valeur 0\n a1 = 1 if mask[line+1,col+9] == 0 else -1\n a2 = 1 if mask[line+1,col+10] == 0 else -1\n a3 = 1 if mask[line+1,col+11] == 0 else -1\n a4 = 1 if mask[line+1,col+12] == 0 else -1\n<\/pre>\n\n\n\n
mask[line+1,col+9]<\/pre>\n\n\n\n
a1 = 1 if mask[line+1,col+9]== 0 else -1<\/pre>\n\n\n\n
Chaque segment est \u00e9valu\u00e9<\/p>\n\n\n\n a1 = 1 if mask[line+1,col+9] == 0 else -1\n a2 = 1 if mask[line+1,col+10] == 0 else -1\n a3 = 1 if mask[line+1,col+11] == 0 else -1\n a4 = 1 if mask[line+1,col+12] == 0 else -1\n\n \n b1 = 1 if mask[line+7,col+20] == 0 else -1\n b2 = 1 if mask[line+8,col+20] == 0 else -1\n b3 = 1 if mask[line+9,col+20] == 0 else -1\n b4 = 1 if mask[line+10,col+20] == 0 else -1\n\n \n c1 = 1 if mask[line+27,col+20] == 0 else -1\n c2 = 1 if mask[line+28,col+20] == 0 else -1\n c3 = 1 if mask[line+29,col+20] == 0 else -1\n c4 = 1 if mask[line+30,col+20] == 0 else -1 \n \n \n d1 = 1 if mask[line+37,col+9] == 0 else -1\n d2 = 1 if mask[line+37,col+10] == 0 else -1\n d3 = 1 if mask[line+37,col+11] == 0 else -1\n d4 = 1 if mask[line+37,col+12] == 0 else -1\n\n \n e1 = 1 if mask[line+27,col+0] == 0 else -1\n e2 = 1 if mask[line+28,col+0] == 0 else -1\n e3 = 1 if mask[line+29,col+0] == 0 else -1\n e4 = 1 if mask[line+30,col+0] == 0 else -1\n \n \n f1 = 1 if mask[line+7,col+0] == 0 else -1\n f2 = 1 if mask[line+8,col+0] == 0 else -1\n f3 = 1 if mask[line+9,col+0] == 0 else -1\n f4 = 1 if mask[line+10,col+0] == 0 else -1\n \n \n g1 = 1 if mask[line+18,col+9] == 0 else -1\n g2 = 1 if mask[line+18,col+10] == 0 else -1\n g3 = 1 if mask[line+18,col+11] == 0 else -1\n g4 = 1 if mask[line+18,col+12] == 0 else -1<\/pre>\n\n\n\n
La somme de la valeur trouv\u00e9e pour chaque pixel pour chaque segment est calcul\u00e9e. Cette somme sera donc comprise entre 4 et -4. <\/p>\n\n\n\n sa=a1+a2+a3+a4 # Segment a\n sb=b1+b2+b3+b4 # Segment b\n sc=c1+c2+c3+c4 # Segment c\n sd=d1+d2+d3+d4 # Segment d\n se=e1+e2+e3+e4 # Segment e\n sf=f1+f2+f3+f4 # Segment f\n sg=g1+g2+g3+g4 # Segment g<\/pre>\n\n\n\n
La valeur de sortie de chaque neurone est calcul\u00e9e. Exemple pour le neurone \u00ab1\u00bb, il faut additionner la valeur des segments \u00abb\u00bb et \u00abc\u00bb.<\/p>\n\n\n\n # Assignement m\u00e9moire pour chaque chiffre \u00e0 repr\u00e9senter\n r=np.array([0,1,2,3,4,5,6,7,8,9])\n \n r[0]=sa+sb+sc+sd+se+sf # valeur de sortie du neurone \u00ab0\u00bb\n r[1]=sb+sc # valeur de sortie du neurone \u00ab1\u00bb\n r[2]=sa+sb+sg+se+sd # valeur de sortie du neurone \u00ab2\u00bb\n r[3]=sa+sb+sg+sc+sd # valeur de sortie du neurone \u00ab3\u00bb\n r[4]=sb+sc+sf+sg # valeur de sortie du neurone \u00ab4\u00bb\n r[5]=sa+sf+sg+sc+sd # valeur de sortie du neurone \u00ab5\u00bb\n r[6]=sa+sf+sg+se+sd+sc # valeur de sortie du neurone \u00ab6\u00bb\n r[7]=sa+sb+sc # valeur de sortie du neurone \u00ab7\u00bb\n r[8]=sa+sb+sc+sd+se+sf+sg # valeur de sortie du neurone \u00ab8\u00bb\n r[9]=sa+sb+sc+sd+sf+sg # valeur de sortie du neurone \u00ab9\u00bb<\/pre>\n\n\n\n
Prochaine \u00e9tape, il faut trouver la valeur maximum<\/p>\n\n\n\n mx=max(r[0],r[1],r[2],r[3],r[4],r[5],r[6],r[7],r[8],r[9])<\/pre>\n\n\n\n
Il reste \u00e0 trouver qui poss\u00e8de la valeur maximum<\/p>\n\n\n\n for i in range(0, 10):\n if r[i] == mx: break\n \n number[m,j]=i<\/pre>\n\n\n\n
<\/p>\n\n\n\nD\u00e9tection de la position du point d\u00e9cimal pour TVOC <\/h5>\n\n\n\n
if m == 1:\n \n tvocD1a = 1 if mask[166,222] == 0 else -1\n tvocD1b = 1 if mask[166,223] == 0 else -1\n tvocD1c = 1 if mask[167,222] == 0 else -1\n tvocD1d = 1 if mask[167,223] == 0 else -1\n \n tvocD1 = tvocD1a+tvocD1b+tvocD1c+tvocD1d\n \n \n tvocD2a = 1 if mask[166,274] == 0 else -1\n tvocD2b = 1 if mask[166,274] == 0 else -1\n tvocD2c = 1 if mask[167,275] == 0 else -1\n tvocD2d = 1 if mask[167,275] == 0 else -1\n \n\n tvocD2 = tvocD2a+tvocD2b+tvocD2c+tvocD2d<\/pre>\n\n\n\n
<\/p>\n\n\n\n print(number[m])\n \n key = cv2.waitKey(1000)\n if key == 27:\n break\n \ncap.release()\ncv2.destroyAllWindows()\n <\/pre>\n\n\n\n
<\/p>\n\n\n\nProgramme complet<\/h2>\n\n\n\n
# -*- coding: utf-8 -*-\n\"\"\"\n Cr\u00e9er le 2022-02-03\n\n@auteur: Richard Morel\n\nAir Quality Monitor\n\n\"\"\"\n\nimport time\n\nimport cv2\nimport numpy as np\n\n\n \n # D\u00e9fini le port de capture vid\u00e9o \ncap = cv2.VideoCapture(0)\n\n# Lecture en boucle de l'image vid\u00e9o\nwhile True:\n time.sleep(1)\n _, frame = cap.read()\n #print(frame.shape)\n\n # Conversion de l'image RGB en HSV\n hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)\n \n lower_blue = np.array([38, 112, 87])\n upper_blue = np.array([179, 255, 255])\n \n mask = cv2.inRange(hsv, lower_blue, upper_blue)\n \n cv2.imshow(\"Frame\", frame)\n cv2.imshow(\"Mask\", mask)\n \n \n \"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\n crop_img = source[y: y+h, x: x+w]\n \"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\n colFin=480\n \n HCHO = mask[69: 125:, 185: colFin]\n cv2.imshow(\"HCHO\", HCHO)\n\n TVOC = mask[127: 183:, 185: colFin]\n cv2.imshow(\"TVOC\", TVOC)\n \n PM2_5 = mask[186: 242:, 185: colFin]\n cv2.imshow(\"PM2_5\", PM2_5)\n \n PM10 = mask[246: 302:, 185: colFin]\n cv2.imshow(\"PM10\", PM10)\n \n CO2 = mask[304: 360:, 185: colFin]\n cv2.imshow(\"CO2\", CO2)\n \n \n # HCHO maximum = 1.999\n HCHO_dot1 = mask[104: 109:, 220: 226]\n cv2.imshow(\"HCHO_dot1\", HCHO_dot1)\n \n # TVOC maximum = 12\n TVOC_dot1 = mask[162: 168:, 220: 226]\n cv2.imshow(\"TVOC_dot1\", TVOC_dot1)\n \n TVOC_dot2 = mask[162: 168:, 272: 278]\n cv2.imshow(\"TVOC_dot2\", TVOC_dot2)\n \n \n \n # Assignement m\u00e9moire pour 5 nombres de 4 \u00abdigits\u00bb\n number=np.array([[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4]])\n \n # Position en y (ligne) du premier pixel de chaque nombre\n posNumber=np.array([69,127,186,246,305])\n # Position en x (col) du premier pixel de chaque \u00abdigit\u00bb\n posDigit=np.array([185,239,289,340])\n\n\n print('-')\n for m in range(0, 5): # 5 Nombres (position 0,1,2,3,4)\n line=posNumber[m]\n for j in range(0, 4): # 4 \u00abdigits\u00bb (position 0,1,2,3)\n col=posDigit[j]\n #print(col)\n #print(line)\n # un pixel blanc sur le mask a comme valeur 255\n # un pixel noir sur le mask a comme valeur 0\n a1 = 1 if mask[line+1,col+9] == 0 else -1\n a2 = 1 if mask[line+1,col+10] == 0 else -1\n a3 = 1 if mask[line+1,col+11] == 0 else -1\n a4 = 1 if mask[line+1,col+12] == 0 else -1\n\n \n b1 = 1 if mask[line+7,col+20] == 0 else -1\n b2 = 1 if mask[line+8,col+20] == 0 else -1\n b3 = 1 if mask[line+9,col+20] == 0 else -1\n b4 = 1 if mask[line+10,col+20] == 0 else -1\n\n \n c1 = 1 if mask[line+27,col+20] == 0 else -1\n c2 = 1 if mask[line+28,col+20] == 0 else -1\n c3 = 1 if mask[line+29,col+20] == 0 else -1\n c4 = 1 if mask[line+30,col+20] == 0 else -1 \n \n \n d1 = 1 if mask[line+37,col+9] == 0 else -1\n d2 = 1 if mask[line+37,col+10] == 0 else -1\n d3 = 1 if mask[line+37,col+11] == 0 else -1\n d4 = 1 if mask[line+37,col+12] == 0 else -1\n\n \n e1 = 1 if mask[line+27,col+0] == 0 else -1\n e2 = 1 if mask[line+28,col+0] == 0 else -1\n e3 = 1 if mask[line+29,col+0] == 0 else -1\n e4 = 1 if mask[line+30,col+0] == 0 else -1\n \n \n f1 = 1 if mask[line+7,col+0] == 0 else -1\n f2 = 1 if mask[line+8,col+0] == 0 else -1\n f3 = 1 if mask[line+9,col+0] == 0 else -1\n f4 = 1 if mask[line+10,col+0] == 0 else -1\n \n \n g1 = 1 if mask[line+18,col+9] == 0 else -1\n g2 = 1 if mask[line+18,col+10] == 0 else -1\n g3 = 1 if mask[line+18,col+11] == 0 else -1\n g4 = 1 if mask[line+18,col+12] == 0 else -1\n\n\n \n \n sa=a1+a2+a3+a4 # Segment a\n sb=b1+b2+b3+b4 # Segment b\n sc=c1+c2+c3+c4 # Segment c\n sd=d1+d2+d3+d4 # Segment d\n se=e1+e2+e3+e4 # Segment e\n sf=f1+f2+f3+f4 # Segment f\n sg=g1+g2+g3+g4 # Segment g\n \n \n # Assignement m\u00e9moire pour chaque \u00abdigit\u00bb \u00e0 repr\u00e9senter\n r=np.array([0,1,2,3,4,5,6,7,8,9])\n \n r[0]=sa+sb+sc+sd+se+sf # valeur de sortie du neurone \u00ab0\u00bb\n r[1]=sb+sc # valeur de sortie du neurone \u00ab1\u00bb\n r[2]=sa+sb+sg+se+sd # valeur de sortie du neurone \u00ab2\u00bb\n r[3]=sa+sb+sg+sc+sd # valeur de sortie du neurone \u00ab3\u00bb\n r[4]=sb+sc+sf+sg # valeur de sortie du neurone \u00ab4\u00bb\n r[5]=sa+sf+sg+sc+sd # valeur de sortie du neurone \u00ab5\u00bb\n r[6]=sa+sf+sg+se+sd+sc # valeur de sortie du neurone \u00ab6\u00bb\n r[7]=sa+sb+sc # valeur de sortie du neurone \u00ab7\u00bb\n r[8]=sa+sb+sc+sd+se+sf+sg # valeur de sortie du neurone \u00ab8\u00bb\n r[9]=sa+sb+sc+sd+sf+sg # valeur de sortie du neurone \u00ab9\u00bb\n \n mx=max(r[0],r[1],r[2],r[3],r[4],r[5],r[6],r[7],r[8],r[9])\n #print (mx)\n \n for i in range(0, 10):\n if r[i] == mx: break\n \n number[m,j]=i\n \n #\n # HCHO_dot1 = mask[104: 109:, 220: 225]\n #\n\n if m == 1:\n \n tvocD1a = 1 if mask[166,222] == 0 else -1\n tvocD1b = 1 if mask[166,223] == 0 else -1\n tvocD1c = 1 if mask[167,222] == 0 else -1\n tvocD1d = 1 if mask[167,223] == 0 else -1\n \n tvocD1 = tvocD1a+tvocD1b+tvocD1c+tvocD1d\n \n \n tvocD2a = 1 if mask[166,274] == 0 else -1\n tvocD2b = 1 if mask[166,274] == 0 else -1\n tvocD2c = 1 if mask[167,275] == 0 else -1\n tvocD2d = 1 if mask[167,275] == 0 else -1\n \n\n tvocD2 = tvocD2a+tvocD2b+tvocD2c+tvocD2d\n \n \n print(tvocD1)\n print(tvocD2)\n\n \n print(number[m])\n \n\n key = cv2.waitKey(100)\n #(curseur dans une des imshow)\n if key == 27:\n break\n \n\ncap.release()\ncv2.destroyAllWindows()\n<\/pre>\n\n\n\n
\n\n\n\nEssais<\/h2>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/validation_02-1024x577.jpg\")
<\/figure><\/div>\n\n\n\n
\n\n\n\nProgramme incluant l’\u00e9criture des donn\u00e9es sur disque<\/h2>\n\n\n\n
posDigit=np.array([[186,236,285,335],[186,234,284,334],[185,235,285,335],[185,235,283,332],[184,233,283,334]])<\/pre>\n\n\n\n
col=posDigit[m,j]<\/pre>\n\n\n\n
# Assignement m\u00e9moire pour 5 nombres de 4 \u00abdigits\u00bb\n number=np.array([[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4]])\n \n # Position en y (ligne) du premier pixel de chaque nombre\n posNumber=np.array([70,125,181,238,294])\n # Position en x (col) du premier pixel de chaque \u00abdigit\u00bb\n posDigit=np.array([[186,236,285,335],\n [186,234,284,334],\n [185,235,285,335],\n [185,235,283,332],\n [184,233,283,334]])\n\n\n print('-')\n for m in range(0, 5): # 5 Nombres (position 0,1,2,3,4)\n line=posNumber[m]\n for j in range(0, 4): # 4 \u00abdigits\u00bb (position 0,1,2,3)\n col=posDigit[m,j]<\/pre>\n\n\n\n
Les donn\u00e9es lues seront \u00e9crites dans un fichier \u00ab.txt\u00bb. Le \u00ab ; \u00bb sert de d\u00e9limiteur entre les donn\u00e9es dans le fichier texte. <\/p>\n\n\n\nDATA_PATH = \"C:\/Documents\/Programmation\/Python\/Travail\/analyse d image\/\"\ndataFileName = DATA_PATH + \"dataDir_00\/dataFile_00.txt\" \ndataFile = open(dataFileName, 'a') #mode \u00abappend\u00bb, ajout\n\n#Ent\u00eate des colonnes\ndataFile.write('Date;'+'Heure;'+'HCHO;'+'HCHO;'+'TVOC;'+'TVOC;'+'PM2.5;'\n +'PM2.5;'+'PM10;'+'PM10;'+'Co2;'+'Co2;'+'\\r')<\/pre>\n\n\n\nNouveaux \u00abmasks\u00bb d’essais<\/h4>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/align-carac_04-1024x628.jpg\")
<\/figure><\/div>\n\n\n\n
<\/p>\n\n\n\nProgramme complet incluant l’\u00e9criture des donn\u00e9es sur disque<\/h2>\n\n\n\n
# -*- coding: utf-8 -*-\n\"\"\"\n Cr\u00e9er le 2022-02-03\n\n@auteur: Richard Morel\n \n Lecture de l'affichage du senseur de qualit\u00e9 de l'air\n\n Air Quality Monitor Rechargeable\n JSM-131 SE\n \n Camera\n ELP-USBFHD01M-MFV\n \n # HCHO maximum = 1.999\n # TVOC maximum = 12\n \n\"\"\"\n\nimport time\nimport datetime\n\nimport cv2\nimport numpy as np\n\nDATA_PATH = \"C:\/Documents\/Programmation\/Python\/Travail\/analyse d image\/\"\ndataFileName = DATA_PATH + \"dataDir_00\/dataFile_00.txt\" \ndataFile = open(dataFileName, 'a') #mode \u00abappend\u00bb, ajout\n\n#Ent\u00eate des colonnes\ndataFile.write('Date;'+'Heure;'+'HCHO;'+'HCHO;'+'TVOC;'+'TVOC;'+'PM2.5;'\n +'PM2.5;'+'PM10;'+'PM10;'+'Co2;'+'Co2;'+'\\r')\n \ncap = cv2.VideoCapture(0)\n\n\n\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\n \n Lecture de l'affichage du JSM-131 SE\n\n\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\" \n\nwhile True:\n _, frame = cap.read()\n # print(frame.shape)\n\n hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)\n \n \n lower_blue = np.array([38, 112, 87])\n upper_blue = np.array([179, 255, 255])\n \n mask = cv2.inRange(hsv, lower_blue, upper_blue)\n \n cv2.imshow(\"Frame\", frame)\n cv2.imshow(\"Mask\", mask)\n \n \n \"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\n crop_img = source[y: y+h, x: x+w]\n \"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\n colFin=480\n \n # HCHO\n \"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\n ligne 69: l+h= 69+42, colonne 186: 480\n \"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\" \n HCHO = mask[69: 111:, 186: 480]\n cv2.imshow(\"HCHO\", HCHO)\n \n \"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\n ligne 69: l+h= 69+42, colonne 186: c+26\n \"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\n \n g=21\n d=38\n HCHO1 = mask[69: 111:, 186: 212]\n cv2.imshow(\"HCHO1\", HCHO1)\n \n HCHO2 = mask[69: 111:, 236: 260]\n cv2.imshow(\"HCHO2\", HCHO2)\n \n HCHO3 = mask[69: 111:, 285: 310]\n cv2.imshow(\"HCHO3\", HCHO3)\n \n HCHO4 = mask[69: 111:, 335: 358]\n cv2.imshow(\"HCHO4\", HCHO4)\n \n HCHO4g = mask[69+g: 111:, 186: 358]\n cv2.imshow(\"HCHO4g\", HCHO4g)\n \n HCHO4d = mask[69+d: 111:, 335: 358]\n cv2.imshow(\"HCHO4d\", HCHO4d)\n \n \n \n # TVOC\n TVOC = mask[125: 167:, 186: colFin]\n cv2.imshow(\"TVOC\", TVOC)\n \n TVOC1 = mask[125: 167:, 186: 210]\n cv2.imshow(\"TVOC1\", TVOC1)\n \n TVOC2 = mask[125: 167:, 234: 260]\n cv2.imshow(\"TVOC2\", TVOC2)\n \n TVOC3 = mask[125: 167:, 284: 308]\n cv2.imshow(\"TVOC3\", TVOC3)\n \n TVOC4 = mask[125: 167:, 334: 358]\n cv2.imshow(\"TVOC4\", TVOC4)\n \n TVOC4g = mask[125+g: 167:, 186: 358]\n cv2.imshow(\"TVOC4g\", TVOC4g) \n \n TVOC4d = mask[125+d: 167:, 334: 358]\n cv2.imshow(\"TVOC4d\", TVOC4d)\n \n \n \n \n PM2_5 = mask[181: 223:, 185: colFin]\n cv2.imshow(\"PM2_5\", PM2_5)\n \n PM2_5_1 = mask[181: 223:, 185: 210]\n cv2.imshow(\"PM2_5_1\", PM2_5_1)\n \n PM2_5_2 = mask[181: 223:, 235: 258]\n cv2.imshow(\"PM2_5_2\", PM2_5_2)\n \n PM2_5_3 = mask[181: 223:, 285: 308]\n cv2.imshow(\"PM2_5_3\", PM2_5_3)\n \n PM2_5_4 = mask[181: 223:, 335: 358]\n cv2.imshow(\"PM2_5_4\", PM2_5_4)\n \n PM2_5_4g = mask[181+g: 223:, 185: 358]\n cv2.imshow(\"PM2_5_4g\", PM2_5_4g)\n\n PM2_5_4d = mask[181+d: 223:, 335: 358]\n cv2.imshow(\"PM2_5_4d\", PM2_5_4d)\n \n \n \n PM10 = mask[238: 281:, 185: colFin]\n cv2.imshow(\"PM10\", PM10)\n \n PM10_1 = mask[238: 281:, 185: 208]\n cv2.imshow(\"PM10_1\", PM10_1)\n \n PM10_2 = mask[238: 281:, 235: 258]\n cv2.imshow(\"PM10_2\", PM10_2)\n \n PM10_3 = mask[238: 281:, 283: 308]\n cv2.imshow(\"PM10_3\", PM10_3)\n \n PM10_4 = mask[238: 281:, 332: 358]\n cv2.imshow(\"PM10_4\", PM10_4) \n \n PM10_4g = mask[238+g: 281:, 185: 358]\n cv2.imshow(\"PM10_4g\", PM10_4g) \n\n PM10_4d = mask[238+d: 281:, 332: 358]\n cv2.imshow(\"PM10_4d\", PM10_4d) \n \n\n\n\n CO2 = mask[294: 336:, 184: colFin]\n cv2.imshow(\"CO2\", CO2)\n \n CO2_1 = mask[294: 336:, 184: 208]\n cv2.imshow(\"CO2_1\", CO2_1)\n \n CO2_2 = mask[294: 336:, 233: 258]\n cv2.imshow(\"CO2_2\", CO2_2)\n\n CO2_3 = mask[294: 336:, 283: 308]\n cv2.imshow(\"CO2_3\", CO2_3)\n\n CO2_4 = mask[294: 336:, 333: 357]\n cv2.imshow(\"CO2_4\", CO2_4) \n \n CO2_4g = mask[294+g: 336:, 184: 357]\n cv2.imshow(\"CO2_4g\", CO2_4g) \n \n CO2_4d = mask[294+d: 336:, 333: 357]\n cv2.imshow(\"CO2_4d\", CO2_4d) \n \n \n \n # HCHO maximum = 1.999\n HCHO_dot1 = mask[103: 109:, 220: 226]\n cv2.imshow(\"HCHO_dot1\", HCHO_dot1)\n \n # TVOC maximum = 12\n TVOC_dot1 = mask[159: 168:, 220: 226]\n cv2.imshow(\"TVOC_dot1\", TVOC_dot1)\n \n TVOC_dot2 = mask[159: 168:, 271: 277]\n cv2.imshow(\"TVOC_dot2\", TVOC_dot2)\n \n\n \n # Assignement m\u00e9moire pour 5 nombres de 4 \u00abdigits\u00bb\n number=np.array([[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4]])\n \n # Position en y (ligne) du premier pixel de chaque nombre\n posNumber=np.array([70,125,181,238,294])\n # Position en x (col) du premier pixel de chaque \u00abdigit\u00bb\n posDigit=np.array([[186,236,285,335],\n [186,234,284,334],\n [185,235,285,335],\n [185,235,283,332],\n [184,233,283,334]])\n\n\n print('-')\n for m in range(0, 5): # 5 Nombres (position 0,1,2,3,4)\n line=posNumber[m]\n for j in range(0, 4): # 4 \u00abdigits\u00bb (position 0,1,2,3)\n col=posDigit[m,j]\n #print(line)\n #print(col)\n\n # un pixel blanc sur le mask a comme valeur 255\n # un pixel noir sur le mask a comme valeur 0\n a1 = 1 if mask[line+1,col+11] == 0 else -1\n a2 = 1 if mask[line+1,col+12] == 0 else -1\n a3 = 1 if mask[line+1,col+13] == 0 else -1\n a4 = 1 if mask[line+1,col+14] == 0 else -1\n\n \n b1 = 1 if mask[line+8,col+22] == 0 else -1\n b2 = 1 if mask[line+9,col+22] == 0 else -1\n b3 = 1 if mask[line+10,col+22] == 0 else -1\n b4 = 1 if mask[line+11,col+22] == 0 else -1\n\n \n c1 = 1 if mask[line+29,col+22] == 0 else -1\n c2 = 1 if mask[line+30,col+22] == 0 else -1\n c3 = 1 if mask[line+31,col+22] == 0 else -1\n c4 = 1 if mask[line+32,col+22] == 0 else -1 \n \n \n d1 = 1 if mask[line+38,col+11] == 0 else -1\n d2 = 1 if mask[line+38,col+12] == 0 else -1\n d3 = 1 if mask[line+38,col+13] == 0 else -1\n d4 = 1 if mask[line+38,col+14] == 0 else -1\n\n \n e1 = 1 if mask[line+29,col+0] == 0 else -1\n e2 = 1 if mask[line+30,col+0] == 0 else -1\n e3 = 1 if mask[line+31,col+0] == 0 else -1\n e4 = 1 if mask[line+32,col+0] == 0 else -1\n \n \n f1 = 1 if mask[line+8,col+0] == 0 else -1\n f2 = 1 if mask[line+9,col+0] == 0 else -1\n f3 = 1 if mask[line+10,col+0] == 0 else -1\n f4 = 1 if mask[line+11,col+0] == 0 else -1\n \n \n g1 = 1 if mask[line+21,col+11] == 0 else -1\n g2 = 1 if mask[line+21,col+12] == 0 else -1\n g3 = 1 if mask[line+21,col+13] == 0 else -1\n g4 = 1 if mask[line+21,col+14] == 0 else -1\n\n\n \n \n sa=a1+a2+a3+a4 # Segment a\n sb=b1+b2+b3+b4 # Segment b\n sc=c1+c2+c3+c4 # Segment c\n sd=d1+d2+d3+d4 # Segment d\n se=e1+e2+e3+e4 # Segment e\n sf=f1+f2+f3+f4 # Segment f\n sg=g1+g2+g3+g4 # Segment g\n \n\n \n r=np.array([0,1,2,3,4,5,6,7,8,9])\n \n r[0]=sa+sb+sc+sd+se+sf # digit 0\n r[1]=sb+sc # digit 1\n r[2]=sa+sb+sg+se+sd # digit 2\n r[3]=sa+sb+sg+sc+sd # digit 3\n r[4]=sb+sc+sf+sg # digit 4\n r[5]=sa+sf+sg+sc+sd # digit 5\n r[6]=sa+sf+sg+se+sd+sc # digit 6\n r[7]=sa+sb+sc # digit 7\n r[8]=sa+sb+sc+sd+se+sf+sg # digit 8\n r[9]=sa+sb+sc+sd+sf+sg # digit 9\n \n mx=max(r[0],r[1],r[2],r[3],r[4],r[5],r[6],r[7],r[8],r[9])\n #print (mx)\n \n for i in range(0, 10):\n if r[i] == mx: break\n \n number[m,j]=i\n \n #print(number[m]) \n \n # D\u00e9tection de la position de la virgule pour TVOC\n if m == 1: \n tvocD1a = 1 if mask[162,222] == 0 else -1\n tvocD1b = 1 if mask[162,223] == 0 else -1\n tvocD1c = 1 if mask[163,222] == 0 else -1\n tvocD1d = 1 if mask[163,223] == 0 else -1\n \n tvocD1 = tvocD1a+tvocD1b+tvocD1c+tvocD1d\n \n \n tvocD2a = 1 if mask[162,272] == 0 else -1\n tvocD2b = 1 if mask[162,272] == 0 else -1\n tvocD2c = 1 if mask[163,273] == 0 else -1\n tvocD2d = 1 if mask[163,273] == 0 else -1\n \n tvocD2 = tvocD2a+tvocD2b+tvocD2c+tvocD2d\n \n #print(tvocD1)\n #print(tvocD2) \n \n \n \"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\n \n Enregistrement des donn\u00e9es dans un fichier\n\n \"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\"\" \n \n \n # D\u00e9buter les donn\u00e9es par un horodatage\n if m == 0:\n timeStamp = datetime.datetime.now()\n date_time = timeStamp.strftime(\"%Y-%m-%d; %H:%M:%S\")\n data = date_time + ';'\n \n # HCHO\n if m == 0:\n # Inclure l'ajout de la virgule pour HVCO\n stHCHO = str(number[m,0]) + '.' + str(number[m,1]) \\\n + str(number[m,2]) + str(number[m,3])\n print(stHCHO)\n HCHO= float(stHCHO)\n #print(HCHO)\n \n if HCHO >= 0 and HCHO <=1.999:\n if HCHO >= 0 and HCHO <= 0.080:\n quality_HCHO = '***** HCHO Excellent;'\n if HCHO >= 0.081 and HCHO <= 0.100:\n quality_HCHO = '**** HCHO Good;'\n if HCHO >= 0.101 and HCHO <= 0.200:\n quality_HCHO = '*** HCHO Mild;' \n if HCHO >= 0.201 and HCHO <= 0.500:\n quality_HCHO = '** HCHO Moderate;' \n if HCHO >= 0.501 and HCHO <= 1.000:\n quality_HCHO = '* HCHO Severe;' \n if HCHO >= 1.001 and HCHO <= 1.999:\n quality_HCHO = 'HCHO Serious;' \n else:\n quality_HCHO = 'HCHO ?;'\n \n #print(quality_HCHO)\n data = data + stHCHO + ';' + quality_HCHO \n \n \n \n #TVOC\n if m == 1:\n stTVOC = str(number[m,0])\n # ajout conditionnel de la virgule pour TVOC\n if tvocD1 == 4 and tvocD2 == -4:\n stTVOC = stTVOC + '.' \n stTVOC = stTVOC + str(number[m,1])\n # ajout conditionnel de la virgule pour TVOC\n if tvocD2 == 4 and tvocD1 == -4:\n stTVOC = stTVOC + '.' \n stTVOC = stTVOC + str(number[m,2])\n stTVOC = stTVOC + str(number[m,3])\n print(stTVOC)\n TVOC= float(stTVOC)\n #print(TVOC)\n \n if TVOC >= 0 and TVOC <=12:\n if TVOC >= 0 and TVOC <= 0.5:\n quality_TVOC = '***** TVOC Excellent;'\n if TVOC >= 0.501 and TVOC <= 0.6:\n quality_TVOC = '**** TVOC Good;'\n if TVOC >= 0.601 and TVOC <= 1.5:\n quality_TVOC = '*** TVOC Mild;' \n if TVOC >= 1.501 and TVOC <= 3.0:\n quality_TVOC = '** TVOC Moderate;' \n if TVOC >= 3.001 and TVOC <= 6.0:\n quality_TVOC = '* TVOC Severe;' \n if TVOC >= 6.001 and TVOC <= 12.0:\n quality_TVOC = 'TVOC Serious;' \n else:\n quality_TVOC = 'TVOC ?;'\n \n #print(quality_TVOC)\n data = data + stTVOC + ';' + quality_TVOC \n \n # PM2.5\n if m == 2:\n stPM2_5 = str(number[m,0]) + str(number[m,1]) \\\n + str(number[m,2]) + str(number[m,3])\n print(stPM2_5)\n PM2_5= float(stPM2_5)\n #print(PM2_5) \n\n if PM2_5 >= 0 and PM2_5 <=1000:\n if PM2_5 >= 0 and PM2_5 <= 35:\n quality_PM2_5 = '***** PM2.5 Excellent;'\n if PM2_5 >= 36 and PM2_5 <= 75:\n quality_PM2_5 = '**** PM2.5 Good;'\n if PM2_5 >= 76 and PM2_5 <= 100:\n quality_PM2_5 = '*** PM2.5 Mild;' \n if PM2_5 >= 101 and PM2_5 <= 150:\n quality_PM2_5 = '** PM2.5 Moderate;' \n if PM2_5 >= 151 and PM2_5 <= 250:\n quality_PM2_5 = '* PM2.5 Severe;' \n if PM2_5 >= 251 and PM2_5 <=1000:\n quality_PM2_5 = 'PM2.5 Serious;' \n else:\n quality_PM2_5 = 'PM2_5 ?;'\n \n #print(quality_PM2_5)\n data = data + stPM2_5 + ';' + quality_PM2_5 \n \n # PM10\n if m == 3:\n stPM10 = str(number[m,0]) + str(number[m,1]) \\\n + str(number[m,2]) + str(number[m,3])\n print(stPM10)\n PM10= float(stPM10)\n #print(PM10)\n \n if PM10 >= 0 and PM10 <=2000:\n if PM10 >= 0 and PM10 <= 50:\n quality_PM10 = '***** PM10 Excellent;'\n if PM10 >= 51 and PM10 <= 100:\n quality_PM10 = '**** PM10 Good;'\n if PM10 >= 101 and PM10 <= 130:\n quality_PM10 = '*** PM10 Mild;' \n if PM10 >= 131 and PM10 <= 200:\n quality_PM10 = '** PM10 Moderate;' \n if PM10 >= 201 and PM10 <= 300:\n quality_PM10 = '* PM10 Severe;' \n if PM10 >= 301 and PM10 <=2000:\n quality_PM10 = 'PM10 Serious;' \n else:\n quality_PM10 = 'PM10 ?;'\n \n #print(quality_PM10)\n data = data + stPM10 + ';' + quality_PM10 \n\n # Co2\n if m == 4:\n stCo2 = str(number[m,0]) + str(number[m,1]) \\\n + str(number[m,2]) + str(number[m,3])\n print(stCo2)\n Co2= float(stCo2)\n #print(Co2)\n \n if Co2 >= 0 and Co2 <=5000:\n if Co2 >= 0 and Co2 <= 450:\n quality_Co2 = '***** Co2 Excellent'\n if Co2 >= 451 and Co2 <=1000:\n quality_Co2 = '**** Co2 Good'\n if Co2 >= 1001 and Co2 <=1500:\n quality_Co2 = '*** Co2 Mild' \n if Co2 >= 1501 and Co2 <=2000:\n quality_Co2 = '** Co2 Moderate' \n if Co2 >= 2001 and Co2 <=3000:\n quality_Co2 = '* Co2 Severe' \n if Co2 >= 3001 and Co2 <=5000:\n quality_Co2 = 'Co2 Serious' \n else:\n quality_Co2 = 'Co2 ?'\n \n #print(quality_Co2)\n data = data + stCo2 + ';' + quality_Co2\n \n dataFile.write(data + '\\r') \n data = ''\n \n print(quality_HCHO)\n print(quality_TVOC)\n print(quality_PM2_5)\n print(quality_PM10)\n print(quality_Co2)\n \n # en seconde\n time.sleep(0)\n \n # en milliseconde\n key = cv2.waitKey(100)\n #(Placer le curseur dans une des imshow)\n if key == 27:\n dataFile.close()\n break\n \n\ncap.release()\ncv2.destroyAllWindows()\n\n<\/pre>\n\n\n\n
<\/p>\n\n\n\n![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/Table-de-reference-sur-la-qualite-de-l-air_00.jpg\")
\n\n\n\nImportation des donn\u00e9es dans Excel<\/h2>\n\n\n\n
![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/Presse-papier01.jpg\")
<\/figure>\n\n\n\n
<\/p>\n\n\n\n![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/Presse-papier02.jpg\")
<\/figure>\n\n\n\n
<\/p>\n\n\n\n![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/Presse-papier03.jpg\")
<\/figure>\n\n\n\n
<\/p>\n\n\n\n![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/resultats-excel_01b.jpg\")
<\/figure>\n\n\n\n
<\/p>\n\n\n\n![\"\"](\"https:\/\/espacerm.com\/webgen\/wp-content\/uploads\/2022\/02\/resultats-excel_01-1024x595.jpg\")
<\/figure>\n\n\n\n
\n\n\n\nConclusion<\/h2>\n\n\n\n