{"id":1668,"date":"2026-06-10T15:55:57","date_gmt":"2026-06-10T15:55:57","guid":{"rendered":"https:\/\/iludis.de\/?p=1668"},"modified":"2026-10-18T17:36:56","modified_gmt":"2026-10-18T17:36:56","slug":"cnn-grayscale-trainer","status":"publish","type":"post","link":"https:\/\/iludis.de\/?p=1668","title":{"rendered":"CNN: Theorie und Anwendung"},"content":{"rendered":"\n<div class=\"wp-block-columns is-layout-flex wp-container-core-columns-is-layout-995f960e wp-block-columns-is-layout-flex\">\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-group has-palette-color-8-background-color has-background is-layout-constrained wp-block-group-is-layout-constrained\">\n<h4>CNN Training interaktiv im Browser<\/h4>\n<p><a href=\"https:\/\/iludis.de\/CNN_sim\/CNN_sim.html\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-1669\" src=\"https:\/\/iludis.de\/wp-content\/uploads\/2026\/06\/Neues-Projekt.png\" alt=\"\" width=\"401\" height=\"324\" srcset=\"https:\/\/iludis.de\/wp-content\/uploads\/2026\/06\/Neues-Projekt.png 401w, https:\/\/iludis.de\/wp-content\/uploads\/2026\/06\/Neues-Projekt-300x242.png 300w\" sizes=\"auto, (max-width: 401px) 100vw, 401px\" \/><\/a><\/p>\n<p><a href=\"https:\/\/iludis.de\/CNN_sim\/CNN_sim.html\">https:\/\/iludis.de\/CNN_sim\/CNN_sim.html<\/a><\/p>\n<p>Datens\u00e4tze:<\/p>\n<ul>\n<li><a href=\"https:\/\/iludis.de\/CNN_sim\/digits_1_3_5_7.csv\">https:\/\/iludis.de\/CNN_sim\/digits_1_3_5_7.csv<\/a><\/li>\n<li><a href=\"https:\/\/iludis.de\/CNN_sim\/digits_2_4_6_8.csv\">https:\/\/iludis.de\/CNN_sim\/digits_2_4_6_8.csv<\/a><\/li>\n<li><a href=\"https:\/\/iludis.de\/CNN_sim\/mnist_train_800.csv\">https:\/\/iludis.de\/CNN_sim\/mnist_train_800.csv<\/a><\/li>\n<\/ul>\n<hr \/>\n<p><strong>CNN-Theorie<br \/><\/strong><a href=\"https:\/\/iludis.de\/Tools\/cnn-labor-Teil1_faltung.html\">https:\/\/iludis.de\/Tools\/cnn-labor-Teil1_faltung.html<\/a><br \/>Grundlagen des Faltungskernels<\/p>\n<p><a href=\"https:\/\/iludis.de\/Tools\/cnn-labor-Teil2_pipeline.html\">https:\/\/iludis.de\/Tools\/cnn-labor-Teil2_pipeline.html<\/a><br \/>Visualisierung der CNN-Pipeline<\/p>\n<p><a href=\"https:\/\/iludis.de\/CNN_sim\/conv2d-parameter-explorer.html\">Conv2D-Parameter-Explorer<\/a><br \/>Wie setzen sich CNN-Netze zusammen?<\/p>\n<p><a href=\"https:\/\/iludis.de\/CNN_sim\/parameter-rechner.html\">Berechnung trainierbare Parameter<\/a><br \/>Interaktive Simulation der Parameteranzahl f\u00fcr CNN-Netze<\/p>\n<p><a href=\"https:\/\/iludis.de\/CNN_sim\/CNN_rezFeld.html\">Visualisierung Rezeptives Feld<\/a><br \/>Was sieht ein CNN-Netz und wie berechnet man das?<\/p>\n<\/div>\n<\/div>\n\n\n\n<div class=\"wp-block-column is-layout-flow wp-block-column-is-layout-flow\" style=\"flex-basis:50%\">\n<div class=\"wp-block-group has-palette-color-8-background-color has-background is-layout-constrained wp-block-group-is-layout-constrained\">\n<h4>CNN f\u00fcr die Mittelstufe<\/h4>\n<p><strong><a href=\"https:\/\/iludis.de\/wp-content\/uploads\/2023\/05\/Conv2D_UIdee.pdf\">Unterrichtsidee<\/a><\/strong><br \/>Erprobt im Hector-Unterricht Kurs &#8218;ML f\u00fcr Mikrocontroller&#8216; und Inf-AG.<\/p>\n<p><strong>1. Bitmap erstellen<br \/><\/strong>Klicke auf die schwarze Fl\u00e4che und erstelle ein Bitmap<br \/>in Form einer 1D-Liste. Diese kann in die Zwischenablage<br \/>kopiert werden.<\/p>\n<p><a href=\"https:\/\/iludis.de\/bitmapMaker\/index.html\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-1336\" src=\"https:\/\/iludis.de\/wp-content\/uploads\/2023\/05\/bitmapM.png\" alt=\"\" width=\"400\" height=\"279\" srcset=\"https:\/\/iludis.de\/wp-content\/uploads\/2023\/05\/bitmapM.png 400w, https:\/\/iludis.de\/wp-content\/uploads\/2023\/05\/bitmapM-300x209.png 300w\" sizes=\"auto, (max-width: 400px) 100vw, 400px\" \/><\/a><\/p>\n<p><strong>2. Jupyter-Notebook f\u00fcr Bitmaps aus 1D-Arrays<br \/><\/strong>Aus den obigen Bitmaps mit numpy und OpenCV<br \/>einfach Bitmaps erstellen:<br \/><a href=\"https:\/\/iludis.de\/bitmapMaker\/Simple_Bitmaps.ipynb\">https:\/\/iludis.de\/bitmapMaker\/Simple_Bitmaps.ipynb<\/a><\/p>\n<p><strong>3. Grayscale-Bitmap im MNIST-Format erstellen.<br \/><\/strong>Erstelle ein 255-Graustufen-Bild mit 28&#215;28 Pixeln,<br \/>exportiere dessen Werte per Knopfdruck in die<br \/>Zwischenablage und f\u00fcge sie ins Conv2D-Applet ein:<\/p>\n<p><a href=\"https:\/\/iludis.de\/GrayscaleMaker\/index.html\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-1342\" src=\"https:\/\/iludis.de\/wp-content\/uploads\/2023\/05\/grayscale.png\" alt=\"\" width=\"400\" height=\"280\" srcset=\"https:\/\/iludis.de\/wp-content\/uploads\/2023\/05\/grayscale.png 400w, https:\/\/iludis.de\/wp-content\/uploads\/2023\/05\/grayscale-300x210.png 300w\" sizes=\"auto, (max-width: 400px) 100vw, 400px\" \/><\/a><\/p>\n<p><strong>4. Optimale Convolution-Kernel entwickeln.<br \/><\/strong>Wie in einem CNN: Teste verschiedene Convolution<br \/>Kernel aus, berechne ein MaxPooling(2,2) und finde<br \/>heraus, welche Pixelmuster man so bestimmen kann.<\/p>\n<p><a href=\"https:\/\/iludis.de\/konvolution\/index.html\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-1337\" src=\"https:\/\/iludis.de\/wp-content\/uploads\/2023\/05\/conv2d.png\" alt=\"\" width=\"400\" height=\"279\" srcset=\"https:\/\/iludis.de\/wp-content\/uploads\/2023\/05\/conv2d.png 400w, https:\/\/iludis.de\/wp-content\/uploads\/2023\/05\/conv2d-300x209.png 300w\" sizes=\"auto, (max-width: 400px) 100vw, 400px\" \/><\/a><\/p>\n<\/div>\n<\/div>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>CNN Training interaktiv im Browser https:\/\/iludis.de\/CNN_sim\/CNN_sim.html Datens\u00e4tze: https:\/\/iludis.de\/CNN_sim\/digits_1_3_5_7.csv https:\/\/iludis.de\/CNN_sim\/digits_2_4_6_8.csv https:\/\/iludis.de\/CNN_sim\/mnist_train_800.csv CNN-Theoriehttps:\/\/iludis.de\/Tools\/cnn-labor-Teil1_faltung.htmlGrundlagen des Faltungskernels https:\/\/iludis.de\/Tools\/cnn-labor-Teil2_pipeline.htmlVisualisierung der CNN-Pipeline Conv2D-Parameter-ExplorerWie setzen sich CNN-Netze zusammen? Berechnung trainierbare ParameterInteraktive Simulation der Parameteranzahl f\u00fcr CNN-Netze Visualisierung Rezeptives FeldWas sieht ein CNN-Netz und wie berechnet man das? CNN f\u00fcr die Mittelstufe UnterrichtsideeErprobt im Hector-Unterricht Kurs &#8218;ML f\u00fcr Mikrocontroller&#8216; und Inf-AG. 1. Bitmap [&hellip;]<\/p>\n","protected":false},"author":2,"featured_media":1669,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[],"class_list":["post-1668","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ki-dqr5"],"_links":{"self":[{"href":"https:\/\/iludis.de\/index.php?rest_route=\/wp\/v2\/posts\/1668","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/iludis.de\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/iludis.de\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/iludis.de\/index.php?rest_route=\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/iludis.de\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1668"}],"version-history":[{"count":7,"href":"https:\/\/iludis.de\/index.php?rest_route=\/wp\/v2\/posts\/1668\/revisions"}],"predecessor-version":[{"id":2025,"href":"https:\/\/iludis.de\/index.php?rest_route=\/wp\/v2\/posts\/1668\/revisions\/2025"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/iludis.de\/index.php?rest_route=\/wp\/v2\/media\/1669"}],"wp:attachment":[{"href":"https:\/\/iludis.de\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1668"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/iludis.de\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=1668"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/iludis.de\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=1668"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}