Kanji-Reader 1.3

Kanji Reader utilizes cutting-edge convolutional neural networks technology to develop an image processing application that analyzes photos of individual Japanese characters, providing their readings, translations, and additional details.

Simply point the camera at a character to observe the real-time display of results. An internet connection is not necessary. It boasts a recognition accuracy of over 90% for more than 2000 Kanji characters.

Toggle between white and black text detection modes. Adjust the camera sensitivity to capture fainter text or reduce sensitivity in high-contrast environments.

Explore additional results generated by the model, with the option to view up to 10 entries ordered by probability.

Access in-depth information about each Kanji, including the "On" and "Kun" readings, radical, and stroke order.