Mirror/AI-on-the-edge-device-docs
1
0
Fork 0
mirror of https://github.com/jomjol/AI-on-the-edge-device-docs.git synced 2025-12-09 21:17:09 +03:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #8 from haverland/main

Browse Source 
#1808 more about collecting images
This commit is contained in:
jomjol
2023-01-13 21:24:27 +01:00
committed by GitHub
parent 9c3264d8e4 e18895ad30
commit 3b0985f7c2
1 changed files with 12 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
docs/Learn-models-with-your-own-images.md
View File

@@ -1,7 +1,16 @@
# Learn a model with your own images
If your device has new, different digits and the existing models don't recognize them well, you can collect your own images and train the model.
But before you do this, please check if your type really is not contained yet in the training data, see [digits](https://jomjol.github.io/neural-network-digital-counter-readout) resp. [pointers](https://jomjol.github.io/neural-network-analog-needle-readout/) for an overview of images used for the training
If your device has new, different digits and the existing models don't recognize them well, you can collect your own images and train the model. Of course, this also applies if you have a model that is already known, but the neural models do not produce good results.
But before you do this, you please check if your type really is not contained yet in the training data, see [digits](https://jomjol.github.io/neural-network-digital-counter-readout) resp. [pointers](https://jomjol.github.io/neural-network-analog-needle-readout/) for an overview of images used for the training. But we also like to take pictures of already known devices.
## Before you start
Poor recognition is often caused by blurred images, low contrast or incorrect setting of the ROIs. Therefore, check these possibilities first, as additional training will bring little improvement here.
You must not train the models by yourself (see section [Share your images](#share-your-images)). In most cases we will integrate you images in the training dataset of the models. Only if we fear a degradation of the models or you need a different behaviour, we will not include the data in the standard models.
## Collecting images
The neural network is trained on base of a set of images, that have been collected over time. If your digits are included or at least very similar to included images, the chance is very high that the neural network is working fine for you as well.
@@ -11,7 +20,7 @@ In order to incorporate new digits a training set of images is required. The tra
![](img/enable_log_image.jpg)
Now wait, until you have an image of each digit of every type on the SD card. Ideally remove the SD card from the camera and search for two to three images of each digit (**not more! :-)**). The format can be jpg.
Now wait, until you have an image of each digit of every type on the SD card. Ideally remove the SD card from the camera and search for two to three images of each digit (**not more! :-)**). The format can be jpg. But don't panic, the collector programmes will sort out duplicate images for you.
## Collecting images for dig-class100/dig-cont/ana-class100