# Can we use image instead of video in createCapture() to add Images in dataset

**URL:** https://discourse.processing.org/t/can-we-use-image-instead-of-video-in-createcapture-to-add-images-in-dataset/7515
**Category:** Coding Questions
**Created:** [January 14, 2019, 6:03pm UTC](https://discourse.processing.org/t/can-we-use-image-instead-of-video-in-createcapture-to-add-images-in-dataset/7515 "2019-01-14T18:03:23Z")
**Posts on this page:** 1
**Page:** 1

<div class="post-metadata">

### Author: ![ankit](https://avatars.discourse-cdn.com/v4/letter/a/cab0a1/32.png) [@ankit](https://discourse.processing.org/u/ankit)
#### Post date: [January 14, 2019, 6:03pm UTC](https://discourse.processing.org/t/can-we-use-image-instead-of-video-in-createcapture-to-add-images-in-dataset/7515/1 "2019-01-14T18:03:23Z")

</div>

Hi,

We are using FeatureExtraction to get classification and regression since it using `video = createCapture(VIDEO);` so can we use images in place of video to add our images in the dataset and also want to train that image.

I’m using [https://ml5js.org/docs/custom-classifier](https://ml5js.org/docs/custom-classifier) and [https://ml5js.org/docs/custom-regression](https://ml5js.org/docs/custom-regression) for reference.

Thanks