Improve image browsing and interface usability + automatic GCP detection

[HeDo88TH]

This PR adds the following features:

• EXIF tag reading: first show the images that are closest to the GCP
• Image retention: no need to re-upload the images for each GCP
• Automatic GCP detection: opencv haar classifier to find GCPs in images

[pierotofy]

Closes #10

[HeDo88TH]

This is ready 🚀 @pierotofy

[pierotofy]

Like the GCP autodetect button. A few improvements are needed:

• It's not clear to the user that images are being sorted by distance to the GCP. There needs to be a button that toggles on/off the ability to display images that are outside of a certain radius. The workflow should be: I drag & drop images, the ones that are outside of a (user specified) radius are not displayed (you can save tons of DOM rendering time). The idea is that I should be able to drag & drop thousands of images and display only the few that are within a radius of the GCP. Sorting them does help me distinguish where one GCP ends and another one begins! (The browser also crashes if I drop thousands of images).
• I should not need to press "Auto Detect" on each image. A global button should try to detect all GCPs for all images I've dropped.
• A text status along with the spinning gear icon would help end users understand what's going on (e.g. "Reading image %(num) of %(total)", "Detecting GCP in %(filename)", etc.)

[pierotofy]

Can we also include in the README a note on how to generate/improve the haar model?

[HeDo88TH]

Added radius filter button and progress dialog @pierotofy I'm adding the readme info soon

[HeDo88TH]

Done @pierotofy 🚀

[pierotofy]

The text status for "Reading coordinates" is always stuck at 100%?

• Create a new GCP file
• Load some images
• Progress indicator is stuck at 100%

[HeDo88TH]

Everything fixed, plus I added pagination to prevent crashes 🚀 @pierotofy