v4.7 - Bug fix.
The app can make object recognition just based on the taken picture by the camera from the device. This can be useful for children, blind people or if you have an object and you really don't know what it is. For blind people the tool has vocal assist too (text-to-speech), this can be activated or deactivated from the menu.
The objects where the recognition had successful result are logged in a list. This way the app can be used to scan items in a supermarket like bar code scanner apps, but it's much easier, because the user needs to take a photo of the product, not to search for its bar code and can be logged items without bar code (like fruits or vegetables).
The object recognition engine does not always work or sometimes makes recognition errors, many times due to the poor quality of the image. Try to focus the image by tapping on the photo area to create good quality images. Focus can be done easily if the phone is moved gradually closer to the object. Sometimes it helps if the object is viewed from close position and there are not too many objects in the background. Sometimes the recognition works better if the flashlight is used. You can always retry to recognize the object, there is a 3 second delay between taking photos, to prevent button spamming.
In order to make the app functional it is needed an internet connection (Mobile or WiFi).
To make the text-to-speech work on your phone, read the tutorial from the http://www.androidauthority.com/google-text-to-speech-engine-659528/ link, and set up Google Text-to-speech Engine with English (United States) language.
The app does collect user data, all the data is saved on the mobile device (taken pictures, and recognized object list), however the taken picture is uploaded on server to make object recognition, but it is deleted immediately after.
To reveal hidden menu long press on any list item.
