Over the years taken photos and my hard drive has accumulated a large number of images. Many of these images by their people, especially my kids. Even the “old days” I’ll scan my negatives or prints in order to have an electronic copy of the images I liked really. Tourism forward 20 years and I am already in my digital camera and seventh continues to add to the pile. Worse, it seems that I’m collecting images from many other electronic devices such as smart phones and my webcam video new-now pretty decent, and images as well. As a result, I have been using Google’s Picasa photo management software to better assist my efforts to organize, locate, tag them all.

Until recently, I haven’t really touched on many jobs face Picasa. To my surprise, I have found some interesting things about it and saw how truly powerful program has become. The larger part of the program works very well, and can help you identify persons quickly and provides compile your images differently or quickly find people you are looking for. Keep in mind, we are talking about face, not only in the face detection. Face detection to answer the question “is this face?,” instead of “facing this?”

To understand why the program might not be able to identify correctly on some ways, we need to understand how does Picasa. In short, let’s consider how the program algorithm performs the face. Most recognition algorithms will create vasibrint derived from measurements of distances between the main features of the face. This Vasibrint, while usually is not unique as a fingerprint, and is getting more accurate in the past few years. Having taken more advanced calculations that can even calculate the slope or angle and direction of the face may be referenced. The athletes are corrected mainly by rotation of the eye, then levels to uniform size tables for comparison with other objects within the database. Even with this ability, it seems Picasa production false positives and there is more room for error.

You can help improve identification, and in many cases, already have found advisable to “ignore” and not a sign of some faces are detected, but is not recognized.

Here are some specific examples where found cases in the ability to identify accurately:

And babies-possibly from the face, chobinis soft skin or the fact that their eyes may not always be open, but it seems that many images of early child is some trouble being recognized and lead to incorrect suggestions for other people. By the time the baby is about 2 months this doesn’t seem to be some issue.

Glasses-this also cause a problem where eye measurements cannot be calculated properly thus breaks the ability to recognize facial.

Partially masked faces-I’ve noticed no identification accuracy with long hair which has covered all or parts of the eye. Similarly, if your hand or other object in front of the face which seems to happen often with photos of children, the person who will not be able to recognize.

Light or low-resolution, low or out of focus-down many times get confused and focus light. Features of least distinguishable and measurements made.

Eyes closed-the unfortunate luck to people taking pictures only in the moment when flashing. It seems that prevent performance of face detection algorithms also recognition accuracy.

Multiple faces-Picasa combines two faces in what is believed to be one. This often occurs when one of the last and look sideways. I avoid identification and tagging these pictures because there is no way, at least until now, divided along or select border of faces.

Toys and dolls and statues-Peksa very good at picking up in even if it was not a real person. I usually just ignore all. On rare occasions, I have seen Picasa even interpretation elements light and shadows to objects in the picture as he even if no one in the picture.

Object rotate-even with advanced recognition, often choose to ignore the faces which have much skew or rotation. Usually people angle in more than 30 degrees off camera trying to ignore. It seems as italic decline should not be recognized as it tends to top face if obscure eyelashes eyes.

There are exceptions where try selecting and if people tag recognized, particularly if they are positioned directly forward looking. These people with beards or wearing glasses-Picasa do very good work in the detection and recognition anyway, but I think the deviation in the face of printing may be worth it, especially if the person who often change “Accessories”.

Also, be sure to check carefully from individuals who have already been identified and not necessarily need to confirm. Having found the brothers and sisters tend to get misused folder individuals more often then you think.

The only thing it can be sure to protect your information. Most people tag, and organize these images for personal use, but with the ability to easily synchronize images and information online sign in Picasa, it is important to protect information vasibrint. Publicly sharing vasibrint can make it easy for anyone to associate with the name for which can be used for malicious activities. Google allows you to control the visibility settings on whether this information is publicly available.