Abstract
Purpose: :
To investigate the recognition of common objects in daily life based on simulated phosphene arrays in visual prosthesis. In this study, the influences of four parameters (grey scale, contrast, dropout percentage, distortion) on objects recognition were investigated.
Methods: :
Thirty volunteers from 21 to 30 years old with normal or corrected to normal visual acuity of 20/20 were recruited. The trial apparatus included a PC computer with self-developed program, a self-designed light shielding device and a 17-inch CRT screen. The experimental materials were objects in common use among blind persons based on the survey from the local blind school. Each object was formed by 25×25 pixel array consisted of Gaussian dots according to the pre-experiment results. For each parameter, there were different levels to be tested. Whatever grey scale, contrast, dropout percentage, distortion changed objects were rendered on the screen one by one against black background. In order to reduce the effects of familiarity and adaptability, the testing sequence is random.
Results: :
According to the experimental results, it showed that the recognition decreased considerably with the increase of dropout percentage and distorted level. On the contrary, the recognition increased with the grey scale increase. Furthermore, the contrast change didn’t show obvious effect on the recognition. Under the same condition, objects with observable shape features have high recognition accuracy.
Conclusions: :
The results indicated that grey scale, dropout percentage and distortion impacted the objects recognition obviously. The individual contrast change has not remarkable effect on the recognition. The various shape features of different objects were the one of the important factors in the recognition.
Keywords: space and scene perception • pattern vision • image processing