Light-emitting diodes text dot-matrix text (LED text) is being widely used for displaying information and announcements. LED display for modernization of society and catch on for its versatile application with many benefits. Existing paper used k-nearest neighbor(k-NN) approach, low computation complexity method for pattern recognition, is used to recognition character component as any class of character and canny edge was used to detect character pixels when appear in led display area from scene images. The drawback of existing system is that it cannot handle text line with non-uniform color and containing less than 3 characters. It also cannot detect continuous LED text. Our proposed system will utilize the probabilistic neural network (PNN) classification to add the robust classification for the higher level of the adaptiveness. Principal component analysis (PCA) is a statistical procedure that uses an orthogonal transformation to convert a set of observations of possibly correlated variables into a set of values of linearly uncorrelated variables called principal components. Our proposed system will achieve better detection and recognition rate than existing system.