Multiscale amplitude modulation – frequency modulation (AM-FM) methods used to identify the difference between normal and abnormal retinal image. The summation of instantaneous amplitude, instantaneous frequency magnitude, and relative instantaneous frequency angle from multiple scales are used as texture feature vectors.It included the similarity span difference between the derived feature vectors to computed inter-structure.The output is presentation of data of normal retinal structure and abnormal region in the tissue based on AM-FM method. So this method make increased usage of automatic diabetic Retinopathy screening.The new appearance based representation technique removes the hard part of direct feature segmentation technique used in current methods for detection of DR. This AM-FM methods used to differentiate the retinal structures. The main aim of this method is to carefully differentiate normal and abnormal retinal structure. It identify diabetic retinopathy in retinal images based on instantaneous amplitude and instantaneous frequency and a huge region under the receiver operating part.An image can be estimated by a addition of AM-FM parts. Firstly, taking out of AM-FM parts from each image scale is done. Then the local frequency quantity is estimated by instantaneous frequency. In terms of cycles per millimetre, the instantaneous frequency magnitude is free from any image rotations or retinal imaging hardware quality, because it gives the original image structure, taken out from each image scale. An alternative to instantaneous frequency angle we can use relative angles. Relative angles are calculated by difference from higher surrounding area angle. AM-FM parts are obtained from various image scales. //To be changeAt lower frequency scales, the magnitude values of the IF are small and the extracted AM-FM features reflect slowly varying image texture. By using equiripple dyadic finite-impulse response filter design, bandpass filters is designed. The passband ripple of 0.001 dB and stop band ripple of 0.0005 dB. AM-FM calculate can be increased by using AM-FM demodulation algorithms by use of equiripple filter bank and a variable spacing linear phase approx. //TO be changeThen, at each pixel, for each combination of scales, uses DCA to select the AM-FM features from the bandpass filter that gave the maximum IA estimate. By using cumulative distribution functions of instantaneous amplitude, instantaneous frequency and the relative angle we can differentiate the retinal structure. Since the scale of values changes based on Cos used, the histograms are calculated from the global minimum value to the global maximum value. The magnitude of instantaneous frequency is inconsiderate to the way of image focus variations. The magnitude of instantaneous frequency is against to the steadily changing the brightness difference in the instantaneous amplitude. So, in a lighter background having a single dark round structure will have same instantaneous frequency distribution as in dark region having same size bright round structure. This show the exudates which is bright tissue damage and MAs which is dark tissue damage. By calculating the instantaneous frequency we can easily distinguish between two region. First region has a single vessel and second region has multiple narrow vessels. The frequency values of both the region will be same but the count on the histogram in the damaged region will be higher. So the higher count of histogram show the higher number of pixels. In neovasucularization region, the histogram have higher frequency distribution peak curve than normal vessel region. .