The single injection for an ICE can be considered not stationary but a sequence of several shots, at the same boundary conditions, can be evaluated as a quasi-stationary or more precisely “pseudo stationary” event. This means that on average the spray will have the same behavior along the considered time window. In this paper, these considerations are applied to sprays for GDI engines. The behavior of the spray is analyzed recording a 3D matrix defined by the x, y, t dimensions (pixel, pixel, time) with the main objective of defining the parameters like cone angle of the single jet, angles between jets, penetration length. These data permit to build a frequency and a morphologic map able to define a univocal “signature” of the device. The analysis has been performed on a large amount of image sequences, keeping constant the injection conditions: pressure, backpressure, gas temperature, fuel and nozzle temperatures and the energizing time. The used fluid was a mono-component fuel (iso-octane). The paper reports an innovative method of analysis based on an advanced statistical techniques applied to images captured by a high-speed camera that allows highlighting phenomena and anomalies hardly detectable by conventional optical diagnostic techniques. The images, previously elaborated by neural network tools in order for clearly identifying the contours, have been analyzed in their time evolution as pseudo-chaotic variables that may have internal periodic components. In addition to the Fourier analysis, tools as Lyapunov and Hurst exponents and average Kω, has permitted to detect the chaos level of the signals. The use of this technique has hallowed to distinguish periodic oscillations from chaotic variations and to detect those parameters that actually determine the spray behavior.