A set of ultrasonic echo signal detection system was constructed based on lowcost ultrasonic sensors, and a quantification test bench was established based on the cylindrical surface leaf blade distribution model. Based on the orthogonal center composite design experiment, a quantitative relationship among the mean value of ultrasonic echo signals, canopy density and detection distance was established, namely the plant canopy density quantitative model. The variance analysis of the established plant canopy density quantification model showed that the plant canopy density quantitative model was very significant, and the loss of quasisimulation was not significant. The R2 and predicted R2 by the plant canopy density quantification model were 0.9885 and 0.9114, respectively, indicating good agreement between the experimental values and the predicted values. In order to verify the reliability of the established plant canopy density quantitative model, four canopy densities were used in indoor to perform verification tests at three different distances. The experiment results showed that the minimum relative error between the measured value and the model measured value was 1.230%, the maximum relative error was 13.650%, and the average relative error was 6.120%. The plant canopy density quantitative model had a good applicability to the canopy density measurement of the indoor test bench. Three Osmanthus trees were selected to verify the model in the outdoor and nine test points were selected for each tree. The experiment results showed that the minimum relative error of the measured density and the model density was 3.959% and the maximum relative error was 20600%, and the average relative error between the measured density and the model measured density of three Osmanthus trees was 11.244%, 12.246% and 9.628%, respectively. Therefore, the plant canopy density quantitative model established was suitable for outdoor density measurement of Osmanthus trees.