Hard and Brittle Materials (HBMs) such as diamond, sapphire, and quartz glass are employed in many applications within the aerospace, military, integrated circuits, biomedicine, and optoelectronic devices sectors. This is due to their high hardness, high thermal stability, chemical inertness, and excellent optoelectronic properties. In particular, hard and brittle materials play an indispensable role in the extreme environment of strong radiation and easy corrosion. Moreover, due to its high hardness and chemical inertness, employing traditional processing techniques to fabricate micro-nano structures is challenging. The pulse duration of femtosecond lasers is significantly shorter than the time scale of electron-lattice interaction, which enables the preparation of micro-nano-structures in a nearly heat-affected region. The existence of the femtosecond Laser Damage Thresholds (LDT) effect enables the spatial resolution of laser processing of hard and brittle materials to be greatly enhanced. Consequently, the study of the LDT of materials is of great significance, as it allows the spatial resolution of femtosecond laser processed materials to be enhanced and provides a reference for the processing scheme.There are three principal methods for obtaining LDT the direct measurement method, the model calculation method and the extrapolation method. The direct measurement method is well-suited to initial material LDT testing due to its simplicity of operation and the ability to obtain data quickly. However, the LDT of the material is related to the parameters of the laser, such as wavelength, pulse width, repetition frequency and the type of focusing objective of the laser process, which requires a significant amount of measurement work. Furthermore, the presence of local defects or inhomogeneities in the material can lead to early damage, introducing uncertainty into the results of experiments conducted using the direct measurement method. Consequently, a significant number of experiments are required to verify the results. Additionally, the discontinuity of processing parameter settings necessitates a substantial number of experimental determinations to accurately obtain the LDT. The model calculation method, which predicts the LDT by computer simulation of the laser-material interaction process, circumvents the necessity for extensive experimental validation, provides precise prediction of the LDT and elucidates the underlying physical mechanism. However, the mechanism of femtosecond laser-material interaction is complex, and it is challenging to develop an accurate physical model. Furthermore, the intricacy of the physical model renders it costly to compute, while its high demand for parameter precision renders it more challenging to align with the actual processing environment. Extrapolation is a method of deriving the LDT by discovering the relationship between laser processing parameters and damage morphology. One of the most commonly used methods is the method of extrapolating the LDT from the diameter of the damage. The extrapolation method combines the advantages of the direct measurement method, which avoids the problem of the existence of an early ablation area of the material by measuring the diameter of the damage area with a certain area, with the simplicity of operation of the extrapolation method. This method then uses the relationship between the processing parameters and the diameter of the damage area to predict the LDT of the material, which avoids a large number of experiments and reduces the cost of measurement. Nevertheless, in the event of manual measurement, there is a risk of inconsistency in measurement standards and manual reading errors, which may result in a reduction in the efficiency, reliability and comparability of threshold results.In this paper, a novel methodology for the automatic measurement of femtosecond laser damage threshold is proposed. The proposed methodology is based on image recognition algorithm and is designed to meet the demand for laser damage threshold measurement on the surface of diamond and other materials. The methodology involves the use of microscopic images of the damage structure in grey scale, the application of a binary Fourier filter to denoise the images, and the automatic extraction of the damage region's boundary. This is achieved by means of an algorithm. The damage structure is then identified automatically, and its area is measured. The damage area and laser power logarithmic curve are then fitted, and the damage threshold value is extrapolated. Utilizing this methodology, the single-pulse damage threshold of the 50-fs femtosecond laser is determined to be 3.1 J/cm² for single-crystal diamond, and the single-pulse damage threshold of microcrystalline diamond is established as 0.55 J/cm² for microcrystalline diamond film. The reliability of the method was verified by the comparison of field emission scanning electron microscope, atomic force microscope and identification results. The method is deemed suitable for materials exhibiting uniform and stable surface morphology and laser damage features that differ from the original surface. The adoption of a uniform and controllable identification standard has been demonstrated to significantly reduce randomness and instability in the measurement process, thereby improving the credibility of the measurement results. Furthermore, the automation of the data processing process enhances measurement efficiency and reduces costs, providing a valuable reference framework for femtosecond laser processing of materials. The development of deep learning, real-time monitoring and high-speed imaging technologies is expected to lead to more intelligent recognition, more efficient real-time feedback and a wider range of material applicability, providing strong support for high-precision processing and quality control of femtosecond lasers in scientific research and industry.