The implement algorithm of adaptive fuzzy image enhancement is proposed in ridgelet transform domain. Making use of the unique superiority for representing the linear singular border, the ridgelet transform can reach the aim of outstanding the border and restraining the noise. Applying the Fourier projection theorem in the frequency domain, the method is proposed for optimizing the ordering of the finite Radon transform coefficients. As can be seen, the “wrap around” effect is reduced in the image disposed by the Radon transform. The algorithm of adaptive fuzzy image enhancement is put forward based on the generalized fuzzy set and the maximum fuzzy entropy. The processed image is the better compromise between enhancing the characteristics and inhibiting the noise. The experiment shows that the edge detection probability of this algorithm is greater than the traditional ones' with approximate 50% under the condition of low signal-noise ratio (2.5～5.5 dB). Applying the algorithm to enhance the linear cracks in the road surface image, we can enhance the crack signals and restrain the discrete noise, such as oil droplet , cobblestone.