This paper proposes a reversible watermarking algorithm for image content authentication based on wavelet transform. The algorithm aims to protect the privacy of image content and enable lossless recovery. The method involves transforming each block of the image using wavelet transform to serve as a carrier for embedding watermark information. Two watermarks are encrypted and scrambled using chaotic mapping with different keys, enhancing security and randomness. Tamper detection uses multiple levels of detection and screening to improve accuracy. When restoring the tampered area, the two watermarks are used to reconstruct the category and feature information of the image block. Experimental results show that the algorithm can accurately detect and locate tampered areas and restore the original image information with high quality after general tampering, constant mean tampering, and paste tampering. Image content authentication is a technology that verifies the authenticity of image content, which has many advantages in protecting the integrity and authenticity of digital images. It originated from data authentication technology in cryptography and is widely involved in communication and cryptography theories. Image content authentication has three main tasks: detecting forgery, locating tampering, and restoring the original image. There are two main technologies for image content authentication: digital signature technology based on cryptography and digital steganography technology based on information hiding methods. Digital watermarking combines the characteristics of cryptography and steganography, offering low overhead and high concealment, which can compensate for the shortcomings of cryptography and help locate tampering. The proposed algorithm divides the image into blocks, each with four pixels. The blocks are transformed into the frequency domain using wavelet transform. The authentication watermark is obtained by averaging the four higher bits of the four pixel values in the block. The correlation features and pixel distribution features between the pixels in each block are selected to generate the feature watermark. After watermark generation, the watermarks are encrypted using chaotic mapping with keys. The authentication detection is achieved by comparing the original authentication watermark with the extracted authentication watermark. Finally, the tamper locating and approximate recovery of the tampered area are achieved by using the feature and authentication watermarks.This paper proposes a reversible watermarking algorithm for image content authentication based on wavelet transform. The algorithm aims to protect the privacy of image content and enable lossless recovery. The method involves transforming each block of the image using wavelet transform to serve as a carrier for embedding watermark information. Two watermarks are encrypted and scrambled using chaotic mapping with different keys, enhancing security and randomness. Tamper detection uses multiple levels of detection and screening to improve accuracy. When restoring the tampered area, the two watermarks are used to reconstruct the category and feature information of the image block. Experimental results show that the algorithm can accurately detect and locate tampered areas and restore the original image information with high quality after general tampering, constant mean tampering, and paste tampering. Image content authentication is a technology that verifies the authenticity of image content, which has many advantages in protecting the integrity and authenticity of digital images. It originated from data authentication technology in cryptography and is widely involved in communication and cryptography theories. Image content authentication has three main tasks: detecting forgery, locating tampering, and restoring the original image. There are two main technologies for image content authentication: digital signature technology based on cryptography and digital steganography technology based on information hiding methods. Digital watermarking combines the characteristics of cryptography and steganography, offering low overhead and high concealment, which can compensate for the shortcomings of cryptography and help locate tampering. The proposed algorithm divides the image into blocks, each with four pixels. The blocks are transformed into the frequency domain using wavelet transform. The authentication watermark is obtained by averaging the four higher bits of the four pixel values in the block. The correlation features and pixel distribution features between the pixels in each block are selected to generate the feature watermark. After watermark generation, the watermarks are encrypted using chaotic mapping with keys. The authentication detection is achieved by comparing the original authentication watermark with the extracted authentication watermark. Finally, the tamper locating and approximate recovery of the tampered area are achieved by using the feature and authentication watermarks.