Steganography is a data-hiding scientific branch that aims to hide secret data in an image, video, or audio. Image steganography methods try to embed a large amount of data into images with high imperceptibility. However, increasing the number of embedded data in the image decreases its quality. Therefore, in this paper, a new method based on Least Significant Bit (LSB) using Hough Transform is proposed to improve the stego image quality with increasing the amount of embedded data. The LSB is the common embedding steganography method due to its simplicity of implementation and low complexity. The proposed method inverts the LSBs of image pixels to enhance the quality of stego image. First, improved edge detection filter is used to detect edges areas. Then, we invert LSBs for the pixel in edge area pixels. Finally, the LSBs smooth area pixels of the cover image are inverted. The performance of the presented method is evaluated for the stego image quality and the amount of embedded data. The results show that the new method has better performance in comparison with the current methods in terms of Peak Signal-to-Noise Ratio (PSNR) and capacity.