Optical wavelet transform (WT) with joint transform correlator (JTC) was described. Haar wavelets and Roberts filters, respectively, were put together with an object image as a joint image. At the output plane of the JTC, the WT of the object was obtained, the features, such as corners, edges, sharpening images, of the object were extracted. These features will be useful in post-pattern recognition. A programmable optical joint wavelet transform system is proposed, which can be used to realize t rue WT of an object image. Digital simulation results are provided.