<p>This paper presents an enhanced spectral conjugate gradient algorithm designed for solving general unconstrained optimization problems. The proposed method builds upon a modified secant condition and incorporates a quasi-Newton-type search direction to improve both convergence behavior and computational efficiency. A novel spectral parameter is introduced, and a dual truncation strategy is employed to regulate both the spectral coefficient and the conjugate parameter, which ensures numerical stability and maintains robust descent properties. It is rigorously proven that the search direction generated by the algorithm satisfies the sufficient descent condition under any line search strategy. Moreover, the global convergence of the method is established under standard assumptions for general nonlinear objective functions. Extensive numerical experiments on benchmark problems demonstrate that the proposed NSHS algorithm outperforms several existing competing optimization methods in terms of convergence speed and solution quality. In addition, the algorithm is applied to image reconstruction tasks within the framework of compressive sensing. The experimental results demonstrate its effectiveness, competitive performance, and strong applicability in practical scenarios.</p>