With the annual increase in agricultural output and the rising cost of labor for picking, fruit and vegetable picking robots gradually become a trend to replace manual labor. This paper proposes an improved target recognition algorithm based on the Hough Circle Transform and a distance estimation method using the L-component of the Lab color space to address the issues of recognition accuracy and depth information acquisition in complex environments for fruit and vegetable picking robots. We select a variety of fruit images, including apples, oranges, and red bayberry, for our experiments. The experimental results demonstrate that the accuracy of the improved target recognition algorithm is significantly higher compared to the traditional method, and the estimated target distance closely matches the actual distance.